MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

performer (or is it per|former?) sequencer build
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author performer (or is it per|former?) sequencer build
sduck
I'm working on this thing, and couldn't find any other build resources here or elsewhere, so thought I would start a thread here. Maybe this will be of use to someone.

More info - https://westlicht.github.io/performer/

PCBs and panels are available from various sources.

Here's where I am currently - posting this so that if anyone out there knows anything about this, maybe they'll spot if I got any of the chip orientations wrong -



Also, if anyone knows a US source for the PJ301CM stereo jacks for this?
djthopa
Im interested on this project too.
Where did you get the pcb?
Any source for a panel?
Mouser cart is
gbiz
I wondered when someone would do this. smile

This is a great sequencer. I've been wanting something like this for a while. I just built one up, have another one to do. Building this to me was like going back to the early days of building the Mutable modules. Use the source Luke. I can see a lot of people who can't work like that having a b0rked build. And as djthopa says, it's quite an expensive BOM.

The IC orientation is marked on the silk screen, though it's not obvious at first glance, or second for that matter. If you look at each device there's a white line on the silk above & below. One of the lines extends alongside an adjacent pad - either top left or bottom right. That pad is pin 1. That only became obvious to me after i opened kicad & looked at the layout wink If you look at the build guide, the images in that are sufficiently hi-resolution that you can detect the device markings, so if you have the same devices you might be able to use that as a double check.

My advice would be to avoid using 0805 2u2 caps & use 0603. That kicad footprint is more like 0605, & theres very little free pad to get an iron on both the device & pad. An 0603 device is a much better fit, provided you ensure it's located centrally on the footprint. I can see some builders struggling to get the 0805 device fitted, making a mistake, trying to correct it & ripping up the power rail to the device it's bypassing - the MCU or DAC.

The rest of the components are reasonably straightforward. Theres a few passives where the reference designator isn't adjacent to the device but they're pretty obvious which is which.

The rest of my suggestions are really down to the panel components. I used a 2mm panel, as i usually would. As this uses PJ301BM's & not Thonkiconns, I'd suggest anyone else getting Alu panels fabbed to go with 1.6mm, that extra -0.4mm will help.

This one is a personal thing. Some will disagree. Flame away. To avoid PCB flex, the encoder needs to be fixed to the panel, not the PCB (ie mounted flush). With the encoder in the BOM, as it has a relatively short bushing, it means the encoder will be mounted probably 5mm off the PCB. I used a split 6mm M6 nylon washer behind the encoder to keep it perpendicular to the panel. 5mm would probably be better, i didn't have one. Like this, the solder tabs on the device will just about fit in the PCB. I looks ugly, but it won't crack the PCB traces after prolonged use. I've got an image of mine if people want that posted here. (Go look back in the original Braids build for discussion on this. It got discussed at length there).

The stereo PJ301 only have 4mm bushing, which means there's very little thread exposed above the panel to get a nut on (you'll see this in the images in the build guide etc). On mine i trimmed the plastic back at the base of the bushing on the socket, & mounted the jack to the panel, again, like the encoder so the device is flush to the panel & slightly raised above the PCB. Another option is to leave the nuts off on these two, but as they're the nearest bracing for the PCB at the USB socket & sd-card, they're really needed. (Stereo Thonkiconns don't have this reduced bushing size. Sigh).

Again, with the extra height with the PJ301's, there isn't a lot of switch cap exposed (not helped by the 2mm panel). So when you fit the switch caps, don't push them fully home.

As for loading the initial code. I'm not someone who uses pre-compiled binary, compilation is part of the module build. It took a while to get the code to build but i got it done eventually. Something i didn't see anywhere aside from in the code is the address to load the sequencer binary to. The bootloader goes to 0x8000000, as usual. The sequencer starts at 0x8010000. (The code compilation generates a standalone sequencer that you can load at 0x8000000, but then you lose the bootloader for future upgrades via the sd-card).
There's only pre-compiled hex files available. If you need a binary, something like hex2bin will convert them for you. I tried this using the hex2bin i provide with dervish, & it works.

I'll edit this if i think of anything else ...
gbiz
djthopa if you're in the UK ping me an email, i've got a PCB, 2mm Alu panel & acrylic window going spare. It'll be at the cost to me + a donation to the designer.

edit: sold
gbiz
Anyone who's going to get panels done at PCBWay, when you add in the notes for them not to add the order number on the top silk, also ask them to put a piece of paper between each board before shipping (like PCBCart do without you having to ask them). It'll help prevent them getting scratched.

The first batch of panels they sent me, the engineer decided to edit the silk text & fill in the enclosed space on the characters. They agreed to send me replacements FoC.

gbiz
I went for a mix of light & dark grey & one red pushbutton caps. All white would be a bit too clean looking for me. I originally planned using light grey for the function buttons, but Mouser were out of stock with those. Now it's built, i think dark works.

This is the PCBWay "enhanced" graphics panel in it's full glory

[edit] now with the window. Annoyingly there's a bit of dust on the window behind the panel. It'll have to stay there for now

sduck
Thanks gbiz! Lots of great info. I noticed those markings on the pcb for the chips, but it was unclear what they meant (to me at least) - I went with the detailed pics in the build guide.

djthopa, I got the pcb from Michigan Synth Works, and the panel from Front Panel Express. There are other sources - I think ZZ Ardoz has them here, there might be others. You can get the gerbers from github and have your own made also. I don't really know the total BOM cost, I had a bunch of other things in my cart when I threw this stuff in there - it would have been somewhere between 100-200$US.

Oh, and I found those stereo jacks on ebay - https://www.ebay.com/itm/25-Pack-PJ-301CM-3-5mm-Stereo-Audio-Jack-with -Knurled-Nut-USA-Seller/232590960218
Dunk_91
Using it since a week and I love it!! So much features and so easy to use, really a pleasure to patch and program! As said build is pretty straight, components are well spaced, just take your time and go on nanners
Pay attention when soldering encoder, I've soldered mine flush to the pcb and now I can't bolt it to the panel. It's fine like that but next time I'll bolt it and then solder grin
OIP
thanks for posting, this has been on my list for a while now and was going to fab some PCBs but the BOM cost put me off. i'll get to it eventually. looks like a great sequencer and the work put in by the designer is awesome.

really curious about the MIDI out possibilities and using it to sequence external gear along with euro too.
djthopa
gbiz wrote:
djthopa if you're in the UK ping me an email, i've got a PCB, 2mm Alu panel & acrylic window going spare. It'll be at the cost to me + a donation to the designer.


Hi! I just saw your reply in here. Have sent you a pm!

Thanks for all the detailed information Guinness ftw!
djthopa
sduck wrote:
Thanks gbiz! Lots of great info. I noticed those markings on the pcb for the chips, but it was unclear what they meant (to me at least) - I went with the detailed pics in the build guide.

djthopa, I got the pcb from Michigan Synth Works, and the panel from Front Panel Express. There are other sources - I think ZZ Ardoz has them here, there might be others. You can get the gerbers from github and have your own made also. I don't really know the total BOM cost, I had a bunch of other things in my cart when I threw this stuff in there - it would have been somewhere between 100-200$US.

Oh, and I found those stereo jacks on ebay - https://www.ebay.com/itm/25-Pack-PJ-301CM-3-5mm-Stereo-Audio-Jack-with -Knurled-Nut-USA-Seller/232590960218


Thanks sduck! Fingers crossed im not too late to the train and can get the pcb and panel from gbiz smile

Thanks for the info and good luck with the built Om
Dunk_91
Here a couple of pics of my build, maybe can be helpful for someone screaming goo yo


sduck
gbiz wrote:

My advice would be to avoid using 0805 2u2 caps & use 0603. That kicad footprint is more like 0605, & theres very little free pad to get an iron on both the device & pad. An 0603 device is a much better fit, provided you ensure it's located centrally on the footprint. I can see some builders struggling to get the 0805 device fitted, making a mistake, trying to correct it & ripping up the power rail to the device it's bypassing - the MCU or DAC.


Good call on this. I was going to do a search on mouser for 0603 2u2 caps, but decided to try one first, to see how bad it was. I managed to get them all soldered, and although it's not my favorite work, they pass beep test. I've already had some practice fitting larger parts - I used the Octopart to buy from mouser, and it didn't fill the 0603 100n caps, but I have a ton of 0805 ones, so I've been using them.
gbiz
sduck wrote:
gbiz wrote:

My advice would be to avoid using 0805 2u2 caps & use 0603. That kicad footprint is more like 0605, & theres very little free pad to get an iron on both the device & pad. An 0603 device is a much better fit, provided you ensure it's located centrally on the footprint. I can see some builders struggling to get the 0805 device fitted, making a mistake, trying to correct it & ripping up the power rail to the device it's bypassing - the MCU or DAC.


Good call on this. I was going to do a search on mouser for 0603 2u2 caps, but decided to try one first, to see how bad it was. I managed to get them all soldered, and although it's not my favorite work, they pass beep test. I've already had some practice fitting larger parts - I used the Octopart to buy from mouser, and it didn't fill the 0603 100n caps, but I have a ton of 0805 ones, so I've been using them.


From Mouser, the TDK part 810-C1608X7R1A225K looks like the best option currently.

I don't envy you having to fit 25 0805 caps on 0603 footprints. Dead Banana
mxmxmx
re build tips. fyi/fwiw, the output stage / in the loop compensation looks a bit odd (the 18p cap):



... don't think it's a big deal, if any (i don't understand op amp circuits well enough...); i soldered the caps across the resistors anyways, so the cap sits in the direct feedback path.
toneburst
Anyone any advice on soldering the MCU and DAC chips?

I'd normally drag-solder for fine-pitch ICs, but the pads for these are very small.

Anyone had success with the drag-solder method, or had success with an alternative method?
kropo
Hi

On BOM leds are LTL1BEKVJNN, is there any red/yellow replacement for these on mouser? Those would be better since screen is yellow.

Regards

Jake
toneburst
kropo wrote:
Hi

On BOM leds are LTL1BEKVJNN, is there any red/yellow replacement for these on mouser? Those would be better since screen is yellow.

Jake


There's a note in Simon's Build Guide about the LEDs.

"Make sure to get the LEDs specified in the BOM. The circuit for driving the matrix is operating at the current sinking limits of the shift registers. For that reason, it is important to get bright LEDs, as reducing the current limiting resistor to increase brightness will damage the shift registers!"

I'd love to use red/blue LEDs, but I'm not going to risk it.
toneburst
The other reason to use the specified LEDs is to avoid having to mentally transpose the colours when reading through the documentation wink
toneburst
Maybe these Red/Blue ones would work..

https://www.digikey.co.uk/product-detail/en/sunled/XLMDKCBD34M/1497-13 07-ND/4901632

They're brighter than the specified parts.

One of my 2 Performer builds will have a blue screen, so red/blue would be cool.
toneburst
kropo wrote:
Hi

On BOM leds are LTL1BEKVJNN, is there any red/yellow replacement for these on mouser? Those would be better since screen is yellow.

Jake


These?
https://www.digikey.co.uk/product-detail/en/bivar-inc/3SBCC-R-Y/3SBCC- R-Y-ND/4925776

(Not in stock at DigiKey right now, but may be possible to source them from somewhere else).
sduck
toneburst wrote:
Anyone any advice on soldering the MCU and DAC chips?

I'd normally drag-solder for fine-pitch ICs, but the pads for these are very small.

Anyone had success with the drag-solder method, or had success with an alternative method?


I drag soldered them, and had to do a bit of clean up, but everything came out ok.
toneburst
Thanks @sduck, that’s encouraging smile
d.simon
ob
d.simon
as far as positioning resistors...I find a stereo microscope gives more precision when fiddling with SMD. Also, really appreciate SMD resistor code and other SMD component markings more with the microscope as opposed to without.

I had a batch of Bourns and a batch of Panasonic SMD resistors - the Bourns had markings, but Panasonic didn't. Really prefer the marked ones.
mxmxmx
d.simon wrote:
mxmxmx wrote:

i soldered the caps across the resistors anyways, so the cap sits in the direct feedback path.


sounds familiar...I did that on some older rev2c ornament+crime boards to bring them up to spec.


it's not really about "up to spec", more like i'm not sure the cap still does what it is supposed to do in this configuration. it probably doesn't matter much, idk; if it did matter much, i suppose people would have noticed by now.

(by rev2c the output stage, btw, had long assumed its present form (?), so it shouldn't sound familiar .. (well, it still had trim pots, but otherwise it was the same))
Altitude909
Here's mine. Im a little torn about the 301B jacks, they are tighter and have a smaller footprint but it really throws the whole height thing into chaos. The stereo ones are shorter (maybe its the Kobis, I didnt have the Qing Pu ones) so those need to "float" and the encoder kinda does too. I dont like pressing on the board like that unless its secure and as is, you'll be lucky to get one thread on. With the amount of jacks on this its a fairly big job even if you go with the "tight" 398S footprint without the hole..

I'm sure someone will, I dont like the buttons almost level with the panel look. Also, the LEDs in the BOM are pretty long and stick out a LOT (i used a 2mm panel and its too much). In hindsight, I wish I would have used some standoffs for those. The ones in the plaits bom would have been perfect
gbiz
That looks great Altitude909

I've got about one threads worth above the nut on the encoder (and that's with a washer), but only by barely having the encoder pins into the PCB. They needed soldering from the top to ensure they were well fastened. It's well fugly, but I'll take that over having the PCB flex.
Altitude909
I do the same thing as i do for all encoders like this: 2 nuts, one underneath, one on top.
Markthom
Just finished a couple of these at the weekend, one for me and one for a pal down in Yorkshire. It's a great module....

I would agree with some of the obervations folks have made, such as gBiz (I now need an acryllic screen cover Graeme!) and Altitude909 - there are certainly one or two things I would tweak if I build one in future such as LED standoff's, but still it's a great thing.....this is a fun wee one handed jam with it filmed last night, I'll try an develop it further as a track but it's nice as is - as an SDIY bonus it also uses gBiz's wonderful uDervish for the shimmer-y sound behind the melody;



PS I really like the fact that song mode means you arent limited to 64 step melodies, this one uses 128 steps and was relatively painless to put in to the module....

PPS The performance page is fabulous too!
djthopa
Wow some great looking builts here!

I just got the pcb for Michigan Synth Works.

Can i ask where did you get the panel done?

Did you have to convert the frontpanel.ai to any format to send it to frontpanel express?

Im going to see if Razorlab also does it, but im not sure if it willl take the file format.

edited: Razorlab will need a converted file.

Any one care to share their files?

Thanks

hmmm.....
iheartmodular
i can't find any

where are pcb/panel sets available please?
Markthom
iheartmodular wrote:
i can't find any

where are pcb/panel sets available please?


The gerbers are available for you to order your own, also watch out for people selling spares after doing that, as the min order is usually 5.

In Europe I know the Pusherman store is expecting some v soon, so keep an eye out for them, for US I'm not 100% sure, maybe Altitude909s Michigan Synthworks....
sduck
ZZ Ardoz had one or two extras as of a few days ago I think - look him up in the BST.

Here's a fpd file - it's kind of pricy, but maybe what you want if you can't find a cheaper option...
iheartmodular
lovely

thank you xx

This is fun!
djthopa
sduck wrote:
ZZ Ardoz had one or two extras as of a few days ago I think - look him up in the BST.

Here's a fpd file - it's kind of pricy, but maybe what you want if you can't find a cheaper option...


Thanks for sharing thumbs up

Edited: Yep, thats a nice panel!
Bartelby
My panel and pcb arrived today. It's going to be a while before I get around to building this. I have 6 other modules to build and a modular project of my own to complete first.
toneburst
waves to @Bartelby.
Virusinstaller
If anyone find a source for Aluminium panels that would be super handy.
I only have the black FR4. Which is still super nice though.
mqmq
Can this do polyphony over midi ?
Virusinstaller
I had someone ask me this today. But I'm really not sure either.
Must study the manual somemore.
mqmq
The manual says it can do polyphony over CV with the MIDI/CV interface but nothing about polyphony to MIDI hmmm.....
sirjpink
sduck wrote:
ZZ Ardoz had one or two extras as of a few days ago I think - look him up in the BST.

Here's a fpd file - it's kind of pricy, but maybe what you want if you can't find a cheaper option...

Thanks for sharing this! It looks absolutely fabulous in silver.

I hope SynthCube (or Pusherman) is able to put together a kit. Building stuff is fun, but I do not like the sourcing process. Dead Banana
Altitude909
sirjpink wrote:
sduck wrote:
ZZ Ardoz had one or two extras as of a few days ago I think - look him up in the BST.

Here's a fpd file - it's kind of pricy, but maybe what you want if you can't find a cheaper option...

Thanks for sharing this! It looks absolutely fabulous in silver.

I hope SynthCube (or Pusherman) is able to put together a kit. Building stuff is fun, but I do not like the sourcing process. Dead Banana


NC License, not going to happen unless Simon changes it
sduck
Arggh, got mine basically finished tonight but it's not working correctly. Buttons not working (all of them). Encoder working, but the button not working. Every time I start it, it goes to a different page before flaking out and locking up. Reflowing/debugging shall commence...
toneburst
Check IC orientation, first, I’d say.
Virusinstaller
sduck wrote:
Arggh, got mine basically finished tonight but it's not working correctly. Buttons not working (all of them). Encoder working, but the button not working. Every time I start it, it goes to a different page before flaking out and locking up. Reflowing/debugging shall commence...


Diode orientation correct? It might be as simple as that.
DozenCrows
PCBs and panels now up on Pusherman - https://pushermanproductions.com/product/westlicht-performer-sequencer  /

IIRC Simon mentioned on FB he was discussing this with Pusherman a little while back.
Altitude909
sduck wrote:
Arggh, got mine basically finished tonight but it's not working correctly. Buttons not working (all of them). Encoder working, but the button not working. Every time I start it, it goes to a different page before flaking out and locking up. Reflowing/debugging shall commence...


595s and the 165. tight pads on those, i had some rework to do there also.
toneburst
Altitude909 wrote:
595s and the 165. tight pads on those, i had some rework to do there also.


There are a Lot of tight pads. Ones for the MCU are pretty tiny, and those for the 0805 caps are practically impossible to solder with the specified parts. 0603 caps work fine here, though.
merlin
DozenCrows wrote:
PCBs and panels now up on Pusherman - https://pushermanproductions.com/product/westlicht-performer-sequencer  /

IIRC Simon mentioned on FB he was discussing this with Pusherman a little while back.


thumbs up

Karl
Altitude909
toneburst wrote:
Altitude909 wrote:
595s and the 165. tight pads on those, i had some rework to do there also.


There are a Lot of tight pads. Ones for the MCU are pretty tiny, and those for the 0805 caps are practically impossible to solder with the specified parts. 0603 caps work fine here, though.


KlingonCad. I didnt even realize he specd 805 parts. I used 603 throughout..
sduck
toneburst wrote:
Check IC orientation, first, I’d say.


Take a look at the first post, help me out - is everything ok? I was checking multiple diagrams/pictures as I soldered them in, and think i have the mall correct.

Virusinstaller wrote:


Diode orientation correct? It might be as simple as that.


This is a possiblility - I may have misread the PCB - I have all the lines on the diodes oriented towards the bottom of the pcb - is this wrong? I think I know diode symbols, but you never know what you don't know...

Altitude909 wrote:

595s and the 165. tight pads on those, i had some rework to do there also.


These were the easy ones of the bunch, compared to the DAC and the ARM. I've gone over them with a really strong magnifying glass, and beep tested everything, seems ok.
gbiz
The ICs all look ok from the image you posted in the OP, but teh resolution isn't that great. Better resolution pic would be useful.

The encoder and it's switch are run directly from one of the MCU ports, not the pushbutton matrix. Have you checked the soldering on the crystal & it's caps ?.
Are the supply rails at the correct voltage & clean/stable ?. The pushbutton matrix runs from +5V, that's then regulated down to 3.3V for the MCU.

And check Q5. That's responsible for the output enable for the pushbutton matrix drivers.
sduck
gbiz wrote:
The ICs all look ok from the image you posted in the OP, but teh resolution isn't that great. Better resolution pic would be useful.

The encoder and it's switch are run directly from one of the MCU ports, not the pushbutton matrix. Have you checked the soldering on the crystal & it's caps ?.
Are the supply rails at the correct voltage & clean/stable ?. The pushbutton matrix runs from +5V, that's then regulated down to 3.3V for the MCU.

And check Q5. That's responsible for the output enable for the pushbutton matrix drivers.


I haven't checked all of the power rails, need to do that. The 5v at least looks ok. The thing fires up, and seems to be working, except for the switches - they're throwing some random stuff, causing it to end up in some weird state that's different every time. Thanks for ideas of other stuff to check, I'll probably get on that more tomorrow.

forestcaver
Not seen this mentioned before but for those who aren’t aware, it is worth using the tester hex to verify function of all i/o jacks, buttons and encoder - it makes testing the function of the module very simple.
toneburst
Performer n.1 of my two mostly working, and playing very nicely with my Launchpad Mini.



The Clock input seems to be dead. I was having too much fun exploring the Launchpad functionality to flash the testing firmware, so don't know if anything else is non-functional, too.
forestcaver
sduck wrote:
gbiz wrote:
The ICs all look ok from the image you posted in the OP, but teh resolution isn't that great. Better resolution pic would be useful.

The encoder and it's switch are run directly from one of the MCU ports, not the pushbutton matrix. Have you checked the soldering on the crystal & it's caps ?.
Are the supply rails at the correct voltage & clean/stable ?. The pushbutton matrix runs from +5V, that's then regulated down to 3.3V for the MCU.

And check Q5. That's responsible for the output enable for the pushbutton matrix drivers.


I haven't checked all of the power rails, need to do that. The 5v at least looks ok. The thing fires up, and seems to be working, except for the switches - they're throwing some random stuff, causing it to end up in some weird state that's different every time. Thanks for ideas of other stuff to check, I'll probably get on that more tomorrow.



Triple check pins 20-23 on the stm32. It has to be the multiplexors, stm32 or q5. I suspect the encoder weirdness is just due to the page buttons acting weird and not the encoder, at a guess. May be worth scoping the multiplexor outputs and see what you are getting....
Virusinstaller
sduck wrote:
toneburst wrote:
Check IC orientation, first, I’d say.


Take a look at the first post, help me out - is everything ok? I was checking multiple diagrams/pictures as I soldered them in, and think i have the mall correct.

Virusinstaller wrote:


Diode orientation correct? It might be as simple as that.


This is a possiblility - I may have misread the PCB - I have all the lines on the diodes oriented towards the bottom of the pcb - is this wrong? I think I know diode symbols, but you never know what you don't know...

Altitude909 wrote:

595s and the 165. tight pads on those, i had some rework to do there also.


These were the easy ones of the bunch, compared to the DAC and the ARM. I've gone over them with a really strong magnifying glass, and beep tested everything, seems ok.


Sduck diodes down should be correct. Everything is populated?
Virusinstaller
toneburst wrote:
Performer n.1 of my two mostly working, and playing very nicely with my Launchpad Mini.



The Clock input seems to be dead. I was having too much fun exploring the Launchpad functionality to flash the testing firmware, so don't know if anything else is non-functional, too.


Clock input has a clock divider which can be found on the BPM page. Check that.

How is using the launchpad?
cnicht
What capabilities does the Launchpad Mini add to the Performer? Does it enable the selection of different sequences? Muting/unmuting? Playing scales? Inputting notes/values?

Very curious!
Bartelby
toneburst wrote:

The Clock input seems to be dead. I was having too much fun exploring the Launchpad functionality to flash the testing firmware, so don't know if anything else is non-functional, too.


Once you get that working you could go really crazy and, you know, use it to make some music... hihi
toneburst
cnicht wrote:
What capabilities does the Launchpad Mini add to the Performer? Does it enable the selection of different sequences? Muting/unmuting? Playing scales? Inputting notes/values?

Very curious!


It doesn't add any capabilities that can't be accessed from the module itself, as far as I can tell.

It does allow editing of most pattern-level properties from the pads.
toneburst
Bartelby wrote:
Once you get that working you could go really crazy and, you know, use it to make some music... hihi


How do you know I wasn't making music, while "exploring the Launchpad functionality".
toneburst
I wasn't, but that in no way invalidates my point wink
sduck
I found the problem. Very minor sort of thing, easily missed (at least for me apparently!). Bad solder on pin 11 of U5. It looked OK from most angles, I finally got it lit up the right way with my magnifying glass and there it was. Took all of 5 seconds to fix (heating up the iron was far longer). Fully functional now.

Thanks to everyone who responded! All your help and suggestions were greatly appreciated.


Virusinstaller
This is excellent & nice looking panel too!

I am finding this module super inspiring lately. Much fun wiggling!
Bartelby
toneburst wrote:

How do you know I wasn't making music, while "exploring the Launchpad functionality".


Just a hunch...

toneburst wrote:
I wasn't, but that in no way invalidates my point wink


applause
Roni
mqmq wrote:
Can this do polyphony over midi ?


I don't think it can. or at least I haven't worked it out. I had a fiddle in the MIDIOUT page, trying to set different events, sources and parameters to different MIDI channels, but it appears that it's a global setting, not per track or per MIDI channel. So you can set different source tracks for notes, gates, CCs etc but you only get one of each that you can output over the channel of your choice either over the MIDI or USB (but not both), if that makes sense.
toneburst
Re. polyphony: you may be able to re-route MIDI in events in your DAW, or whatever is receiving MIDI from the Performer to the same channel.

Alternatively, if the receiving MIDI device is mono-timbral, you may be able to put it into OMNI mode, so it effectively responds to notes etc. on all channels.

You will still need one Performer channel per stacked note.
sneak-thief
toneburst wrote:
cnicht wrote:
What capabilities does the Launchpad Mini add to the Performer? Does it enable the selection of different sequences? Muting/unmuting? Playing scales? Inputting notes/values?

Very curious!


It doesn't add any capabilities that can't be accessed from the module itself, as far as I can tell.

It does allow editing of most pattern-level properties from the pads.


You can get an idea of what the Launchpad does by looking at the raw code functions:

Quote:


From LaunchpadController.h:

// Sequence mode
void sequenceEnter();
void sequenceExit();
void sequenceDraw();
void sequenceButtonDown(const Button &button);
void sequenceButtonUp(const Button &button);

void sequenceUpdateNavigation();

void sequenceSetLayer(int row, int col);
void sequenceSetFirstStep(int step);
void sequenceSetLastStep(int step);
void sequenceEditStep(int row, int col);
void sequenceEditNoteStep(int row, int col);
void sequenceEditCurveStep(int row, int col);

void sequenceDrawLayer();
void sequenceDrawStepRange(int highlight);
void sequenceDrawSequence();
void sequenceDrawNoteSequence();
void sequenceDrawCurveSequence();

// Pattern mode
void patternEnter();
void patternExit();
void patternDraw();
void patternButtonDown(const Button &button);
void patternButtonUp(const Button &button);

// Performer mode
void performerEnter();
void performerExit();
void performerDraw();
void performerButtonDown(const Button &button);
void performerButtonUp(const Button &button);

// Navigation
void navigationDraw(const Navigation &navigation);
void navigationButtonDown(Navigation &navigation, const Button &button);

// Drawing
void drawTracksGateAndSelected(const Engine &engine, int selectedTrack);
void drawTracksGateAndMute(const Engine &engine, const PlayState &playState);

void drawStepRange(int first, int last, int highlit);



From LaunchpadController.cpp:

static const LayerMapItem noteSequenceLayerMap[] = {
[int(NoteSequence::Layer::Gate)] = { 0, 0 },
[int(NoteSequence::Layer::GateProbability)] = { 1, 0 },
[int(NoteSequence::Layer::Retrigger)] = { 0, 1 },
[int(NoteSequence::Layer::RetriggerProbability)] = { 1, 1 },
[int(NoteSequence::Layer::Length)] = { 0, 2 },
[int(NoteSequence::Layer::LengthVariationRange)] = { 1, 2 },
[int(NoteSequence::Layer::LengthVariationProbability)] = { 2, 2 },
[int(NoteSequence::Layer::Note)] = { 0, 3 },
[int(NoteSequence::Layer::NoteVariationRange)] = { 1, 3 },
[int(NoteSequence::Layer::NoteVariationProbability)] = { 2, 3 },
[int(NoteSequence::Layer::Slide)] = { 2, 0 },
};


static const LayerMapItem curveSequenceLayerMap[] = {
[int(CurveSequence::Layer::Shape)] = { 0, 0 },
[int(CurveSequence::Layer::Min)] = { 0, 1 },
[int(CurveSequence::Layer::Max)] = { 0, 2 },
};




- https://github.com/westlicht/performer/blob/master/src/apps/sequencer/ ui/controllers/launchpad/LaunchpadController.h

- https://github.com/westlicht/performer/blob/master/src/apps/sequencer/ ui/controllers/launchpad/LaunchpadController.cpp
cnicht
Very useful - thank you.

Does it require any programming of the Launchpad?
sneak-thief
cnicht wrote:
Very useful - thank you.

Does it require any programming of the Launchpad?


Hopefully toneburst can answer that. Haven't built one yet.
toneburst
Anyone know what MIDI DIN > TRS standard is used?

I can't seem to get either MIDI in or out working with the adapter that came with my Beatstep Pro, which leads me to suspect it's wired the wrong way round.

Anyone got MIDI I/O out working?
toneburst
cnicht wrote:
Does it require any programming of the Launchpad?


No, it's plug-and-play smile
cnicht
That’s great! Thank you.
toneburst
I’m not sure if this has been documented elsewhere, but for the record, I’ve discovered that the Performer utilises the Make Noise type of DIN > TRS MIDI adapter, rather than the Novation/Arturia type.

Make Noise sell replacement adapters for their O-Coast module that will do the trick. Or you could make up your own cables or adapters, of course.
toneburst
Oh, and a gotcha when using MIDI output; it’s possible to to set a constant output note velocity (in fact it’s necessary to do this, to get useable output from a single Track), But the velocity of the output note has a value 1 greater than the value set on the Performer.

Also (and this had me scratching my head for ages, until I ran a MIDI monitor on the Performer MIDI out), the value “wraps” at 126, so setting a constant value of 127 (max velocity) on the Performer produces a stream of MIDI Note Off messages (notes with 0 velocity).

Ive flagged up the bug (if it is a bug) with Simon, but I have a feeling he’s swamped with messages, now that Performer boards have popped up in all the popular sDIY web-shops.
toneburst
I’ve said this before, but Launchpad integration is Very Cool!

I’d strongly suggest, if you’re making one of these, and you don’t already have a Launchpad S or Mini, get one! They’re not expensive on the 2nd-hand market, and really bring another dimension to the Performer.
toneburst
toneburst wrote:
Oh, and a gotcha when using MIDI output; it’s possible to to set a constant output note velocity (in fact it’s necessary to do this, to get useable output from a single Track), But the velocity of the output note has a value 1 greater than the value set on the Performer.

Also (and this had me scratching my head for ages, until I ran a MIDI monitor on the Performer MIDI out), the value “wraps” at 126, so setting a constant value of 127 (max velocity) on the Performer produces a stream of MIDI Note Off messages (notes with 0 velocity).

Ive flagged up the bug (if it is a bug) with Simon, but I have a feeling he’s swamped with messages, now that Performer boards have popped up in all the popular sDIY web-shops.


This has been fixed in the latest 0.1.6 firmware release, apparently.
sneak-thief
Midi RX: Tip is current sink, Ring is current source
Midi TX: Tip is current sink, Ring is current source



so, as toneburst mentioned it's:

TRS To MIDI Male Type A

(Akai Pro, IK Multimedia, Korg, Line 6, littleBits, Make Noise)

https://www.midi.org/articles-old/updated-how-to-make-your-own-3-5mm-m ini-stereo-trs-to-midi-5-pin-din-cables
toneburst
I know from experience that it’s not the same as the MIDI > TRS Arturia use (unless they've changed it since late last year when I bought my black BSP).

Simon made the right choice, going with that’s now become the official standard wiring, I think.

It's quite easy to convert between the two formats by making up a little adapter cable to flip the TRS tip and ring connections (GND is the same in both types).
Altitude909
toneburst wrote:
I know from experience that it’s not the same as the MIDI > TRS Arturia use (unless they've changed it since late last year when I bought my black BSP).

Simon made the right choice, going with that’s now become the official standard wiring, I think.

It's quite easy to convert between the two formats by making up a little adapter cable to flip the TRS tip and ring connections (GND is the same in both types).


Or just put a pair of BAT54Ss on the opto input so you can use either type. We should really ask for a hardware rev, he's almost got it perfect but that and those damn 11.5mm need to go
toneburst
Altitude909 wrote:
Or just put a pair of BAT54Ss on the opto input so you can use either type.


That would be cool. I was thinking the same, in fact. You'd need a dual-channel option-isolator for the input, too, I think. You can get those in the same footprint as the specified single-channel ones.

Quote:
We should really ask for a hardware rev, he's almost got it perfect but that and those damn 11.5mm need to go


There are a few things that would be improved, for the DIYers, I think, including:

    Thonkiconns for all sockets
    Bigger pads for passives, especially 0805 caps (of spec 0603 instead, and use same footprints as other 0603 parts)
    Bigger pads for MCU and DAC


Having said that, I'm not sure we're in a position to request anything (including support and troubleshooting advice) from Simon.
Altitude909
toneburst wrote:
..
That would be cool. I was thinking the same, in fact. You'd need a dual-channel option-isolator for the input, too, I think. You can get those in the same footprint as the specified single-channel ones.
...


Its already done and tested. This is Neutron sounds "PSImidi". Already in use on Threshold, mBrane, and my midi expander. It also allows for male or female midi to TRS cables since that's just a pin swap.



It's up to him to take advice from people, community developed things tend to end up much better.. The jack thing is a deal breaker though, I thought it would be OK but that was on the premise that the jacks are 0.5mm taller, they are 1.5mm taller and those buttons he's using are straight up too short. Having them them "barely on" isnt a viable solution. Same goes for the stereo jacks, good 1mm shorter than the others.
toneburst
Probably not the right place for this, given this is a hardware build thread, primarily, but I wonder if adding support for this cheap Akai device, in addition to the Launchpad, might be cool.

https://d2r1vs3d9006ap.cloudfront.net/s3_images/1143703/apc_mini_midi. jpg?1421447412

It scores over the Launchpad because of the builtin sliders, but has the same tri-state illuminated 64-pad button matrix, which could work in the same way as the Novation controller, while the sliders could be mapped to various global or track-specific parameters.
toneburst
Altitude909 wrote:
It's up to him to take advice from people, community developed things tend to end up much better..


Well, he can if he wants to. He's under no obligation to do so.
If he were selling a commercial product, the situation would be different.

Quote:

The jack thing is a deal breaker though, I thought it would be OK but that was on the premise that the jacks are 0.5mm taller, they are 1.5mm taller and those buttons he's using are straight up too short. Having them them "barely on" isnt a viable solution. Same goes for the stereo jacks, good 1mm shorter than the others.


I agree 110%.

I suspect Simon's primary interest is in the firmware, since that's where the real innovations lie, so I'm sure he would take advice on improvements to the hardware.

I (perhaps stupidly) bought an extra Performer PCB/panel set, with the intention of selling the spare, to try and claw back some of the money spent on the build of the two (I cleared it with Simon, first).

These hardware issues cause me moral qualms about charging the kind of money I'd normally expect to pay for a module with the capabilities of the Performer. Not that there really is anything else on the market with the feature-set of this module, but I'm sure you get my point..
sneak-thief
Altitude909 wrote:
In hindsight, I wish I would have used some standoffs for those. The ones in the plaits bom would have been perfect


Do you mean these?

https://www.mouser.de/ProductDetail/Bivar/ERM-1-300?qs=RaByzyhmEWh15rH m3WtL0g==


For all the fresh new BOM-hunters, don't forget that ST and TI have f r e e samples of the cpu and $$$ DAC.
toneburst
I usually just use tape for the LEDs. Stick a strip of masking tape over the hole on the top surface of the panel, poke the LEDs through from the bottom, they stick to the tape, and you can solder them easily.

Of course, if you want them to stick up a little, that method won't work..
toneburst
Getting the button caps the right height is more challenging.

I've actually resorted to raising the 5 switches below the display slightly above the surface of the board. This allowed me to push the rectangular caps all the way down, avoiding the issue of potentially pressing the button too hard, and ending up pushing the button down below the level of the top of the panel.

I haven't re-fitted the panel yet, so I don't know yet if this has worked.

I've been thinking of adding plastic washers over the actuators of the other switches, below the round caps, to act as spacers, so I can make sure all the round caps are pushed down exactly the same amount (and again, preventing them being pushed down too far).

I'm wary of attempting to remove the already-fitted round caps from my first Performer build, for fear of breaking the delicate switches, so I'll probably just try this with the 2nd build.
Altitude909
sneak-thief wrote:


Do you mean these?

https://www.mouser.de/ProductDetail/Bivar/ERM-1-300?qs=RaByzyhmEWh15rH m3WtL0g==


For all the fresh new BOM-hunters, don't forget that ST and TI have f r e e samples of the cpu and $$$ DAC.


yeah, those are the ones. slick design
toneburst
sneak-thief wrote:
For all the fresh new BOM-hunters, don't forget that ST and TI have f r e e samples of the cpu and $$$ DAC.


Oh, really? That's interesting. Any tips on strategies for acquiring these?
OIP
oh snap, that could push me over the edge. the DAC is a bit eye watering
sneak-thief
@Altitude909 - Thanks!

@toneburst & OIP - sign up for an account with ST & TI. Now, I'm not sure if there are new specific criteria for samples that I'm not aware of, but usually when you're logged in, simply look at the respective product page and there's a link that says SAMPLE & BUY (ST) or Order Now (TI). There will be a "Sample" button next to whichever product that has samples available.

http://www.ti.com/info-store/help/free-sample-program.html

Note: they may occasionally refuse to send the samples after you order them but it's never happened to me in the past 15 years.
sneak-thief
That encoder is begging for a bigger knob. How about this 20mm Davies?



https://www.digikey.de/product-detail/en/1202CY/1722-1068-ND/6566291/? itemSeq=287533246
sneak-thief
toneburst wrote:
Getting the button caps the right height is more challenging.


OK, that's it. I'm going to resort to bending the 4 bottom metal flaps of the jacks which brings the total height from pcb to panel to 10.5mm (including the lumpy plastic bezel at the bottom of the threads which never ends up going through the panels).

I've had to do this before:



angry
toneburst
Eek...
toneburst
sneak-thief wrote:
@toneburst & OIP - sign up for an account with ST & TI. Now, I'm not sure if there are new specific criteria for samples that I'm not aware of, but usually when you're logged in, simply look at the respective product page and there's a link that says SAMPLE & BUY (ST) or Order Now (TI). There will be a "Sample" button next to whichever product that has samples available.

http://www.ti.com/info-store/help/free-sample-program.html

Note: they may occasionally refuse to send the samples after you order them but it's never happened to me in the past 15 years.


I may give that a go, thanks for the tip!

I'm not worried about the MCU, but that TI DAC is Really pricey...

a|x
toneburst
sneak-thief wrote:
That encoder is begging for a bigger knob. How about this 20mm Davies?

itemSeq=287533246


Might go a bit close to the left edge of the display, though.

I'm going to go for a Rogan encoder knob. Looks nice on my TSNM and MI modules.
gbiz
sneak-thief wrote:
toneburst wrote:
Getting the button caps the right height is more challenging.


OK, that's it. I'm going to resort to bending the 4 bottom metal flaps of the jacks which brings the total height from pcb to panel to 10.5mm (including the lumpy plastic bezel at the bottom of the threads which never ends up going through the panels).

I've had to do this before:



angry


I can't remember what module it was now, where the recommended fix for the extra height with these jacks was remove the metal tabs completely, not bend them. It leaves a jagged edge, so needs care if there's any trace underneath them.
That coupled with removing the plastic from the base of the bushing would probably be more than enough.
Less work than refitting switches or pulling caps.
sneak-thief
@sduck - what knob is that?

It looks like it'll be the same as sduck's - which seems acceptable to me at first glance:



sneak-thief
gbiz wrote:


I can't remember what module it was now, where the recommended fix for the extra height with these jacks was remove the metal tabs completely, not bend them. It leaves a jagged edge, so needs care if there's any trace underneath them.
That coupled with removing the plastic from the base of the bushing would probably be more than enough.
Less work than refitting switches or pulling caps.


Ack, cutting would be bad. Bending is quick and doesn't seem to mess up the jack.

But how would you remove that plastic bezel on the base? Utility knife? That would be.. uhh, irritating.
sduck
That knob is from Small Bear Electronics - http://smallbear-electronics.mybigcommerce.com/aluminum-large-black-wh ite-index/

Protip - if you want this knob to rotate evenly, place a small piece of plastic drinking straw on the encoder shaft.
gbiz
sneak-thief wrote:
gbiz wrote:


I can't remember what module it was now, where the recommended fix for the extra height with these jacks was remove the metal tabs completely, not bend them. It leaves a jagged edge, so needs care if there's any trace underneath them.
That coupled with removing the plastic from the base of the bushing would probably be more than enough.
Less work than refitting switches or pulling caps.


Ack, cutting would be bad. Bending is quick and doesn't seem to mess up the jack.

But how would you remove that plastic bezel on the base? Utility knife? That would be.. uhh, irritating.


I didn't say cutting the tabs was good, just that was the fix people adopted on here for this problem with another module some time back wink (IIRC in that case, i removed the tabs by bending them back & forward a few times until the tab broke). From looking at my performer here, i'm not sure if there's enough distance between the jacks for bent out tabs.

Yes, a knife would be one way to remove that plastic at the base of the bushing. You'll find you probably have to do that anyway for the two stereo jacks, they only have 3.5mm of bushing otherwise. Yes it's irritating. At least the plastic is soft & easily cut with a sharp knife.


12mm soft touch Sifam knob from Thonk for me. I put a Davies pointerless one on there but hard plastic isn't great for encoder knobs.
Altitude909
Ok, enough with the PJ301 tab bending talk. Board rev is in the works now.

So for the USB there are two options:

A) Keep same part but open up the hole in the panel
B) Switch to they Kycon which is 10mm but needs a different footprint

discuss.
sduck
Opening up the hole a little seems like an easier choice.
Altitude909
sduck wrote:
Opening up the hole a little seems like an easier choice.


Unless you're changing a bunch of shit on the board already, which we are.

This the before and after:

Tough call. The Kycon is half the price too..

mangros
Altitude909 wrote:
Ok, enough with the PJ301 tab bending talk. Board rev is in the works now.


Ah shit, anyone want the PCB/panel/load of PJ301BM jacks I've just got in? smile

Seriously though - is the current state of play bordering on unusable with a 1.6mm PCB panel or is it mostly cosmetic? I don't mind floating a few components off the board, and if the large buttons are almost flush I can live with it.

Though if it's physically hard to use I'd rather cut my losses and wait for a new board before soldering in £60 worth of parts.
sneak-thief
mangros wrote:
Altitude909 wrote:
Ok, enough with the PJ301 tab bending talk. Board rev is in the works now.


Ah shit, anyone want the PCB/panel/load of PJ301BM jacks I've just got in? smile

Seriously though - is the current state of play bordering on unusable with a 1.6mm PCB panel or is it mostly cosmetic? I don't mind floating a few components off the board, and if the large buttons are almost flush I can live with it.

Though if it's physically hard to use I'd rather cut my losses and wait for a new board before soldering in £60 worth of parts.


Same boat here - too late now, let's just plow through this Mr. Green
patrickod
sneak-thief wrote:
mangros wrote:
Altitude909 wrote:
Ok, enough with the PJ301 tab bending talk. Board rev is in the works now.


Ah shit, anyone want the PCB/panel/load of PJ301BM jacks I've just got in? smile

Seriously though - is the current state of play bordering on unusable with a 1.6mm PCB panel or is it mostly cosmetic? I don't mind floating a few components off the board, and if the large buttons are almost flush I can live with it.

Though if it's physically hard to use I'd rather cut my losses and wait for a new board before soldering in £60 worth of parts.


Same boat here - too late now, let's just plow through this Mr. Green


Likewise. I put in a pcb and panel order just recently. I'm hopeful that with a 1.6mm panel and some of the component placement advice here that I can reach a good result. Certainly there appear to be some solid builds out there in the community already love

Separately, if anyone is interested in v1.0 panels and PCBs at cost or trade in the USA feel free to reach out, as they're otherwise going to sit idle on a shelf.
sduck
[quote="mangros"]
Altitude909 wrote:


Seriously though - is the current state of play bordering on unusable with a 1.6mm PCB panel or is it mostly cosmetic? I don't mind floating a few components off the board, and if the large buttons are almost flush I can live with it.



No, not at all. It works fine as is. Currently, with a 1.6mm panel, the bigger buttons are flush with the panel, but are still perfectly useable - they only require a tiny amount of downward pressure to activate.
Altitude909
Yeah, its not terrible. I super glued all the caps on and its fine since they stick out about the same as a "right" panel but I dont think I would consider using it with the buttons pushed all the way down like they are supposed be. Especially with a 2mm panel. There is a lot of "hold this button while you press this button/rotate the encoder" action and that would drive me nuts if the caps were pretty much flush with the panel
toneburst
The other issue with the buttons is that, unless you glue them on, it's possible to push them a bit hard, and actually accidentally push a cap down a bit further on the switch, so some caps end up lower than others.
toneburst
Glad you're taking this on, @Altitude909!

c wrote:


If you're moving stuff around, can I suggest shifting the USB or SD slot a tad, so the panel cutouts are centred in relation to each other? Very minor aesthetic thing, but looks neater, I think.

Also, I found the transistor Q1 to be very close to the hole for the top-left display standoff.

And, of course, those teeny tiny 0805 cap pads.

Do you envisage any issues with the display not fitting under the lowered panel?
sneak-thief
toneburst wrote:
The other issue with the buttons is that, unless you glue them on, it's possible to push them a bit hard, and actually accidentally push a cap down a bit further on the switch, so some caps end up lower than others.


Oh ho ho, I had already superimposed the two datasheets and deduced as much:



I'm gonna find some plastic bits of junk I can wedge up the holes to prevent exactly that.

In addition, I have thin plastic washers I'm going to stick under the buttons when soldering them in.

(This project is gonna kill me: I developed a nasty chemical intolerance to flux and have to wear a full plexiglass mask and use an exhaust fan. But it's that good.)
sneak-thief
toneburst wrote:
Do you envisage any issues with the display not fitting under the lowered panel?


If you don't mind checking: how much space is between the display and the panel?

And between the display and the header socket below?

The standoffs can always be substituted - the header, not so much.
forestcaver
Just another perspective: I had no issues with button cap height (1.6mm panel thickness). They seem fine to me. Likewise jack and encoder height were ok - I raised them off the board a little when soldering.

I also didn’t notice the pad sizes for passives or ics being small or tricky to solder... not claiming to be any good at building this stuff but there didn’t seem to be these issues on my board....
toneburst
forestcaver wrote:
Just another perspective: I had no issues with button cap height (1.6mm panel thickness). They seem fine to me. Likewise jack and encoder height were ok - I raised them off the board a little when soldering.


Interesting. Maybe we're all getting a bit hung-up on the caps issue, so it's good to get another perspective. I did find them a bit low, though, to be honest, and must harder to get even heights than I expected.

Quote:
I also didn’t notice the pad sizes for passives or ics being small or tricky to solder... not claiming to be any good at building this stuff but there didn’t seem to be these issues on my board....


The pads for the MCU and DAC could definitely be bigger. Having said that, despite my nerves, I didn't have any problems drag-soldering either.

More of an issue for me were the 0805 caps. The pads for these are way too small for me to comfortably solder this size of part. I managed one, just, but used 0603 versions for the others, and the pads are fine for the smaller package.

I'm not sure why Simon speced 0805 caps here, really. Maybe he just couldn't find that value in 0603 at the time, due to the ongoing global MLCC shortage.

Maybe you have a steadier hand than I do.
Altitude909
where does he actually say use 805? My bom says all 0603 except for those 2u2 parts

I'm not going to move anything around and now that I really think about it, the USB has to change to the short one as well. I've seen firsthand what happens when there are two panels for one project and its a VERY big mess. Everything has to stay exactly as is so the new board will be compatible with the panel. As far as display height, Its a non-issue; plenty thin
toneburst
Altitude909 wrote:
where does he actually say use 805? My bom says all 0603 except for those 2u2 parts


It's just those 4. Annoying if you source 0805 parts per the BOM, and can't solder them, though.

You could just change the BOM for 0603 parts, but if you're going to do that no reason not to also change the pads to the same footprint as the other 0603 MLCCs. It's not a massive issue, but would bug me, if I were tweaking the board layout.

Fair enough re. moving the USB and SD slots. Again, it's no biggie. Bigger pads for the high-density chips would be welcome, though, I think (though I've heard that too long pads can be a problem for automated assembly, so maybe don't go overboard).

Out of interest, how are you pairing the GNDs for the Thonkiconns, with vertical pairs sharing the same hole, or horizontal ones?
Altitude909
vertical pairs sharing, horizontal no. The board will be a bit taller as a result (like a hair) but it was the most logical approach since the midi in shouldnt have the ground tied to anything
toneburst
I'm declaring my first Performer build "finished" smile

I'll do some things differently with build no.2, but I'm generally happy with the way the first one has turned out.

It's a great module, and really inspiring to play with! I especially like that it sends out MIDI. It's great to have my MIDI hardware playing along with my Eurorack modules, and all without a computer...
sduck
Just in case anyone sticking their head into this thread wonders what we're going on about, here's a pic of what the buttons look like when they're pushed down all the way onto the switches currently, with a 1.6mm thick panel. And note that I've got the leds sticking out as far as they go, which I suspect some people don't like, but I do. They're very useable like this; however, I can certainly imagine them being even easier to use if they were sticking out a bit more (the switches, not the leds).

merlin
sneak-thief wrote:
@Altitude909 - Thanks!

@toneburst & OIP - sign up for an account with ST & TI. Now, I'm not sure if there are new specific criteria for samples that I'm not aware of, but usually when you're logged in, simply look at the respective product page and there's a link that says SAMPLE & BUY (ST) or Order Now (TI). There will be a "Sample" button next to whichever product that has samples available.

http://www.ti.com/info-store/help/free-sample-program.html

Note: they may occasionally refuse to send the samples after you order them but it's never happened to me in the past 15 years.


Do ST charge postage for samples? I'm in the UK and they quote $37 for delivery. I know that other manufacturers drop the delivery charge for samples (I have ordered the DAC for this build) but have never used ST before.

Thanks,

Karl
Altitude909
ok, off to fab. It all works pretty well except for the midi in which will need the ground lead cut off, small price to pay

Altitude909
..
windspirit
Ok, where are people getting metal panels?
toneburst
Nice work @Altitude909!
sneak-thief
merlin wrote:

Do ST charge postage for samples? I'm in the UK and they quote $37 for delivery. I know that other manufacturers drop the delivery charge for samples (I have ordered the DAC for this build) but have never used ST before.

Thanks,

Karl


I've never had a postal charge to Germany. I can't say if that's normal or if there's a free UK shipping option.
sneak-thief
windspirit wrote:
Ok, where are people getting metal panels?


So far in terms of commercial options, Pusherman UK are collaborating with Westlicht:

https://pushermanproductions.com/product/westlicht-performer-sequencer -black-aluminium-panel-only/
Altitude909
toneburst wrote:
Nice work @Altitude909!


I farmed it out. oops allergic to kicad
toneburst
Altitude909 wrote:
toneburst wrote:
Nice work @Altitude909!


I farmed it out. oops allergic to kicad


Ah, fair enough. Looks good, anyway.
sneak-thief
toneburst wrote:
Probably not the right place for this, given this is a hardware build thread, primarily, but I wonder if adding support for this cheap Akai device, in addition to the Launchpad, might be cool.

https://d2r1vs3d9006ap.cloudfront.net/s3_images/1143703/apc_mini_midi. jpg?1421447412

It scores over the Launchpad because of the builtin sliders, but has the same tri-state illuminated 64-pad button matrix, which could work in the same way as the Novation controller, while the sliders could be mapped to various global or track-specific parameters.


I agree - it's not much larger and gives it much more potential.

One would first get the vendor and product IDs, add it to this:

https://github.com/westlicht/performer/blob/master/src/apps/sequencer/ ui/ControllerRegistry.h

Then add the new controller folder and re-map the launchpad files using this chart:



The extra row of 8 buttons would also be nice.
OIP
all thonkiconns is a great development, thanks!
autodafe
I jusr realized I didn't order Y1, the Abracon Crystal

I have plenty of "standard" SMD 8Mhz Crystals, do you think I can replace it??

Altitude909
autodafe wrote:
I jusr realized I didn't order Y1, the Abracon Crystal

I have plenty of "standard" SMD 8Mhz Crystals, do you think I can replace it??



Should be fine, take it off that plastic holder and bend the legs into little feet so you have room to solder it
merlin
sneak-thief wrote:
windspirit wrote:
Ok, where are people getting metal panels?


So far in terms of commercial options, Pusherman UK are collaborating with Westlicht:

https://pushermanproductions.com/product/westlicht-performer-sequencer -black-aluminium-panel-only/


Yes, my PCB/Front Panel arrived from Pusherman on Monday- see here for options.

Karl
Virusinstaller
I have started pushing the switches up on my builds so that the legs on the rear of the pcbs are flush with the surface. This then allows the switch caps to sit higher regardless of whether they have been pushed all of the way down.
Feels much nicer using it this way, more tactile.
sneak-thief
Virusinstaller - unless you're putting something under the switch like a plastic washer, every button-press stresses the solder joints.

There's a real risk the joints will crack over time.
toneburst
sneak-thief wrote:

I agree - it's not much larger and gives it much more potential.


And they're cheap, too.

Quote:
One would first get the vendor and product IDs, add it to this:

https://github.com/westlicht/performer/blob/master/src/apps/sequencer/ ui/ControllerRegistry.h

Then add the new controller folder and re-map the launchpad files using this chart:




That was my thought, too. Shouldn't be too hard wink

Off course, I don't have one of these controllers to test, but may have to get one..


Quote:
The extra row of 8 buttons would also be nice.


Same number of buttons as the Launchpad, in fact. It's really just the additional sliders, and that single non-illuminated button at bottom right that distinguish it from the LP.

I'm (possibly naively) assuming that the CCs sent out by the sliders will just be picked up by the Performer, and can be mapped to anything you want on the Routing page.
sneak-thief
@toneburst: I misinterpreted the lack of midi note numbers on the top row in the Launchpad programmer's guide - it looked like they didn't transmit anything useful. After looking elsewhere, it appears they send and react to midi CC 104-111. Mystery solved!

Do you think the short vertical height of the APC mini buttons is an ergonomic issue?
toneburst
sneak-thief wrote:
@toneburst: I misinterpreted the lack of midi note numbers on the top row in the Launchpad programmer's guide - it looked like they didn't transmit anything useful. After looking elsewhere, it appears they send and react to midi CC 104-111. Mystery solved!


They're used extensively in Performer Launchpad integration.

Quote:
Do you think the short vertical height of the APC mini buttons is an ergonomic issue?


It's not ideal, but you're getting 9 sliders, in addition to the 64 pads, in more-or-less the same form-factor as the Launchpad, so I'm OK with the oblong buttons, personally.
sduck
Hey, folks discussing non-build related things, I don't mean to chase you away, but it would be nice if we could share some of the love with the neglected performer thread over in the euro forum - https://www.muffwiggler.com/forum/viewtopic.php?t=213251
autodafe
Just finished building this. Flashed OK, but this is what I get...Screen goes berserk after a few seconds...same stuff also in "tester" mode...

I must confess I made a terrible job on the screen, I didn't have suitable pin headers so I soldered it directly on the board, just to discover I had to desolder it to clean some small bridges on the STM...(I know, I am an idiot)
I damaged some pads (pins 7 and 8 of the screes) but I was able to trace the connections and I have exposed the VIAs and soldered a couple of wires directly...It was working but then started doing like this...

Any idea?



Image of the mess I made with the display pins:

cnicht
Could it be thermal expansion of one of the solder joints?

I’ve built a couple and wouldn’t like to directly solder the display given the price of them!
autodafe
cnicht wrote:
Could it be thermal expansion of one of the solder joints?


you mean the ones that I have soldered on the vias?

cnicht wrote:

I’ve built a couple and wouldn’t like to directly solder the display given the price of them!


I know, please don't tell me, I am a complete idiot, I was hurrying to get it finished....
cnicht
autoldafe wrote:
cnicht wrote:
Could it be thermal expansion of one of the solder joints?


you mean the ones that I have soldered on the vias?


Yes - maybe a dry joint heating up and expanding and losing connectivity? Try powering up and applying some freezer spray to the soldered joints.
toneburst
sneak-thief wrote:
One would first get the vendor and product IDs, add it to this:

https://github.com/westlicht/performer/blob/master/src/apps/sequencer/ ui/ControllerRegistry.h

Then add the new controller folder and re-map the launchpad files using this chart:




I just noticed the round buttons below and to the right of the pad matrix are single-colour (red across the bottom, green down the side), rather than bi-colour (as their equivalents on the Launchpad are).

That might throw a spanner in the works a bit.
toneburst
@autodafe could the bridges on the MCU still be there? If you have one, have a Really good look with a loupe or other illuminated magnifier, just to make sure there isn't still something between some pins there.
cnicht
autoldafe wrote:
cnicht wrote:
Could it be thermal expansion of one of the solder joints?


you mean the ones that I have soldered on the vias?


Yes - maybe a dry joint heating up and expanding and losing connectivity? Try powering up and applying some freezer spray to the soldered joints.
autodafe
toneburst wrote:
@autodafe could the bridges on the MCU still be there? If you have one, have a Really good look with a loupe or other illuminated magnifier, just to make sure there isn't still something between some pins there.


I am pretty sure there aren't any now. I have checked with my microscope and I can flasf/reflash the MCU as I like...
autodafe
looks like it's solved...there was somethin' strange on pin 19 and 20, they are supposed to be on GND but I measeured some resistance on those pins. connected bot together to GND and now the screen is stable :-)

Let's debug the rest and see if everything works as expected...
autodafe
yeah!

cnicht
The Tester firmware is very useful for debugging the hardware and testing its functionality!
autodafe
cnicht wrote:
The Tester firmware is very useful for debugging the hardware and testing its functionality!


yeah, right, I will flash it tonight, time permitting...
However I have been playing a bit this afternoon and I created a simple drum pattern using the 1st 3 CV/gates and they worked perfectly ;-)
DozenCrows
Finished my build yesterday and tested with the test firmware - all was good, but when the clock out, reset out or any of the gates were set high, there was some ripple on the output - occasional little 'v' shaped drops in the signal on my scope of around 0.3 V to 0.4 V.

Anyone else experienced this?
cnicht
Make sure you’re getting a signal from the microcontroller to R19 then through Q3.

I hadn’t soldered a transistor connection and had a similar problem.
toneburst
Altitude909 wrote:


The fact you've got a little MIDI keyboard plugged into the USB port there suggests it can be used for programming step notes. Is that the case?
Altitude909
toneburst wrote:
..
The fact you've got a little MIDI keyboard plugged into the USB port there suggests it can be used for programming step notes. Is that the case?


Yes, its a bit awkward but yes
toneburst
Altitude909 wrote:
toneburst wrote:
..
The fact you've got a little MIDI keyboard plugged into the USB port there suggests it can be used for programming step notes. Is that the case?


Yes, its a bit awkward but yes


OK. In what way is it awkward?
toneburst
I'm guessing you don't always get the notes you specify, if the current scale doesn't include the note.
DozenCrows
cnicht wrote:
Make sure you’re getting a signal from the microcontroller to R19 then through Q3.


Thank you for the tip! I'll check that.

However I see the same issue on clock out, reset out and all the gates - so wonder if it might be something with the 5V supply.
cnicht
The +5V rail is derived from the DC-DC converter and should be measurable on Pin 3 of it.
sneak-thief
@DozenCrows - are you sure your psu has enough power for this module? Did you try powering it all by itself?
DozenCrows
sneak-thief wrote:
@DozenCrows - are you sure your psu has enough power for this module? Did you try powering it all by itself?


Yup, this is running by itself off a Frequency Central FC Power PSU, which should have plenty of oomph for this module.

One idea I had is that I may have a bad joint on one of the associated caps for the 5V convertor, or the wrong value for a cap, or a bad cap.
sneak-thief
@DozenCrows - If all the DAC channels are getting a blip, then as cnicht suggested, check to see if the 5V output of U16 is also doing the same thing.

If not, another possibility could be something messing with the DAC's 3.3V analog reference voltage supply.

Start by checking all the components around U10
DozenCrows
Just checked it with the scope, and this noise is present continually on the 5V supply.

The DAC outputs are fine; the noise appears on the clock out, reset out and gates - which all ultimately derive from the 5V via various means (3904, shift register).

Here's what it looks like:

[/img]
Altitude909
toneburst wrote:
Altitude909 wrote:
toneburst wrote:
..
The fact you've got a little MIDI keyboard plugged into the USB port there suggests it can be used for programming step notes. Is that the case?


Yes, its a bit awkward but yes


OK. In what way is it awkward?


You have to hold the step down while playing the key, not a terribly musical input method
sneak-thief
Anybody else here able to scope their 5V rail and/or gate outputs?

@DozenCrows - I see; I hadn't imagined you testing ripple on the gate outs. I don't think I've ever scoped my gates heh.

What happens when you 1) use the tester firmware and 2) disconnect the OLED and erase the firmware completely?

I also wonder if your Recom 5V switcher is defective. Still waiting for some parts to build mine so I can't scope my rails to compare.
toneburst
Altitude909 wrote:
toneburst wrote:
Altitude909 wrote:
toneburst wrote:
..
The fact you've got a little MIDI keyboard plugged into the USB port there suggests it can be used for programming step notes. Is that the case?


Yes, its a bit awkward but yes


OK. In what way is it awkward?


You have to hold the step down while playing the key, not a terribly musical input method


I think you can persistently select a step with a key combo. But I take your point. Awkward if your keyboard and Performer aren't within easy reach of each other.
DozenCrows
sneak-thief wrote:
@DozenCrows - I see; I hadn't imagined you testing ripple on the gate outs. I don't think I've ever scoped my gates heh.

What happens when you 1) use the tester firmware and 2) disconnect the OLED and erase the firmware completely?


@sneak-thief - I've done all the testing using the tester firmware. I'll try without OLED and no firmware and see. The OLED is powered from 3.3V though.

Once difference with the standard BOM is that I'm using a blue version of the same OLED.
toneburst
DozenCrows wrote:
Once difference with the standard BOM is that I'm using a blue version of the same OLED.


I have blue and yellow OLEDs, and can confirm that both work fine on both my Performer builds, so I think you can discount that.
DozenCrows
toneburst wrote:
I have blue and yellow OLEDs, and can confirm that both work fine on both my Performer builds, so I think you can discount that.


Cool - thank you@toneburst for the confirmation. Does your 5V appear smooth on scope?
toneburst
DozenCrows wrote:
Cool - thank you@toneburst for the confirmation. Does your 5V appear smooth on scope?


Ah.. apologies. I was mixing your messages up with someone else's about another problem.

I don't have a scope, unfortunately. I have heard in the past that OLEDs can cause difficult-to-eradicate noise issues though, unless they're very carefully isolated.

Having said that (and I may be playing Devil's advocate here), if the gate output is below a reasonable trigger threshold for the kind of modules it's likely to be connected to when 'off', and comfortable above it when 'on', I wonder if it's that important if there's some noise.
DozenCrows
Yes, noise on the gate/clock/reset may not be an issue in most situations.

The ripple I see looks to be about 300mV peak to peak, which is higher than I was expecting. In some circumstances that might cause issues - e.g. if using the gate signal as CV into a VCA
Altitude909
DozenCrows wrote:
Yes, noise on the gate/clock/reset may not be an issue in most situations.

The ripple I see looks to be about 300mV peak to peak, which is higher than I was expecting. In some circumstances that might cause issues - e.g. if using the gate signal as CV into a VCA


That's huge. My guess is either you have a supply issue (confirm with another supply/isolate) or there are some components that are not soldered all the way (capacitors)

Its way too slow for switching noise (that will be in the hundred of kHz) and OLED noise doesnt sound/look like that
toneburst
@Altitude909 did you have to change the footprint for the USB socket, to use a part that would fit under the panel with the Thonkiconn sockets?

I've removed all the tabs on my 'PJ' sockets, and (belatedly) worried the original specced USB socket won't fit, now.
DozenCrows
Altitude909 wrote:
That's huge. My guess is either you have a supply issue (confirm with another supply/isolate) or there are some components that are not soldered all the way (capacitors)

Its way too slow for switching noise (that will be in the hundred of kHz) and OLED noise doesnt sound/look like that


This is what I was thinking; thank you for the corroboration!
elodin
Altitude909 wrote:
toneburst wrote:
Nice work @Altitude909!


I farmed it out. oops allergic to kicad


Is there a github page for the edited board?

Other than the jack changes, did anything else change? (i assume the USB changed as well).
Altitude909
toneburst wrote:
@Altitude909 did you have to change the footprint for the USB socket, to use a part that would fit under the panel with the Thonkiconn sockets?

I've removed all the tabs on my 'PJ' sockets, and (belatedly) worried the original specced USB socket won't fit, now.


Yeah, there's a new one that fits, old one is too tall. I'll publish the files once i build and test it
autodafe
Just in case someone else needs it...A "standard" SMD quartz works perfectly (and a TH should too), just bend their legs a bit and solder it to the pads. Attention to the height because it shoud fit behind the LCD...
toneburst
Altitude909 wrote:
Yeah, there's a new one that fits, old one is too tall.


Do you happen to have a part no. for the smaller socket? I have one of the original boards half-populated, and have bent all the tabs of the sockets now, so I'm about to hit a USB socket roadblock in my 2nd build.
toneburst
Of course, I'm assuming it has an identical footprint. It's back to the drawing board, if it doesn't.
Altitude909
No, it doesnt. heres my new BOM

https://docs.google.com/spreadsheets/d/1Ij7JMWFCJfHYTZfkfIStLHR7a7g3nU JE4kBo3WPE5Zw/edit?usp=sharing

Old parts in Red with replacements underneath

Phil made a gigantic cutout in his panels so the old one should be fine with the shortened jacks
toneburst
Altitude909 wrote:
Phil made a gigantic cutout in his panels so the old one should be fine with the shortened jacks


Ah, OK. So the socket can poke through the hole in the panel... that's good.

Can't source the shorter socket here in the UK, anyway, and it looks like it would be offset in relation to the panel cutout, even if I did manage to hack the socket to fit the footprint of the old one.
Altitude909
toneburst wrote:
Altitude909 wrote:
Phil made a gigantic cutout in his panels so the old one should be fine with the shortened jacks


Ah, OK. So the socket can poke through the hole in the panel... that's good.

Can't source the shorter socket here in the UK, anyway, and it looks like it would be offset in relation to the panel cutout, even if I did manage to hack the socket to fit the footprint of the old one.


It needed to be rotated 180 degs to match the panel cut out
patrickod
my PCB and panel order arrived so I have sets left spare which I'm hoping to trade or sell on to people. I'm USA based and should be able to ship most places. The panels are 1.6mm aluminium with black silkscreen and white text as shown.

Really excited to start building this, as I hope it's going to be the sequencing heart of my first rig which I'm making all DIY.
gbiz
I've had a few people email me asking for the design file i used to create my Performer windows. You can download a zipfile containing it plus a readme detailing how to upload it to Ponoko/Razorlab from here.

This is CC-by-SA-NC, so this is for personal use only. As you'll likely have some windows spare (this P1 sheet gets you 12 windows), i appreciate you'll need to sell any extras on. It'd be nice if you add a small amount to the price of each so you can donate the cost of a decent beer/coffee to Simon at his Performer donation page.

I haven't rounded the corners on these windows, so you might need to do a minimal amount of filing/sanding of the corners to get the window to fit. I personally prefer that, they don't require any adhesive to hold them in place.

OB1 has tested this file at Razorlab & the result works for him.
westlicht
I've only really discovered this build thread today. It seems demand for my module is pretty big, cool to see so many people building one! Anyway, I would like to clarify some things:

There seems to be a debate about some of my choices with regard to the mechanical design, e.g. the height of frontpanel components.

First of, I was a bit torn between the jacks I use and the Thonkiconns. At the time of designing this, there was no matching stereo jack to go with the Thonks. And I still think mechanically the jacks I've selected are way better. The stereo jacks being lifted off the PCB was a tradeoff I had to make. In terms of the buttons not sticking out enough. I think the current design, when using a 1.6mm frontpanel is about perfect. I don't like the buttons to stick out too much as it prevents you from sliding over the buttons to press a whole row. I do mount the LEDs almost flush to the panel, so they don't stick out much. This is what's best in my taste, but of course your mileage may vary. It's cool that there is a redesign with the Thonks (but please, share the files publicly, otherwise you are violating the license), but I built modules having two jacks sharing the GND pin and I didn't like it. It's quite fiddely to mount these jacks.

There were some complaints about the footprints. I did make the mistake to only use hand-soldering footprints on the 0603 components. The 0805 are a bit small, I agree. But I never found it to be annoying. The ICs I find are easy to solder and I have not really seen footprints that are much larger in other DIY builds. But I will consider changing these for a revision 1.1.

There was also a suggestions to make the MIDI IN jack type agnostic. It's kind of a neat idea, but I'm not sure it's worth the extra components. I think we are about to converge to the new standard (finally), so give it a little time and type B will be gone anyway.

I was also a bit shocked to see the ripple on the 5V of that one build, but I don't think it's due to my design. I've never seen ripple like that with different PSUs.

I will do a revision 1.1 for DIY, but I will have to put in some thought first as to what things I will change. It's a lot of work, and I also want to do more work on the firmware and eventually work towards a commercial release.

Anyway, hope that clears some things up. Happy building everyone! And happy performing later on!!

PS: Please use the github issue pages to inform me about bugs or feature requests. I will not take everything into account, but it helps me to keep priorities organized.[/quote]
pixelforest
Altitude909 wrote:
DozenCrows wrote:
Yes, noise on the gate/clock/reset may not be an issue in most situations.

The ripple I see looks to be about 300mV peak to peak, which is higher than I was expecting. In some circumstances that might cause issues - e.g. if using the gate signal as CV into a VCA


That's huge. My guess is either you have a supply issue (confirm with another supply/isolate) or there are some components that are not soldered all the way (capacitors)

Its way too slow for switching noise (that will be in the hundred of kHz) and OLED noise doesnt sound/look like that


for an additional data point, zooming in on my clock out pulses i'm seeing a ripple of 64mV @ 276.24kHz. on the 5V pin of the USB port i'm seeing a ripple of ~100mV @ ~300hZ
toneburst
Good to see you here @westlicht! I'm sure I can speak for everyone here when I say a big "thank you" for releasing this great project into the public domain.
Altitude909
Welcome Simon!

westlicht wrote:
There was also a suggestions to make the MIDI IN jack type agnostic. It's kind of a neat idea, but I'm not sure it's worth the extra components. I think we are about to converge to the new standard (finally), so give it a little time and type B will be gone anyway.
...

Its not even a question of A or B, it enables you to use male or female connections. Its one extra part.

My boards came in yesterday and I'll post the files once they check out.
pixelforest
in case anyone else is like me and buying a launchpad just to use with this: get a launchpad mini. the documentation at this point all just says "launchpad" which suggested to me that meant original launchpad, but that is not the case. I've tried recompiling with an additional entry into the device registry (\src\apps\sequencer\ui\ControllerRegistry.h) for the original launchpad (device id 0x000E) and it does not work. the repo has two entries in the controller registry with ids 0x0036 and 0x0036 which both are commented "launchpad mini" so the mini should work.
Roni
westlicht wrote:

There seems to be a debate about some of my choices with regard to the mechanical design, e.g. the height of frontpanel components.

[/quote]

I don't mind the chosen jacks. I was concerned that such a large module would seem flaky with a pcb panel, but it feels rock solid and I suspect the reinforced jacks contribute to that. It's difficult to find aluminium at 1.6mm though, so that's a shame.

Thank you so much for this unit. I'm getting a lot of enjoyment out of it.

.
toneburst
pixelforest wrote:
in case anyone else is like me and buying a launchpad just to use with this: get a launchpad mini. the documentation at this point all just says "launchpad" which suggested to me that meant original launchpad, but that is not the case. I've tried recompiling with an additional entry into the device registry (\src\apps\sequencer\ui\ControllerRegistry.h) for the original launchpad (device id 0x000E) and it does not work. the repo has two entries in the controller registry with ids 0x0036 and 0x0036 which both are commented "launchpad mini" so the mini should work.


The original Launchpad isn’t MIDI Class-Compliant, which is why it’s not supported, I believe. The Mini works fine.

Adding support for the 2nd-generation Launchpad S probably wouldn’t be too hard, I imagine.

The full-size Launchpad Mk.2 has RGB pads, and may also work with some tweaking, though there would be a bit more work involved in adding support for a wider range of pad colours. It would be pretty cool to be able to use, say, a different colour-palette for each track.
DozenCrows
When I've sorted my ripple, I'll test with my Launchpad S and report back.
toneburst
DozenCrows wrote:
When I've sorted my ripple, I'll test with my Launchpad S and report back.


I think the device has to be recognised by it's manufacturer and model IDs to work.

There are two devices listed in this file, and both are tagged in the comments as "Novation Launchpad Mini". The model numbers are consecutive, so I'd be surprised if they weren't Launchpad Mini and Launchpad Mini Mk.2.

As I understand it, any non-recognised controller (even if it has the right number of pads and other buttons) is ignored, so it's unlikely the Launchpad S will work, out of the box.

I'm sure adding support for it would be very easy, though, as I think the S (but not the original full-size Launchpad) and Mini are functionally identical.

I much prefer the smaller form-factor of the Mini models, but would love to see a Mk.3 version with full RGB pads. I'm surprised Novation hasn't already produced one.
merlin
I'm just about to order the parts for this build, has anyone used the white display as opposed to the yellow - see here.

The current consumption is slightly higher which I can't see as a problem; any other issues anyone thinks I may encounter?

Thanks,

Karl
pixelforest
toneburst wrote:
DozenCrows wrote:
When I've sorted my ripple, I'll test with my Launchpad S and report back.


I think the device has to be recognised by it's manufacturer and model IDs to work.

There are two devices listed in this file, and both are tagged in the comments as "Novation Launchpad Mini". The model numbers are consecutive, so I'd be surprised if they weren't Launchpad Mini and Launchpad Mini Mk.2.

As I understand it, any non-recognised controller (even if it has the right number of pads and other buttons) is ignored, so it's unlikely the Launchpad S will work, out of the box.

I'm sure adding support for it would be very easy, though, as I think the S (but not the original full-size Launchpad) and Mini are functionally identical.

I much prefer the smaller form-factor of the Mini models, but would love to see a Mk.3 version with full RGB pads. I'm surprised Novation hasn't already produced one.


got my hands on a friend's broken launchpad mini (not a mkII) and got it working on a computer long enough to pull the device id, its 0x0036 which leads me to believe that neither id that is already supported in the repo corresponds to mkII mini. it's simple enough to add lines to the controller registry but i'm curious if the mapping would be drastically different on the newer/other iterations or if there are other issues that would prevent a simple registry change from allowing base functionality on other launchpads.

if anyone has other (class compliant) launchpads but isn't set up for compiling the firmware i'm more than happy to provide hex files to test this. I just don't have other launchpads around. it'd be good to figure out the full range of devices that will already work if we just add them.

@simon - will you accept pull requests for stuff like added usb midi devices?
toneburst
LP Mini Mk.1 and 2 are functionally identical, I believe.
pixelforest
toneburst wrote:
LP Mini Mk.1 and 2 are functionally identical, I believe.


cool, so my offer stands if someone needs help testing firmware with a non supported device id. i just don't have the launchpads to test with myself.
DozenCrows
Done some further testing on the 5V ripple on my build with the following findings:

- Without the OLED connected, there is no ripple.
- With OLED connected:
-- Small ripple on the 12V direct from the power connector
-- More pronounced ripple on +12V after the protection diode (pin 1 of 5V converter)
-- Same ripple on +5V
-- No ripple on output of either 3.3V converters or on -12V
-- Ripple frequency is around 330 Hz
-- Size of ripple on +12V and +5V is around 300mV

My current hypothesis is that the additional current draw of the OLED is exposing a problem with the +V side of the supply (either +12 or +5) going through the module.

I've not seen this ripple powering other modules from the PSU.
DozenCrows
I have also tried powering the module from a different power supply (almost empty Rackbrute) and still get the ripple.
westlicht
Altitude909 wrote:
Welcome Simon!

westlicht wrote:
There was also a suggestions to make the MIDI IN jack type agnostic. It's kind of a neat idea, but I'm not sure it's worth the extra components. I think we are about to converge to the new standard (finally), so give it a little time and type B will be gone anyway.
...

Its not even a question of A or B, it enables you to use male or female connections. Its one extra part.

My boards came in yesterday and I'll post the files once they check out.


I don't get the male/female thing. MIDI cables are pin-to-pin so how would you switch them? Or do you mean of someone uses a stereo jack cable that has leads crossing? I don't get it. But looking at the schematics, it actually is just one part, didn't realize the BAT54S were dual schottkys. So I might just add that for revision 1.1, don't see any downsides to it really. Thanks for sharing!
westlicht
Quote:
@simon - will you accept pull requests for stuff like added usb midi devices?


I will have a look at it and then decide smile
Altitude909
DozenCrows wrote:
I have also tried powering the module from a different power supply (almost empty Rackbrute) and still get the ripple.


I really think there is something wrong with your build. I just checked mine and I dont see anything like you describe, my noise floor is about 20 mV on all the CV out/clock/etc
toneburst
westlicht wrote:
But looking at the schematics, it actually is just one part, didn't realize the BAT54S were dual schottkys. So I might just add that for revision 1.1, don't see any downsides to it really. Thanks for sharing!


That's cool! I was quite dismayed to find that there were two different standards for TRS MIDI (and that the adapter that came with my Beatstep Pro was the wrong kind).
toneburst
westlicht wrote:
Quote:
@simon - will you accept pull requests for stuff like added usb midi devices?


I will have a look at it and then decide smile


Does it currently support both Launchpad Mini models?
westlicht
Quote:
Does it currently support both Launchpad Mini models?


I only tested with the MK2. Not sure about the MK1 and the other launchpads. The first one will not be supported as it is not conforming to USB MIDI. The Pro will not work because it definitely uses a different mapping. The other ones I think should work but I didn't find info about the PID for those, which I use for auto-detection. If they use the same, it should hopefully just work. In any case, if someone wants to send other launchpads to me for integration feel free to do so. I would also look into supporting the APC Mini if someone wants to lend me one or donate one to me.
toneburst
westlicht wrote:
Quote:
Does it currently support both Launchpad Mini models?


I only tested with the MK2. Not sure about the MK1 and the other launchpads.


The LP Mini Mk.1 I have works fine. Maybe it has the same ID as the Mk.2.
I always thought it was a bit cheeky of Novation to call it th "Mk.2", as I think all they did was change the packaging, and the colour of some of the text on the front.

Quote:
I would also look into supporting the APC Mini if someone wants to lend me one or donate one to me.


That's cool. I'm planning to get one of these, myself, so might be able to lend it to you for a while, if it would help.
Roni
westlicht wrote:

I think we are about to converge to the new standard (finally), so give it a little time and type B will be gone anyway.


toneburst wrote:

That's cool! I was quite dismayed to find that there were two different standards for TRS MIDI (and that the adapter that came with my Beatstep Pro was the wrong kind).


Perhaps it's the tyranny of distance suffered by the country that I live in, but I find it mildly amusing that I have no shortage of factory-made Type B adapters lying around, yet here I am soldering together some Type A's, the new "Standard", to use with my Performer. God bless the MIDI org.

.
toneburst
Launchpad integration is now documented.
https://westlicht.github.io/performer/manual/
pixelforest
toneburst wrote:
Launchpad integration is now documented.
https://westlicht.github.io/performer/manual/


yeah just saw that on github, simon is really going full-force on updates and improvements which is absolutely fantastic.

i have a launchpad mini mk2 arriving today, in theory.
toneburst
He’s definitely on a roll.

The enhancements in 0.1.9 make working with the Launchpad even better!

Keep up the great work, @westlicht!
DozenCrows
Well, my 5V ripple saga continues :(

Reflowed joints on C5, C13, C46 and U16 - still get ripple.

Replaced U16 and C13 - still get ripple.

Reflowed other components on +5V net - still get ripple.
DozenCrows
Further update on 5V ripple.

So far I've been running with the test firmware. I flashed the latest sequencer firmware, and had a play around. The ripple varies from not present to as bad as my previous post with a picture. It seems to depend on what the OLED display is showing.

Ripple is at its worst on the steps page.
sneak-thief
@Dozencrows - some random thoughts:

1. I wonder if C5 should be higher, eg. 47uF or 100uF to handle the ripple on the 5V rail?

2. did you check the continuity to C24? If the OLED has some serious voltage swings that overstep what 22uF cap on the digital 3.3V rail can smooth, I wonder if the 5V rail load would be hit more than it could smooth?

On that note, soldering a 47uF or 100uF here could be worth checking.

3. An experiment (aka Drastic Measures): take out the 5V regulator altogether (U16) and send 5V from your psu to U16 V-Out.

4. Did you try a different power cable between the Performer and your distro board?

What's your grounding situation like? Shitty ground connections could also cause ripple like that.
DozenCrows
Thank you again for the extra suggestions, @sneak-thief!

C25 continuity is good, and joints look fine.

Different power cabling makes no difference.

Grounding is ok. This noise is only present when the OLED is on the board and is influenced noticeably by OLED activity. If grounding were an issue, I would expect it to be present without the OLED.

I hacked up a temporary shield out of tin foil and kapton tape and tried that behind the OLED both grounded and floating; also made no difference. Not very scientific, though.

Before I do any more desoldering & resoldering, I'm going to try another OLED. Less risky and messy that way smile
sneak-thief
@Dozencrows - which colour OLED are you using? If it's the white one, that explains everything (max 250mA instead of 146mA for the yellow/blue).
DozenCrows
sneak-thief wrote:
@Dozencrows - which colour OLED are you using?


I'm using a blue one. IIRC that draws a bit more than the yellow (10mA?), but not as dramatically much as the white.
merlin
sneak-thief wrote:
@Dozencrows - which color OLED are you using? If it's the white one, that explains everything (max 250mA instead of 146mA for the yellow/blue).


According to the data sheets, you draw 146mA with the display at 50% brightness for the yellow or blue. The white display is 9mA higher. The differences in the figures at 100% brightness are similar (typically 250mA).

I checked the numbers as I want to use the white display.

Can you adjust the display brightness via the software?

Karl
Altitude909
merlin wrote:
sneak-thief wrote:
@Dozencrows - which colour OLED are you using? If it's the white one, that explains everything (max 250mA instead of 146mA for the yellow/blue).


According to the data sheets, you draw 146mA with the display at 50% brightness for the yellow or blue. The blue display is 9mA higher. The differances in the figures at 100% are similar (250mA).

I checked the numbers as I want to use the white display.

Can you adjust the display brightness via the softwear?

Karl


I think you're both barking up the wrong tree here. The current draw isnt anywhere near those limits, I just measured mine and its 72/12 mA total for the whole device, I can push it to 125 on the + on certain pages AND with launchpad plugged in but that's still a long ways off from the limit of the OLED
sneak-thief
@Altitude909 - of course the OLED won't reach max current as none of the screens will be fully lit.

Clearly I'm barking up the wrong tree if he's using a blue OLED. However, the extra 71% current usage of the white one *might* have explained it.

It is an interesting mystery when something is polluting the power rails so heavily, particularly if it's a device after a 3.3V regulator that's hammering said regulator's incoming 5V rail.

For example, last week I had a 4ms RCD/SCM combo that I lazily connected to my psu's 12V/GND using flimsy low-gauge dupont wires. All the other modules on that PSU were spitting out noise that corresponded to the RCD/SCM LEDs blinking.

Once connected properly (well, if you call eurorack ribbon cables "proper" lol), all noise was gone.

If it were me and I was at this point, I'd also try connecting the OLED's 3.3V in (pin 2) to a separate 3.3V supply.
toneburst
sneak-thief wrote:
Clearly I'm barking up the wrong tree if he's using a blue OLED. However, the extra 71% current usage of the white one *might* have explained it.


I was considering the white OLED, too, so interested to see if this is the source of the issue.
Altitude909
toneburst wrote:
sneak-thief wrote:
Clearly I'm barking up the wrong tree if he's using a blue OLED. However, the extra 71% current usage of the white one *might* have explained it.


I was considering the white OLED, too, so interested to see if this is the source of the issue.


I have one here that I intend to swap my yellow out with so I'll be able to look at this soon
sneak-thief
I have a white one but haven't had the time to assemble it yet. I'll also scope it and report back.
toneburst
Cool. I’m interested to know how white it is, too. I’ve seen nominally “white” OLED displays that are distinctly bluish, and others that look yellowish.

I’m thinking it might be cool to add a coloured lighting gel sheet over a white display, to get any colour I want, but would work best if the display is pure white, and quite bright.
toneburst
toneburst wrote:
westlicht wrote:
I would also look into supporting the APC Mini if someone wants to lend me one or donate one to me.


That's cool. I'm planning to get one of these, myself, so might be able to lend it to you for a while, if it would help.


Just bought an APC mini on eBay smile

The offer stands, @westlicht.
DozenCrows
DozenCrows wrote:
When I've sorted my ripple, I'll test with my Launchpad S and report back.


Haven't sorted the ripple, but have now built the firmware and tried out the Launchpad S USB IDs (0x1235, 0x0020) - it pretty much works, but the LEDs are dim. Am looking further into that now.
DozenCrows
The colour codes in the Performer firmware are currently either 0 or 1 for the red and green LEDs in the Launchpad. Launchpad S actually has four levels per LED, so by temporarily setting the values to 0 or 3, I get full brightness.
toneburst
DozenCrows wrote:
The colour codes in the Performer firmware are currently either 0 or 1 for the red and green LEDs in the Launchpad. Launchpad S actually has four levels per LED, so by temporarily setting the values to 0 or 3, I get full brightness.


LP Mini also has dimable LEDs. I'm not sure the APC Mini has, though. :(
sneak-thief
APC Mini:

Quote:
For the grid, the note on velocity values are:
00 = off
01 =green
02 = green blink
03 = red
04 = red blink
05 = yellow
06 = yellow blink

For the single color buttons, the values are:
00 = off
01 = on
02 = blink
toneburst
I wonder if that will make it less useable, as a Pweformer control-surface.
sneak-thief
@toneburst - where/when are dimmed leds used with the Launchpad?
toneburst
sneak-thief wrote:
@toneburst - where/when are dimmed leds used with the Launchpad?


Not sure. I'm not sure they currently are, but I can think of situations where they could usefully be used.
Altitude909
Ok, new panel is in and everything is on point alignment wise. Just need to wrap the build and I'll put the files up in my repo. This is the new 10mm tall USB with the PJ398 jacks installed


toneburst
Looks good!

You're presumably going to use Thonkiconns for the rest of the jack sockets.

Have you managed to source a header that's low-profile enough to allow the display to sit comfortably under the panel?

That was my primary concern about lowering the panel, I have to say, and what made to decide to instead attempt to raise the buttons slightly, on my 2nd Performer build.
Altitude909
Clears fine as is. Just need to use a 2mm thick window
toneburst
Altitude909 wrote:
Clears fine as is. Just need to use a 2mm thick window


Good to know smile
Altitude909
toneburst wrote:
Altitude909 wrote:
Clears fine as is. Just need to use a 2mm thick window


Good to know smile


This is with a 5mm spacer and a 3mm thick window. It was maybe enough with the 6mm/3mm window but too close for comfort when you add in acrylic tolerances etc and 2mm clear is everywhere anyway (except here, hence the 3mm thick one)

toneburst
And that was with the display header from Simon's original BOM?
Altitude909
toneburst wrote:
And that was with the display header from Simon's original BOM?


*edit*

i didnt even consider that. I'd never use a header for something that you attached once, especially a 4 dollar a pop one.
toneburst
That's good to know..

I didn't know it was possible to get even more low-profile headers (thought the specced one was pretty slim).

I may have to try and track down a 5 or 6mm one, in that case.

Thanks for the tip!
toneburst
Accidental duplicate.
Altitude909
reread my post smile, it wont clear, 6.34mm tall if you mount it like youre supposed to. You could probably mount the male header on top and make it work if you MUST use it
toneburst
Ah... I hate desoldering (too many bad experiences with lifted pads and horrible bodged repairs), so I'd much prefer to mount the display in a socket, to be honest.
Altitude909
toneburst wrote:
Ah... I hate desoldering (too many bad experiences with lifted pads and horrible bodged repairs), so I'd much prefer to mount the display in a socket, to be honest.


Then just mount it on the top of the display, solder underneath and trim off the extra
sneak-thief
Altitude909 wrote:
toneburst wrote:
And that was with the display header from Simon's original BOM?


*edit*

i didnt even consider that. I'd never use a header for something that you attached once, especially a 4 dollar a pop one.


There's no way I'd ever solder an OLED to a board after Crystal Fonts sold me 6 defective OLEDs that died within 2 months. I don't trust the OLED lifespan as of yet - even if it's ten years, that's too short. Also, most displays in commercial synths are socketed.
Altitude909
sneak-thief wrote:
Altitude909 wrote:
toneburst wrote:
And that was with the display header from Simon's original BOM?


*edit*

i didnt even consider that. I'd never use a header for something that you attached once, especially a 4 dollar a pop one.


There's no way I'd ever solder an OLED to a board after Crystal Fonts sold me 6 defective OLEDs that died within 2 months. I don't trust the OLED lifespan as of yet - even if it's ten years, that's too short. Also, most displays in commercial synths are socketed.


Not sure I would agree with that. I've been converting/changing displays for this for the better part of 10 years now and have yet to fine one commercial device that had a socketed display.

I think your problem is the fact you dealt with crystalfontz, they are garbage peddlers. Not that brand name displays (which technically means 2 different manufacturers since they are all CMd from two places) dont have bad parts but out of probably 200 different displays from winstar and raystar, I've seen one bad one out of the box and that's not a big enough number to worry about.
sneak-thief
I guess I'll reserve my judgment until these OLEDs from the last 4 years reach 10 years old (speaking as an ex-synth repair guy)
toneburst
Simon has flagged up 5V power ripple issues in the notes for the latest release.

I've personally never had any issues using my Launchpad Mini with Performer, but have a suspicion my new APC Mini may have crashed out a couple of times, while connected.

https://github.com/westlicht/performer/releases/tag/v0.1.11
Altitude909
Yeah, I havent noticed anything either but I havent really gave the launchpad a good workout either.

I did finish the new board and everything is perfect. I'll post pics shortly.

Also looked at the white and yellow displays and the current consumption difference is negligible, maybe 10 mA. I did a quick and dirty comparison on the bench supply only and that was the difference when on the same page.

Also took some numbers on the clock output with the white and yellow. The gate outs are much quieter so the clk out is probably a better place to measure the background noise..

so Yellow OLED CLK Out:


yellow OLED Gate out for comparison


White OLED Clk out



and I also looked at the noise vs the output of the 5V Reg so can see the switching noise out of the reg is what you are seeing on the clock (note the p2p output V from the reg, still no where near the .3V that was reported, spec is 50 mV IIRC)



Altitude909
and pics of the build:









sneak-thief
fwiw the Recom 5V regulator's specsheet says max 120mVp-p ripple @ 20MHz.


They also have a low ripple version (20-30mVp-p) for more than twice the price:

https://www.mouser.com/ProductDetail/RECOM-Power/R-7825-10?qs=XF8hdbuH JAXTOhvC8te%252B4A==
Altitude909
yeah, the E parts are pretty noisy but at the end of the day, if its only there to supply power to the USB, its plenty fine since the target device should be appropriately filtered for USB power anyway. Next time I have mine apart i'll measure it again with the launchpad plugged in and see what it looks like and how much it pukes into the other power rails.

I put the rest of my boards up in the store.

https://michigansynthworks.com/collections/circuit-boards/products/per former-sequencer-pcb-pj398-366-edition

I wont be restocking so when they are gone they are gone until Phil snatches them.
Boris_Yeltsin
Man, I really wanted one of those and they are gone already lol
westlicht
@Altitude909 nice work! Great to have a proof of concept for the Thonkiconns smile I will definitely consider it for the revision 1.1. There are a few other small changes I plan to do.
Altitude909
westlicht wrote:
@Altitude909 nice work! Great to have a proof of concept for the Thonkiconns smile I will definitely consider it for the revision 1.1. There are a few other small changes I plan to do.


The only sticking point is the socket on the display, it wont clear as is. If you want to keep that a different solution will be needed. It is probably a good idea for a kit since there are lots of places for people to screw up under the display and I can already see the unpleasentness of people murdering a $40 display trying to get it off to fix something underneath.
toneburst
Altitude909 wrote:
The only sticking point is the socket on the display, it wont clear as is. If you want to keep that a different solution will be needed. It is probably a good idea for a kit since there are lots of places for people to screw up under the display and I can already see the unpleasentness of people murdering a $40 display trying to get it off to fix something underneath.


A couple of socketing options for the display that might work:

• Put the pin-header in upside-down, with the plastic at the top of the display board and solder on the bottom
• Extract the pins from the male header with pliers, and stick them into the female header, stick the display onto them and solder from above

I haven't tried either method, so don't know if they will lower the display enough to clear the lower panel. They will lower the display by the depth of the plastic part of the male pin-header.
Altitude909
Either should work but both are a little hacky. I like the stick the male pins in the socket idea

I measured all this out today:

Female pin socket + Male Header = Min 6.34mm
Display + 7mm spacer = 10.4ish mm

so really, you need to lose a half mm on the header/socket combo to get it under 6mm
westlicht
@Altitude909 please share all the files (KiCad project) not just Gerbers, as this is pretty much useless for any derivative work. The license requires to share all changed material, hope you understand.
Altitude909
westlicht wrote:
@Altitude909 please share all the files (KiCad project) not just Gerbers, as this is pretty much useless for any derivative work. The license requires to share all changed material, hope you understand.


Working on it..
DozenCrows
A further data point on my ripple-y build - I have tried it with a different blue OLED and with a yellow OLED, but still get the same ripple behaviour as before. It is most obvious on the steps page (Vpp ~240mV) and with a Launchpad S connected and all grid buttons lit it is Vpp ~360mV.

There's one ripple of around 300 to 500 Hz, and another higher frequency around 10x higher when the Launchpad has full grid lit.
Altitude909
DozenCrows wrote:
A further data point on my ripple-y build - I have tried it with a different blue OLED and with a yellow OLED, but still get the same ripple behaviour as before. It is most obvious on the steps page (Vpp ~240mV) and with a Launchpad S connected and all grid buttons lit it is Vpp ~360mV.

There's one ripple of around 300 to 500 Hz, and another higher frequency around 10x higher when the Launchpad has full grid lit.


Try swaping the Recom for a 7805
toneburst
@Altitude909 what are those LED spacers?
westlicht
@Altitude909 thanks for putting the files up! I will do a revision 1.1, just figured out a way for placing the Thonks without having to put two ground leads into one hole.
Altitude909
toneburst wrote:
@Altitude909 what are those LED spacers?


The best standoffs ever:
https://www.mouser.com/ProductDetail/749-ERM1-300
mongrol
Great work guys. I have a v1 set here ready to build but think I'll wait for the design to settle. Looks like a killer sequencer.
westlicht
@Altitude909 I was just checking your adjusted USB footprint. Is this working? Because it seems like the pins are reversed.
secrethero303
westlicht wrote:
@Altitude909 I was just checking your adjusted USB footprint. Is this working? Because it seems like the pins are reversed.


Comparing the 2 data sheets it looks like he flipped the footprint 180° so the pins match. I’ll let @Altitude909 confirm though.
Altitude909
secrethero303 wrote:
westlicht wrote:
@Altitude909 I was just checking your adjusted USB footprint. Is this working? Because it seems like the pins are reversed.


Comparing the 2 data sheets it looks like he flipped the footprint 180° so the pins match. I’ll let @Altitude909 confirm though.


Yeah, its flipped and it doesnt have the 5th pin
toneburst
@westlicht can I suggest you move the card slot a touch further away from the USB socket, in v1.1? It’s fiddly to insert or remove the card when there’s a USB cable plugged in, I find.
westlicht
@toneburst It would be nice if it was a little less cramped yes, however, it wont happen for two reasons. 1st, I would like revision 1.1 to be compatible with the old frontpanel layout. Second, this area is really dense already and there really is no space to move stuff around much.
Lemmy
Can any users comment on how this compares with Beatstep Pro, features-wise and in terms of usability?

I like my Beatstep Pro, but this OLED looks like it might make the workflow much clearer... and I like the potential for customisation that open source brings.

(Is there a thread about this in a non-DIY forum?)
forestcaver
Lemmy wrote:
Can any users comment on how this compares with Beatstep Pro, features-wise and in terms of usability?

I like my Beatstep Pro, but this OLED looks like it might make the workflow much clearer... and I like the potential for customisation that open source brings.

(Is there a thread about this in a non-DIY forum?)


Much nicer in almost every way! Except not as thin, flat or white! Not used my bsp since I built this.
sneak-thief
For anyone considering red/blue LEDs instead of red/green ones:

I tested 3 different brands of 3mm diffused red/blue and despite experimenting with different resistors, if you set the LED brightness to a reasonable level it's often hard to distinguish between red vs red+blue.

The diffusion is also often troublesome with the blue being concentrated along one edge of the LEDs from some manufacturers.

Note: I didn't try that expensive brand, SunLED from Digikey that was mentioned earlier in this thread.
toneburst
westlicht wrote:
@toneburst It would be nice if it was a little less cramped yes, however, it wont happen for two reasons. 1st, I would like revision 1.1 to be compatible with the old frontpanel layout. Second, this area is really dense already and there really is no space to move stuff around much.


Fair points.
toneburst
sneak-thief wrote:
For anyone considering red/blue LEDs instead of red/green ones:

I tested 3 different brands of 3mm diffused red/blue and despite experimenting with different resistors, if you set the LED brightness to a reasonable level it's often hard to distinguish between red vs red+blue.

The diffusion is also often troublesome with the blue being concentrated along one edge of the LEDs from some manufacturers.

Note: I didn't try that expensive brand, SunLED from Digikey that was mentioned earlier in this thread.


Thanks for testing those.

Having been using a Performer with the specified LEDs for a while now, I’m quite happy with those, to be honest, so I won’t risk overloading the shift-registers by switching to red/blue for future builds.

Having said that, I’m really happy with the 5mm red/blue LEDs on my Nonlinear Circuits Triple Sloth any Hypster builds. On these, you never see both elements lit at the same time, though, and obviously, with the Performer, it’s important to be able to distinguish this state from others where only one of the two colours is lit.
toneburst
sneak-thief wrote:
if you set the LED brightness to a reasonable level it's often hard to distinguish between red vs red+blue.


I wonder if this might be to do with the relative brightness of the two elements. It could be that one of the colours appears to drown out the other, when both are lit, where this is not the case with the red/green parts.
sneak-thief
@toneburst - I used many combinations of different resistors for both the blue and the red. If you boost the blue enough so that the red becomes magenta, then when the red is off, the blue is too bright.

And we all love blinding blue LEDs on the modular, right?

IMHO 5mm LEDs often have better diffusion characteristics than 3mm ones.

Anyhow, the Performer's shift register outputs are split into the red column and the green column, so tuning is possible if a different suitable bicolor LED is found.
djthopa
Im in smile

djthopa
Hi!

Any one got a spare plastic screen protector for the lcd they want to part from?

I dont feel like ordering one full sheet just for one screen.

Thanks!
PopGoblin
I just finished a build (actually only the thru-hole parts - the SMD parts was taken care of by a New Zealand friend smile ) - and everything seems ok, except Clock Out.

Clock Out sends a constant voltage, no matter what subdivision etc. i setup.

Where should I look for errors? I can't seem to find anything with the through-hole parts...

Help greatly appreciated!
sneak-thief
@PopGoblin - check Q3 (and maybe R10). My guess is that pin 1 or 2 of Q3 isn't properly soldered. I've seen that happen often with ICs in such packages.
PopGoblin
@sneak-thief thank you for your reply!

I checked the Q3 and R10. They look ok to me. I've tried checking everything else... seems fine...
sneak-thief
@PopGoblin - A visual inspection may not cut it here - even checking it with a continuity meter may not suffice because simply pushing on the leg could temporarily reconnect a bad solder joint.

It should be easy enough to just reheat all of Q3's joints.
cnicht
@PopGoblin Are you getting a clock signal at both ends of R19 and have you tried the test firmware?
PopGoblin
@sneak-thief & @cnicht thankyou both for helping out!

I did reheat q3 and a lot of other places - to make sure. I’m going to check your suggestion with r19 @cnicht. I havent run the tester .hex (i haven’t got a jtag flasher - the board came with smd’s soldered and the firmware flashed).

Otherwise, to me, things looks good so its pretty frustating with the high signal continously on at clock out...
sneak-thief
@PopGoblin - actually, I meant R19. Sorry about that. cnicht had it right!

You don't need a JTAG flasher to update the firmware - just put the tester (or firmware) on an SD card: https://westlicht.github.io/performer/manual/#appendix-firmware-update
PopGoblin
allright, thanks @sneak-thief! Both sides of R19 show roughly the same (-0.4 to -0.5'ish), when clock is running or not. Q3 pins are 2.8v, -0.4-5'ish and 0.

I found a hot-air SMD rework station and made sure the R19 was good, as well as the Q3 - even heated the arm chip...

still. Constant high at clock out?

I searched - cant figure out how to run the tester.hex - even putting it on the flash card, it doesn't seem to do anything?

Thanks for all your help!



[img][/img]
sneak-thief
oh, there's no .DAT file for the tester firmware:

"- Download the latest UPDATE.DAT file from https://github.com/westlicht/performer/releases.
- Copy the UPDATE.DAT file to the root directory of the SD card.
- Insert the SD card into the sequencer’s SD card slot."


1. Do you have continuity between the bottom of R19 and pin 28 of the MCU?

Pin 32 is the top-right corner, so pin 28 is 4 pins down from that.

2. The clock pulse is configurable from 1-20ms. Not all multimeters will register such a short blip!! So your R19 readings may not be showing this. Also, if the click out circuit is incorrect, it may be dragging down the MCU's clock output pulse voltage so that you can't notice it. -0.4 to -0.5 around R19 doesn't sound right at all.


3. All I can think of is:

- clock high should be close to 5V, not 2.8V, so clearly there's something amiss with that circuit

- there's no connection between pin 28 of the MCU and R19 - or pin 28 is shorted with pin 29 or 27 (Reset in/out) - can you check if there's a short there?

- R19 or R18 are the wrong values - what codes are on those resistors?

- the 2N3904 transistor is bad or not the correct component - what code does it say on it?
Virusinstaller
I have a problem with a Performer Build. All of the LEDs are on when the module powers on.
When I press PAGE the LEDs do not correctly change. seriously, i just don't get it

Where too from here!
cnicht
@PopGoblin All of sneak-thief’s suggestions are good ideas.

It may be worth using an oscilloscope to view the actual state of the signal dynamically rather than trying to interpret it from a multimeter!
cnicht
Virusinstaller wrote:
I have a problem with a Performer Build. All of the LEDs are on when the module powers on.
When I press PAGE the LEDs do not correctly change. seriously, i just don't get it

Where too from here!


Do the LEDs go through the start-up sequence then remain on?

Have you tried the test firmware?
sneak-thief
@Virusinstaller:

The LEDs are driven by U12 and U13. Start by checking those first.

- Is it possible to test if the gate outs work?
- Do any/all of the buttons work?

If neither gates nor any of the buttons are working, then check out U5 which connects to the gates, LEDs and buttons.
Virusinstaller
Thanks Guys,

Buttons & gates are working, LEDs go through a start up, but not like the module typically working. Otherwise the module appears to be working correctly, except the LEDs. They do react when you press Page, but not as they should. Just a little strange.
I bought some Aliexpress 74HC595, so I have followed up with buying from Arrow.
Not sure that it is as simple as a reflow as I have two with the same behaviour. Which would be very unusual if it was just a reflow problem.
Virusinstaller
Test firmware is a good idea though. I haven't installed that on any builds as of yet, but might check it out for interest & problem solving!
PopGoblin
@sneak-thief; thanks!! pin28 HAD a short... testing for continuity and your hints about that made all the difference... now my clock out is happy-blinky!!

so happy.... smile
sneak-thief
@PopGoblin Rockin' Banana!
forestcaver
Retro.... :-)

djthopa
forestcaver wrote:
Retro.... :-)



Alternative firmware? applause
forestcaver
djthopa wrote:


Alternative firmware? applause


No - easter egg - I was poking around in the code, trying something out and saw this.... :-) (First time I had sat down to try to build it and work out the code - especially the hardware and drivers etc)
djthopa
Do you know what it does? How do you trigger it? Miley Cyrus
forestcaver
djthopa wrote:
Do you know what it does? How do you trigger it? Miley Cyrus


Yes - it allows you to play asteroids ! (Function keys rotate, thrust, fire and hyperspace)
djthopa
Guinness ftw! thanks for sharing!
OB1
I've got a couple spare... EDIT: all gone now, sorry.

djthopa wrote:
Hi!

Any one got a spare plastic screen protector for the lcd they want to part from?

I dont feel like ordering one full sheet just for one screen.

Thanks!
djthopa
pm! thanks a bunch Rockin' Banana!
Handmedown
After searching I have found no info on the USB problem I am having. I can use the USB from my Akai MPK49 keyboard but not from my Arturia Keystep. Akai keyboard is recognized but not the Arturia.

Is this a bug?
Altitude909
Handmedown wrote:
After searching I have found no info on the USB problem I am having. I can use the USB from my Akai MPK49 keyboard but not from my Arturia Keystep. Akai keyboard is recognized but not the Arturia.

Is this a bug?


Is the keystep class compliant or does it need a driver?
keninverse
keystep is class compliant
djthopa
Finished my second and last one today.

Probably on of the best builds with the interactive bom webpage feature.

Thanks to Simon_ Westlicht for sharing this great project!





edited:

Maybe for those that are using eagle and would like to use the interactive bom function i mentioned (i did not know about it before) for Eagle:

https://github.com/oceanofthelost/InteractiveBOM
toneburst
Look great, djthopa!
elodin
Altitude909 wrote:
No, it doesnt. heres my new BOM

https://docs.google.com/spreadsheets/d/1Ij7JMWFCJfHYTZfkfIStLHR7a7g3nU JE4kBo3WPE5Zw/edit?usp=sharing

Old parts in Red with replacements underneath

Phil made a gigantic cutout in his panels so the old one should be fine with the shortened jacks


Is D42 supposed to be 1N5819HW? It is not on the BOM.
Altitude909
elodin wrote:
Altitude909 wrote:
No, it doesnt. heres my new BOM

https://docs.google.com/spreadsheets/d/1Ij7JMWFCJfHYTZfkfIStLHR7a7g3nU JE4kBo3WPE5Zw/edit?usp=sharing

Old parts in Red with replacements underneath

Phil made a gigantic cutout in his panels so the old one should be fine with the shortened jacks


Is D42 supposed to be 1N5819HW? It is not on the BOM.


BAT54-S. Forgot about those, BOM updated
windspirit
Ok, looks like 1N5819HW-7-F schottkey diodes are out from mouser. They are in stock at digikey but I'd like to avoid multiple orders if possible.

Is this a suitable replacement?

https://www.mouser.com/ProductDetail/STMicroelectronics/STPS1L40ZFY?qs =U58vgkJzipT31AgUgXQl%2Fg%3D%3D

Same footprint, forward current and repetitive reverse voltage. The forward voltage is slightly higher (.5V instead of .45V) and the reverse current is also much lower (35 uA vs 1000 uA). Not sure if that represents a possible safety issue with a diode that is used for clamping.

Here is a link to the 1n5819 for reference:

https://www.mouser.com/ProductDetail/621-1N5819HW-F
keninverse
windspirit wrote:
Ok, looks like 1N5819HW-7-F schottkey diodes are out from mouser. They are in stock at digikey but I'd like to avoid multiple orders if possible.

Assuming you don't want to pay multiple shipping charges you could just get them from Arrow. Free shipping.
sneak-thief
keninverse wrote:
Assuming you don't want to pay multiple shipping charges you could just get them from Arrow. Free shipping.


Arrow don't have any stock - unless you know of a close equivalent.
cygmu
this any good?

https://www.arrow.com/en/products/1n5819hw1-7-f/diodes-incorporated
Altitude909
cygmu wrote:
this any good?

https://www.arrow.com/en/products/1n5819hw1-7-f/diodes-incorporated


Mouser has the F ones too. They will be fine, they are wider and shorter but will fit the footprint
Virusinstaller
Virusinstaller wrote:
I have a problem with a Performer Build. All of the LEDs are on when the module powers on.
When I press PAGE the LEDs do not correctly change. seriously, i just don't get it

Where too from here!


I bought common Anode LEDs rather than common Cathode LEDs. Easy fix!
Don't make the same mistake kids.
pix
I'm stuck uploading the firmware. I have vagrant open in the performer-master folder. I can see the Makefile, Vagrantfile and build folders but none of the commands I usually use with MI stuff works.

Can one of you code wizes share the magic words?
gbiz
pix wrote:
I'm stuck uploading the firmware. I have vagrant open in the performer-master folder. I can see the Makefile, Vagrantfile and build folders but none of the commands I usually use with MI stuff works.

Can one of you code wizes share the magic words?


Easiest way is to follow the upgrade details in the doc (see here)


Performer uses cmake, so building is slightly different. I used this script to build from scratch & upload using "st-flash" rather than openocd. This works on OSX with macports & on Fedora. I don't see why it wouldn't work if you run it in the Mutable vagrant image, aside that provides an old version of the st-link tools so the upload will fail. You'd need to update that.

Code:

git clone https://github.com/westlicht/performer
cd performer
git submodule update --init
make setup_stm32
cd build/stm32/release
make
st-flash write bootloader/bootloader.bin 0x8000000 && st-flash  write sequencer/sequencer.bin 0x8010000 && st-flash reset


If you get a "md5" error from one of the python scripts, change line 1 in scripts/makeupdate to
#!/usr/bin/env python2
forestcaver
1. You could use openocd in the mutable dev environment and grab the prebuilt firmware from the performer github. eg:

openocd -f interface/stlink-v2.cfg -f target/stm32f4x.cfg -c “program performer.hex verify reset exit”

and the same for bootloader.hex

(assuming you use a stlink v2)

Find the compiled hex files here
https://github.com/westlicht/performer/releases



2. You could also use the vagrant virtual machine that Simon uses - the vagrantfile is also in the github and then build and flash it from that virtual machine.

3, You could use the stutility program on windows

4. You could do as gbiz suggests

I’ve done both 1 and 2

On a new build I’d suggest once you confirm it’s running, try the tester hex to confirm and test all the functions to ensure the hardware build is ok
sduck
I've used #3 above, easy peasy.
Fractaltctrl
Hi guys, I have an issue with mine. I tripped on my main power cable and when I turned the power back on, the screen was all scrambled. There is no led startup sequence and the buttons don't activate them either. I tried a different psu, same thing. Checked the voltages at the components referring to this , they seem to check out. What do you guys think? Do you recon the ARM chip is pooped?

sad banana

sneak-thief
The first thing to try is to reflash it with a JTAG unit or USB-serial adapter.
Fractaltctrl
sneak-thief wrote:
The first thing to try is to reflash it with a JTAG unit or USB-serial adapter.


Ok thanks will give it a go. Will the PGM-W-10053 Mini ST-LINK/V2 work fine for this purpose or do I need the bigger PGM-W-5822 ST-LINK/V2?
sneak-thief
I don't see why that programmer wouldn't work - it says it's compatible with STM32 chips. Go for it!
Fractaltctrl
sneak-thief wrote:
I don't see why that programmer wouldn't work - it says it's compatible with STM32 chips. Go for it!

Chugging Beers
pix
firmware uploaded! thanks everyone! we're not worthy
pix
btw, I think this sequencer deserves a better interface, if anyone would be willing to make a new PCB version with double the width to be able to use proper buttons (and in a row of 16) it'd be fantastic. Amazing sequencer!
sneak-thief
pix wrote:
btw, I think this sequencer deserves a better interface, if anyone would be willing to make a new PCB version with double the width to be able to use proper buttons (and in a row of 16) it'd be fantastic. Amazing sequencer!


Isn't that the whole point of connecting a Launchpad?
djthopa
I personally think its fantastic the way it is.

I have tried it with the launchpad, and sure it is more user friendly, but i feel like if you use an external midi controller it kind of takes ways some of the "hands on".... if im using a midi controller for a eurorack device i might as well use a midi sequencer.

I think the performer is very hands on and all modes are very accessible with not too many button combos, but i guess its a matter of taste.

Just my 2 cents
forestcaver
pix wrote:
btw, I think this sequencer deserves a better interface, if anyone would be willing to make a new PCB version with double the width to be able to use proper buttons (and in a row of 16) it'd be fantastic. Amazing sequencer!


All the files are there, why not do it yourself? It wouldnt be hard and would be a good learning opportunity.....
smiler
This arpeggiator mode demo was posted by the author today in the Facebook group:
https://www.facebook.com/performerseq/videos/vb.498075893924338/288727 178704452/?type=3&theater

Pretty slick!
cnicht
It’s the gift that just keeps giving!
pix
forestcaver wrote:
pix wrote:
btw, I think this sequencer deserves a better interface, if anyone would be willing to make a new PCB version with double the width to be able to use proper buttons (and in a row of 16) it'd be fantastic. Amazing sequencer!


All the files are there, why not do it yourself? It wouldnt be hard and would be a good learning opportunity.....


yeah that's fair...but at this point in my life I really have to squeeze all the extra couple of minutes I can find here and there...I don't even have time to play with these things after I build them.
sneak-thief
Regarding red/blue LEDs - I found a combination that works.








I experimented with different resistor values and 5K (4.99k) resistors instead of 1K resistors for R76, R74, R72 & R70 make a very discernible combination of Red, Blue & Magenta. The blue by itself is actually comfortable and not blinding.

Now I can't guarantee that you'll get the exact same batch as me, but these work great:

https://www.ebay.de/itm/NEW-3mm-Dual-Bi-Color-Red-Green-Blue-Super-bri ght-Diffused-Bright-3Pin-Led-diode/262039334653?ssPageName=STRK%3AMEBI DX%3AIT&var=560824316678&_trksid=p2057872.m2749.l2649



I also found a USB to +/-12V power supply that seems to be relatively low-noise based on my little scope (>20mVpp ripple?):

https://www.ebay.de/itm/HIFI-Low-Noise-Single-Voltage-To-Positive-Nega tive-Regulated-Power-Supply-Module/332677482180?ssPageName=STRK%3AMEBI DX%3AIT&_trksid=p2057872.m2749.l2649



I didn't install the Recom 5V regulator at all - instead I passed through 5V directly from the USB psu.

If anybody wants to try this, I highly recommend a top-quality USB psu. I measured the ripple of various ones I had lying around and Samsung USB psu's seem to be the best so far.
toneburst
@sneak-thief that looks great!
toneburst
Does anyone happen to have a spare one of @altitude909s Performer PCBs (with Thonkiconn jack sockets) they'd be willing to sell me?
sneak-thief
@toneburst - thanks! The white OLED is fabulous.

Here's a very basic example of the Performer's ratcheting - it's triggering Braids and a drum loop from Radio Music. TC Fireworx is making the weird scratching/vocal effect:

http://sneak-thief.com/modular/performer%20test%201.mp3
toneburst
@sneak-thief I notice you’ve bent/removed the tabs on the bottom of the sockets. How did you get the usb socket to fit through the panel slot?
sneak-thief
@toneburst - After I bent all the tabs, I took those thin washers that usually go on top of the panel (under the nut) and put them under the panel.
toneburst
@sneakthief hehe.. so you lowered the panel, then lifted it up again...

Sorry for all the questions, but did you manage to socket the display, or did you have to solder it directly to the board?

Did you raise the tact switches up at all?
toneburst
@sneak-thief also, I'd be interested in finding out how the shift-register chips hold up with the red/blue LEDs, given the warnings in the Build Guide about the potential for them failing, if LEDs other than the specced ones are used.

I've ordered 50 of the LEDs you linked to, anyway. They will take a while to arrive, so time to find out if they do work reliably for you wink
djthopa
@sneakthief

Looking great ! applause
sneak-thief
toneburst wrote:
@sneakthief hehe.. so you lowered the panel, then lifted it up again...

Sorry for all the questions, but did you manage to socket the display, or did you have to solder it directly to the board?

Did you raise the tact switches up at all?

...

I'd be interested in finding out how the shift-register chips hold up with the red/blue LEDs, given the warnings in the Build Guide about the potential for them failing, if LEDs other than the specced ones are used.



Chopped and raised. Guilty as charged. Only way to get the round buttons the perfect height. Used 5 nylon washers under the square buttons. I can hear Altitude909 scowling all the way from Michigan grin

The OLED is socketed, nothing changed there at all and there's 1mm of space between the panel and the OLED. OB1 kindly sold me a window that's 3mm high, so it elegantly protrudes similar to my Virus TI razz

Re. shift register & alternate LEDs: I take it the biggest concern is too much current consumption. I just measured the DC current draw of the red/blue LEDs on a breadboard: when set to a similar brightness as on my Performer, the blue used 0.3mA and the red was 0.6mA.

A typical 74HC595 can easily deliver 35mA (it's max rating) and likely twice as much if you want to walk on the wild side: https://electronics.stackexchange.com/questions/286055/how-is-the-curr ent-limited-through-the-leds-in-this-led-matrix-design

Even with all 32x red/blue LEDs on, my estimate is that it's under 30mA. This is ignoring the fact that the Performer is also driving them with PWM, so even less current.

I promise to let you know if I blow up my 595's grin
toneburst
sneak-thief wrote:
Used 5 nylon washers under the square buttons. I can hear Altitude909 scowling all the way from Michigan grin


Hehe..
Where did you add the washers, may I ask? Around the switch actuator, under the button, or did you raise the switches off the board, with the washers underneath?

Quote:
The OLED is socketed, nothing changed there at all and there's 1mm of space between the panel and the OLED. OB1 kindly sold me a window that's 3mm high, so it elegantly protrudes similar to my Virus TI razz


I think I will experiment with lowering the display, as I like the display window to be completely flush with the panel.

Quote:
Re. shift register & alternate LEDs: I take it the biggest concern is too much current consumption. I just measured the DC current draw of the red/blue LEDs on a breadboard: when set to a similar brightness as on my Performer, the blue used 0.3mA and the red was 0.6mA.

A typical 74HC595 can easily deliver 35mA (it's max rating) and likely twice as much if you want to walk on the wild side: https://electronics.stackexchange.com/questions/286055/how-is-the-curr ent-limited-through-the-leds-in-this-led-matrix-design

Even with all 32x red/blue LEDs on, my estimate is that it's under 30mA. This is ignoring the fact that the Performer is also driving them with PWM, so even less current.

I promise to let you know if I blow up my 595's grin


That's reassuring smile
mongrol
I've still got the original v1 pcb and panel. Are these still ok to build? no glaring issues other than the button heights?
sneak-thief
@toneburst - the washers are between the 5 function buttons and the pcb. Before soldering, of course wink

Do you already have a clear window? If not, a 2mm would be ideal and much better than a sunken display.

@mongrol - yeah, V1 is fine. Button heights are a matter of personal preference. The only real important workaround is raising the encoder enough before soldering it in so you can get a nut around the top. I had a thicker nylon washer that did the trick. Maybe 2mm?
Altitude909
if anyone wants the thonk version boards, I'll have more in the shop next week.


(scowling at Sneak)
toneburst
sneak-thief wrote:
@toneburst - the washers are between the 5 function buttons and the pcb. Before soldering, of course wink


I raised all the switches up on my 2nd build. Thought I'd done it OK, but looking at it in the cold light of day, the button cap heights are horribly uneven. It's 100% functional, but doesn't look great :(

Quote:
Do you already have a clear window? If not, a 2mm would be ideal and much better than a sunken display.


I do, though on my #2 Performer, it sticks up fractionally on one side, for some reason.

I have one of these panels on the way from Canada:

https://www.modulargrid.net/e/other-unknown-westlicht-per-former-perfo rmer-alternate-panel

which apparently comes with a display window.

Quote:
The only real important workaround is raising the encoder enough before soldering it in so you can get a nut around the top. I had a thicker nylon washer that did the trick. Maybe 2mm?


I used a thick steel washer I had lying around. Did the job OK. I don't like having the encoder hovering above the board, but it's supported by the nut, and I made sure I used plenty of solder, so I think it will be OK.

Same applies to the MIDI I/O sockets, too, of course.
toneburst
Incidentally @sneak-thief how long did you red/blue LED order take to turn up?

My order (from the same eBay seller you got yours from) is expected on the 26th June, which seems excessive, even for freight shipping from China...
elodin
My Per|former is nearly complete (still need to calibrate and make sure everything is working). I built Altitude909's version with the PJ398SM jacks.

I was concerned about possible errors under the display, so I used the socket in the original BOM. I connected a normal sized header and *carefully* pulled off the plastic bit. I then put the socket with header posts into the pcb, and screwed down the display to solder in the posts. I then gave it a haircut to keep the posts clear from the panel. The image below shows it at this stage.

For extra caution, I added a thin strip of electrical tape on top of the soldered posts.

[/img]
toneburst
@elodin that's cool! I was thinking on similar lines myself. Definitely not going to solder the display directly to the board!

Can I ask what you used to get the plastic strip off? I thought of trying to pry it off with a flat-head screwdriver, but was concerned might damage the display board.
toneburst
Ah, disregard the above. I was assuming you'd already soldered the header to the display board, before attempting to pry off the plastic header strip (which is what I started to do, but them stopped for fear of damaging the expensive display).

When you say you "connected a normal sized header", what did you connect it to, may I ask?
elodin
toneburst wrote:
When you say you "connected a normal sized header", what did you connect it to, may I ask?


I connected 68002-120HLF (or another header strip that I had handy) to 929870-01-20-RA. I then used needle nose pliers to remove the plastic covering. Once removed, I gently pushed down on each of the posts to make sure that they were properly inserted prior to placing the 929870-01-20-RA on the PCB.
sneak-thief
toneburst wrote:
Incidentally @sneak-thief how long did you red/blue LED order take to turn up?

My order (from the same eBay seller you got yours from) is expected on the 26th June, which seems excessive, even for freight shipping from China...


4-5 weeks Dead Banana
toneburst
Thanks @elodin and @sneak-thief.
elodin
If anyone would be willing to send me a display window I will make a $25.00 (CAD) donation to Westlicht in your name.
AlanP
Where is the best place to get both a PCB and panel?
elodin
Depends on which pcb you want. Thonk and Michigan Synth carry the official prototype. If you want the thonkikon version, you may have to print it yourself.
sneak-thief
No - if anybody wants the pcb that uses thonkiconns, Altitude909 is the dude. Just have to wait another week.

https://michigansynthworks.com/collections/circuit-boards/products/per former-sequencer-pcb-pj398-366-edition?variant=19844534829142

Altitude909 wrote:
if anyone wants the thonk version boards, I'll have more in the shop next week.


(scowling at Sneak)


I definitely would have bought the thonkiconn pcb had it been available then. And gladly avoided the ugly kludges!
toneburst
sneak-thief wrote:
I definitely would have bought the thonkiconn pcb had it been available then. And gladly avoided the ugly kludges!


Ditto. Which is why I'm going to end up making 3 of these things...
Altitude909
I've been thinking about how to still use the shorter risers for the display while keeping the socket for easy removal if needed and the simplest way to do it that doesnt change the BOM is to simply pull the pins out of the male header, solder in the female header to the board. Mount the display on the standoffs and then just solder in the male pins one by one using the female header as a jig. They will all sit perfect vertical and even that way..

It's a bit more work but really the least complicated and most effective way to deal with it. I'll post some pics once the boards arrive
toneburst
@Altitude909 hey, get your own ideas, that one was mine wink

Joking.

I haven't got round to trying it myself, having opted last time round for raising the switches, rather than lowering the panel, so you can be the guineapig for that one.
Lemmy
@Altitude909 do you have any more PCBs? If not I may get some fabbed.
Altitude909
Lemmy wrote:
@Altitude909 do you have any more PCBs? If not I may get some fabbed.


They are ordered, be here next week
Altitude909
toneburst wrote:
@Altitude909 hey, get your own ideas, that one was mine wink

Joking.

I haven't got round to trying it myself, having opted last time round for raising the switches, rather than lowering the panel, so you can be the guineapig for that one.


Its the right call. I do the same thing already where there is super low clearance and I need a removable part. I use those machined m/f Chinese millmax clones and can get the height down to almost that of a solder header. The trick is to put Kapton tap between the header and the board when youre soldering in the pins so you dont solder to the header when solder goes down the PTH. That wont be needed here though since the board sits up high
toneburst
Interesting. I've found machined pin headers a bit delicate in the past. Very easy to break them off when removing the part. I guess it's the only way sometimes, though.
elodin
i initially tried to make a machined pin header work found it to be too finicky with the gap between the socket and display. I think the socket in the BOM is fine.
Altitude909
this is what I mean:


toneburst
Nice.
Staticcharge
Just finished building and now I'm at the calibration stage, 0v is fine on all CV outputs but from 0v to +5 and -5 float around th 0.61v mark and the manual dialing in doesn't seem to do much. Jumps around a bit but very little change, any ideas?
In addition, with the meter in the cv1 socket I can change to calibrate cv8 and the meter readings seem to be common to all cv outputs. ie, +5 shows .61v or there abouts and 0v is 0v......hmmmm.
sneak-thief
Check all the pins of the cpu - also, post a hires pic of the area around the DAC. I bet something is up in either of those areas.
elodin
Handmedown wrote:
After searching I have found no info on the USB problem I am having. I can use the USB from my Akai MPK49 keyboard but not from my Arturia Keystep. Akai keyboard is recognized but not the Arturia.

Is this a bug?


Hello Handmedown. Were you able to find a solution to this? I seem to be having the same problem with my Keystep. I will dig out my M-Audio Axiom this weekend and see if it works.
Staticcharge
sneak-thief wrote:
Check all the pins of the cpu - also, post a hires pic of the area around the DAC. I bet something is up in either of those areas.

My bad, dodgy leads on my meter. Calibrated fine once I'd swapped leads. Thanks for the tip though, I used a hot plate method and first I thought I'd fried the DAC.
Staticcharge
This is my first major SMD build using the heat plate method, strategically carrying out the build in three steps. Larger components that need a little more heat was the first step and the final step was the ICs and components that are temperature sensitive.
I managed to sort out my meter lead issue so calibration is all done. Firmware update was done via a ST-LINK v2, because I don't know what I'm looking at with code....much to learn.
It's my second build using a microprocessor.
Once the panel is back in stock at Pusherman the build will be complete.







Additionally, the display header. I found using a standard single row header but puting the longer end into the display PCB works fine, once soldered in place you can trim off the excess then tidy the solder points up.

Warpus
I just finished my Per|former build, but mine does not start up. The display is blank, no LEDs light up, etc.... I've successfully flashed the bootloader.hex and sequencer.hex files as the tester.hex. The display and LED pins are getting power, so I'm not sure what's going on. Any ideas? Thanks!
Staticcharge
Warpus wrote:
I just finished my Per|former build, but mine does not start up. The display is blank, no LEDs light up, etc.... I've successfully flashed the bootloader.hex and sequencer.hex files as the tester.hex. The display and LED pins are getting power, so I'm not sure what's going on. Any ideas? Thanks!

From what I've read and I'm sure others will request..pictures. Can you take pictures of the display headers and check for dry joints. LEDs, Check to make sure you have the LEDs soldered in the correct way, ie short leg to the square pad on the pcb. Additionally, check the soldering around pin 11, 33-36 on your STM32 (U9)
This may help, it may be worth scanning through the earlier replies to see if anyone else has had the same issue.
If you find a solution, post it up here so it may be useful for others.
My fingers are crossed for you..
Warpus
Thanks for the response. Yeah, I’m sure the diodes are all oriented correctly and the LEDs are soldered in with the short leg in the square pad. I was actually able the successfully flash the STM32 as well.

Here are some photos. I’m stumped.



Altitude909
i see probably 5 bad solder joints on U5-U14 in that pic and R36 is straight up missing
sduck
^what he said^. I'd reflow everything on U5 and U6 and U11-14. Can't really see the chips to the right, so don't know.
Handmedown
elodin wrote:
Handmedown wrote:
After searching I have found no info on the USB problem I am having. I can use the USB from my Akai MPK49 keyboard but not from my Arturia Keystep. Akai keyboard is recognized but not the Arturia.

Is this a bug?


Hello Handmedown. Were you able to find a solution to this? I seem to be having the same problem with my Keystep. I will dig out my M-Audio Axiom this weekend and see if it works.


Well I came to the conclusion after deciphering the manual and chatting with others that it is setup for the USB Novation Launchpad and if you get a USB MIDI keyboard working with it you are lucky. I know it does does not say this specifically in the recording paragraph on page 38 because all it says is MIDI keyboard can be used. It does NOT say USB MIDI keyboard. Also there maybe something odd with the Arturia keystep USB protocol and possibly other Auturia products. My Akai USB MIDI keyboard works fine though so thats why I say this about Arturia. I do not have a chance to try other keyboards.
Warpus
Altitude909 wrote:
i see probably 5 bad solder joints on U5-U14 in that pic and R36 is straight up missing


I can't believe I missed that R36! d'oh! Fixed that and reflowed and all good. Thanks! thumbs up
Warpus
sduck wrote:
^what he said^. I'd reflow everything on U5 and U6 and U11-14. Can't really see the chips to the right, so don't know.


Thanks! Yeah, looking back at the photos revealed stuff I just couldn't see with my own eyes! Thanks for having a look and the advice!
t-b
Hi, I had Warpus same problem, flashing stm ok then nothing, stupid I realized that I used a wrong crystal, the one with 4 pads i had in spare, I changed with the good one and magic, it's alive!
Maybe it will help, or help looking to stupid things we do when assembling diy projects.
Staticcharge
That is good news, fresh eyes and the wealth of knowledge found here and it's alive. This forum has god me out of some trouble in the past. Glad it's working for you now though, It's an amazing module and if your not familiar with the Eloquencer or indexing it's well worth flicking through the manual. What it needs is a get started section like the ER-101 manual.
Staticcharge
A question I have regarding the firmware, the area of indexing. Is there a way of changing the not number into related letters, ie 1=C, 2=C# and so on. I'm downloading Vagrant as we speak just so I can have a look and see what I can understand from using Arduino, if anything...
sneak-thief
Staticcharge wrote:
A question I have regarding the firmware, the area of indexing. Is there a way of changing the not number into related letters, ie 1=C, 2=C# and so on. I'm downloading Vagrant as we speak just so I can have a look and see what I can understand from using Arduino, if anything...


Do you have version 0.1.19 installed?
https://github.com/westlicht/performer/releases


Quote:

v0.1.19

@westlicht released this 3 days ago

Improvements

Added arpeggiator to MIDI/CV track (#47)
Show note names for chromatic scales (#55)



"Next release [0.1.19] will have note names again. It takes root note and scales into account. Transposition and octave switching is not taken into account though, as this might lead to pretty weird results when automating those parameters."
Staticcharge
sneak-thief wrote:
Staticcharge wrote:
A question I have regarding the firmware, the area of indexing. Is there a way of changing the not number into related letters, ie 1=C, 2=C# and so on. I'm downloading Vagrant as we speak just so I can have a look and see what I can understand from using Arduino, if anything...


Do you have version 0.1.19 installed?
https://github.com/westlicht/performer/releases


Quote:

v0.1.19

@westlicht released this 3 days ago

Improvements

Added arpeggiator to MIDI/CV track (#47)
Show note names for chromatic scales (#55)



"Next release [0.1.19] will have note names again. It takes root note and scales into account. Transposition and octave switching is not taken into account though, as this might lead to pretty weird results when automating those parameters."


Just run the Update.DAT file?
(EDIT) Scrap that, Just ran the update....Perfect. This sequencer evolves over night....
toneburst
@Altitude909 what height are the display and panel spacers for your Performer PCB version?

They don't seem to be on the BOM.
toneburst
Ah, scrolling back through the (very long) thread, it seems you used 5mm standoffs for the display.

I was being a bit lazy. I will look at the data sheet for the Thonkiconns to work out the best height for the panel standoff, so please ignore the above.
toneburst
Staticcharge wrote:
This sequencer evolves over night....


It certainly does. The updates are coming thick and fast.
Altitude909
toneburst wrote:
@Altitude909 what height are the display and panel spacers for your Performer PCB version?

They don't seem to be on the BOM.


10mm for the board spacer, 6mm for the display, there is a hardware tab on the sheet
toneburst
Ah.. apologies- I completely missed the extra tab.

sorry to waste your time, @Altitude909.

Parcel arrived safely, by the way.
toneburst
Incidentally, if you find the 20-way 3M low-profile female header in the BOM is unavailable at your usual supplier, 2 of the 10-way versions (part no. 929870-01-10-RA) work just as well.

I find it useful when combining multiple female headers to stick a male header in them to hold them together and straight, before soldering them.
Altitude909
ok, boards back in stock.
Boris_Yeltsin
Yay! got the parts in yesterday, cant wait to get the pcb and try this thing out. Thanks Altitude
Staticcharge
toneburst wrote:
Staticcharge wrote:
This sequencer evolves over night....


It certainly does. The updates are coming thick and fast.


I saw your comment on Github regarding chord mode, what's the word on that, seems logical to me.
elodin
Handmedown wrote:
elodin wrote:
Handmedown wrote:
After searching I have found no info on the USB problem I am having. I can use the USB from my Akai MPK49 keyboard but not from my Arturia Keystep. Akai keyboard is recognized but not the Arturia.

Is this a bug?


Hello Handmedown. Were you able to find a solution to this? I seem to be having the same problem with my Keystep. I will dig out my M-Audio Axiom this weekend and see if it works.


Well I came to the conclusion after deciphering the manual and chatting with others that it is setup for the USB Novation Launchpad and if you get a USB MIDI keyboard working with it you are lucky. I know it does does not say this specifically in the recording paragraph on page 38 because all it says is MIDI keyboard can be used. It does NOT say USB MIDI keyboard. Also there maybe something odd with the Arturia keystep USB protocol and possibly other Auturia products. My Akai USB MIDI keyboard works fine though so thats why I say this about Arturia. I do not have a chance to try other keyboards.


Thanks for the reply.

Yes, I came to the same conclusion. The Keystep comes with a midi to TRS cable. Running that cable into the midi in jack is working perfectly for me, and it means that the USB jack is free to plug in a launchpad once I get one.
d.simon
Altitude909 wrote:
ok, boards back in stock.


I searched/read this thread but couldn't quite tell...does the Michigan Synth Works board need a different panel? I noticed some panel gerbers in your repository...
djthopa
I have been thinking the novation launch control would be great for note sequencing. 16 knobs, 8 pads..lots of customizable banks.

Im not a big fan of one encoder plus a push button for inputing data.

Where should i look for the midi implementation in the code?

Where do you request a feature? Through github i guess?

Many thanks
elodin
d.simon wrote:
Altitude909 wrote:
ok, boards back in stock.


I searched/read this thread but couldn't quite tell...does the Michigan Synth Works board need a different panel? I noticed some panel gerbers in your repository...


It uses the same panel as the original.
autodafe
getting some strange behaviour on my first unit.
If I power it up, it boots incorrectly and sometimes the display jumps in any strange direction.
If I then press the "reset" switch, it always boots "normally",
If i turn it off and on again, ith booots strangely again...
Then pressing the reset butto multiple times, it boots up totally fine...

seems to be related to power up/reset commands

any thoughts?

see it:

elodin
@autodafe are you pushing the limits of your power supply? Some modules draw significantly more power on startup, this can cause abnormal behavior if there is insufficient power to go around.
autodafe
elodin wrote:
@autodafe are you pushing the limits of your power supply? Some modules draw significantly more power on startup, this can cause abnormal behavior if there is insufficient power to go around.



nope. in the video I was using a Doepfer DIY kit 1 PSU with just 2 Per|formers connected...same issue with an uZeus and my bigger Meanwell RT-65B...
elodin
Maybe try reflashing the firmware? seriously, i just don't get it
autodafe
i was thinking about it, I just upgraded to 0.1.19 with its SD card...
toneburst
Staticcharge wrote:
I saw your comment on Github regarding chord mode, what's the word on that, seems logical to me.


I don't know any more info than what's in the Issue on GitHub https://github.com/westlicht/performer/issues/58

It does sound quite cool, now I've got my head around it.
toneburst
I have a bag of jack sockets for the 1.0 PCB if anyone wants them.

Complete set of mono sockets, stereo MIDI I/O sockets Not included.

I have either bent the metal tabs flat at the bottom, or cut them off. In some cases, I've filed down the edges of the metal, where I cut off tabs.

I had thought I'd try and lower the panel on my 2nd Performer build, but changed my mind in favour of attempting to raise the switches (probably a bad decision, in hindsight).

Free to a good home. You pay postage.
EU/UK-only, as probably not worth sending long-distance, given relatively low value.

Drop me a PM if interested.
Staticcharge
autodafe wrote:
i was thinking about it, I just upgraded to 0.1.19 with its SD card...

Did it behave the same before the update? Maybe try running the 1.18 update to revert back to the previous state.
Is it a fresh build, ie first flash? Could be worth erasing the firmware and running a clean flash.
Lemmy
Altitude909 wrote:

10mm for the board spacer, 6mm for the display, there is a hardware tab on the sheet


Hi @Altitude909, I have one of your boards on order - it looks on the hardware tab of the BOM that 11mm and 7mm are specified. Or am I looking at the wrong doc? I have a box full of M3 standoffs, I guess I could use these instead?

I'm also wondering how well a 2mm panel would work with the thonkiconn jacks. Do you think that would allow for button clearance without any hackery? I don't have a local option for a 1.6mm panel. I'm personally not keen on buttons which protrude a lot in any case...
Altitude909
Lemmy wrote:

Hi @Altitude909, I have one of your boards on order - it looks on the hardware tab of the BOM that 11mm and 7mm are specified. Or am I looking at the wrong doc? I have a box full of M3 standoffs, I guess I could use these instead?

I'm also wondering how well a 2mm panel would work with the thonkiconn jacks. Do you think that would allow for button clearance without any hackery? I don't have a local option for a 1.6mm panel. I'm personally not keen on buttons which protrude a lot in any case...


Youre looking at the wrong bom, this is the one you should use:

https://docs.google.com/spreadsheets/d/13_oJvbGlHlTXc8rjvBFxwA8uypjMhi ENA5-VM2RqNHc/edit?usp=sharing

2mm thick panel is fine with the PJ398 based board, that's what I used and the thonkiconn jacks are almost tailor made for 10mm spacing/2mm thick panel. The buttons wont protrude any more than they do on the hundreds of modules that use the same formula/parts.

M3 or M2.5 is fine for the height spacer, it doesnt really matter what you use as long as its 10mm tall
Lemmy
Great, thanks
Staticcharge
Is there anyone in the UK that has any panels spare? Phil at pusherman has no stock.
Payment via PayPal, I'm after 2.
Warpus
Staticcharge wrote:
Is there anyone in the UK that has any panels spare? Phil at pusherman has no stock.
Payment via PayPal, I'm after 2.


I believe Pusherman is getting more in next week.
pixelforest
anyone have a pair of the stereo jacks for the midi i/o they're willing to sell or know where to get them in a hurry? thonk is out of stock and i realized i really want to try and drive my volca via midi for a show in a couple weeks.
sneak-thief
@pixelforest: you already built a whole module from the ground up, so why not take 1x midi cable and 1x stereo cable and chop them in half?

Then connect the TRS tip to "pin 4" and ring to "pin 5". I put the scare-quotes because of the odd numbering assigned to the midi DIN pins, as you can see in the diagram:

pixelforest
sneak-thief wrote:
@pixelforest: you already built a whole module from the ground up, so why not take 1x midi cable and 1x stereo cable and chop them in half?

Then connect the TRS tip to "pin 4" and ring to "pin 5". I put the scare-quotes because of the odd numbering assigned to the midi DIN pins, as you can see in the diagram:



i was looking for the jacks - i hadn't sourced them when i built the module originally. ended up finding them at synthcube so we're good. and amazon sells the midi-trs adapter. sweet diagram though.
sneak-thief
Aha - so jack as in socket.

It's interesting how wikipedia claims that jack predominantly refers to a socket - but then look up TRS jack in google images and you get the opposite:

https://www.google.com/search?q=trs+jack&source=lnms&tbm=isch
pixelforest
I added some feature requests to the github that I think would make performer even more powerful in a live setting if implemented.

interested to see what people think so have a look and chime in if you're involved in that side of things.
AlanP
Can I suggest a fairly prominent note on the build guide that the very subtle line going past one of the solder pads on each IC indicates pin 1?

Because I don't think it's mentioned there, and I wound up aping other people's build photos as far as orientation goes (and installing KiCAD, to check the nets there.) I only noticed that very subtle silkscreen pin 1 indicator after populating them, and I don't think I've ever seen this style of Pin 1 indicator before.
smiler
AlanP wrote:
I don't think I've ever seen this style of Pin 1 indicator before.
It's the default style KiCad IC footprints. I agree that they are a bit too subtle but I've kinda learnt to like them. On my stuff I add an extra dot inside the footprint to make it a bit more obvious.
patrickod
Finished up my build the other week and am having a lot of fun getting to know this sequencer. It's a really fantastic project - major props to Simon.



I followed the advice of others here and trimmed the metal skirt around the bottom of the jacks which allowed them to sit flush with the board. The space saving and the fact that I'm using a 1.6mm panel means that the buttons have plenty of clearance and prevented me from having to raise the encoder. Definitely the tightest fit of any module I've built so far. Make sure to friction fit everything before you start soldering!

pixelforest
just did two shows with the performer, one where it was the only sequencer and one with it as well as my usual er-101/metron/cr/metropolis rig. absolutely phenomenal especially with the launchpad.
LeftyStrat
I've purchased this sequencer on Reverb. I'm a little confused by the various models of Launchpad. Will the Launchpad MKII work as well as an original Launchpad?
pixelforest
LeftyStrat wrote:
I've purchased this sequencer on Reverb. I'm a little confused by the various models of Launchpad. Will the Launchpad MKII work as well as an original Launchpad?


original launchpad will not work and is never intended to be made to work. i have a launchpad mini but i believe there is some discussion about other models earlier in this thread.
forestcaver
LeftyStrat wrote:
I've purchased this sequencer on Reverb


That’s a bit concerning - it’s licenced as cc-by-nc-sa !! You realise you can never sell it now ? And the vendor was violating the licence? (Unless, of course, Simon has started selling them...)
pixelforest
forestcaver wrote:
LeftyStrat wrote:
I've purchased this sequencer on Reverb


That’s a bit concerning - it’s licenced as cc-by-nc-sa !! You realise you can never sell it now ? And the vendor was violating the licence? (Unless, of course, Simon has started selling them...)


that's not really what the NC part means in general. someone can't run a business selling them, but there's nothing commercial about selling your own.


from creativecommons.org
Quote:
CC’s NonCommercial (NC) licenses prohibit uses that are “primarily intended for or directed toward commercial advantage or monetary compensation.”


if you build a NC thing for yourself and then at some point decide you no longer want it, i don't think that selling your thing is outside the bounds of the license intention. the definitions are intentionally vague (see further on the page linked) but the spirit of it is more about not turning someone's open source project into your commercial business.
zarar
Any guides on how to Flash the STM the first time around after build? Do I need any equipment to do this or can it be done via USB/SD-Card? How do i do it?
smiler
zarar wrote:
Any guides on how to Flash the STM the first time around after build? Do I need any equipment to do this or can it be done via USB/SD-Card? How do i do it?

You need a proper STM32 programmer the first time. There are various options out there, I used the ST-LINK/V3 together with the STM32CubeProgrammer software on my build.
Staticcharge
pixelforest wrote:
just did two shows with the performer, one where it was the only sequencer and one with it as well as my usual er-101/metron/cr/metropolis rig. absolutely phenomenal especially with the launchpad.


Did you use the Performer as your master clock or another module, I found the ER-101 is a bit fussy with clock inputs and can sometimes glitch.
Staticcharge
zarar wrote:
Any guides on how to Flash the STM the first time around after build? Do I need any equipment to do this or can it be done via USB/SD-Card? How do i do it?


I use this:
https://www.mouser.co.uk/ProductDetail/STMicroelectronics/ST-LINK-V2?q s=H4BOwPtf9MC1sDQ8j3cy4w==&vip=1&gclid=Cj0KCQjwjYHpBRC4ARIsAI-3GkGImcq BAIdTvURCENnXpYljMDHSIVC7RT1b19AvwJ3Trzd48ysD46caAmn1EALw_wcB
Software is downloadable from the ST website.
I found this video helpful:
https://www.youtube.com/watch?v=l300o8UsYOE
cnicht
The STLINKV2 needs a 10pin adapter to connect to the Performer - or does it now come with one?
Staticcharge
cnicht wrote:
The STLINKV2 needs a 10pin adapter to connect to the Performer - or does it now come with one?

No your right, forgot to add that. I bought mine from The PI hut on Ebay but I'm sure there are others more local to whoever needs to buy one.
https://www.ebay.co.uk/itm/Adafruit-JTAG-2x10-2-54mm-to-SWD-2x5-1-27mm -Cable-Adapter-Board-ADA2094/263111779590?epid=1543518694&hash=item3d4 2af2506:g:cNMAAOSwgv5ZeyUd

https://www.ebay.co.uk/itm/Adafruit-10-pin-2x5-Socket-Socket-1-27mm-ID C-SWD-Cable-150mm-long-ADA1675/263111588002?hash=item3d42ac38a2:g:wwAA AOSwNC5Zev6c

You'll find the cable will be suitable for the MI clones that use the Jtag connection as well, so it's worth getting them. I've been using the HEX code method on all my builds for a while but once I have the time to sit down I'll have a look at using the Vagrant environment just for curiosities sake.
cnicht
Or this at Mouser:

https://www.mouser.co.uk/ProductDetail/Olimex-Ltd/ARM-JTAG-20-10?qs=sG AEpiMZZMtE4ePzUE8d2AvNxheVcyoX
Staticcharge
cnicht wrote:
Or this at Mouser:

https://www.mouser.co.uk/ProductDetail/Olimex-Ltd/ARM-JTAG-20-10?qs=sG AEpiMZZMtE4ePzUE8d2AvNxheVcyoX


Which is much cheaper... applause
zarar
The 3 left-bottom pins on the JTAG connector wants to stay together. Is that a problem or should i continue trying to work my soldering braid? very frustrating
cnicht
I think you have a solder bridge to the square pad (Pin 1)

Pins 3, 5 and 9 are connected to Ground on the schematic but they are not adjacent on the pin header.
selfdestroyer
Finished one of mine and it booted right up! This is the PCB & Panel from Pusherman's store.

This was the first time I soldered an STM32



Just realized I'm missing my MIDI out nut.. damn.

I am using the launchpad mini with it and for anyone on the fence about grabbing one, it super fast for live stuff and track manipulation on the fly. I highly recommend it. I fell in love with the interface also, reminds me of my Akai MPC sampler days.

Time to start on my second one for another case.
AlanP
I'm guessing it can be programmed with a FTDI Friend, like Mutable STM32 stuff, but I really don't recommend using the serial pins. Get a ST-Link V2 and a Mini-JTAG adaptor cable.
toneburst
Altitude909 wrote:
ok, off to fab. It all works pretty well except for the midi in which will need the ground lead cut off, small price to pay


If (theoretically, obviously, ahem) I were to forget to cut off the ground leg of the MIDI In socket, and solder the ground for both MIDI In and Out sockets to one of your boards, could that do something to permanent to kill both TRS MIDI In and Out, assuming I had powered up the module for a while, before realising my mistake?
Lemmy
Got my Performer up and running yesterday. What an amazingly powerful module! Great job Simon.
toneburst
Can't get the TRS MIDI input on my Per|former to work at all.

This is one of @Altitude909's Thinkiconn boards (rev.1.1).

I've replaced D2, U1, R1 and R2, and reflowed all pins of the MCU.

MIDI monitor page shows nothing when I plug in a DIN MIDI controller via a Make Noise DIN to TRS adapter.

The same controller, cable and MIDI adapter shows what I would expect on the Monitor page on one of my other Per|formers (original board) when I turn a knob or press a key.

There seems to be something else odd about my 3rd build, too. It won't boot to the bootloader, or run the update from the SD card.

It's currently running firmware v.0.1.21.

I was able to format an SD card, so I assume the card slot is working OK.

Could the inability to boot to the bootloader be related at all to the MIDI input error?

MIDI output now seems to work fine, incidentally.

As noted before, I forgot to remove the GND leg from the MIDI in socket. Could this have made anything Bad happen?
toneburst
The boot loaded issue is unrelated, I think.

I realised there was an error in the little script I made to flash the Bootloader and Sequencer firmware.

I corrected the error, and re-flashed, and now I can boot to the Bootloader, and update from a file on SD card.

Still no TRS MIDI input, though.

Is it possible that the MCU is damaged?

I need to sleep now, but tomorrow I will try to test for continuity between R2 and pin 38 of the MCU.

If that comes back positive, I'm stumped...
toneburst
I think I’ve discovered the source of the problem!

I bought the wrong opto-coupler IC!

I have a VO0630T, instead of the required VO0600T.
The two have different pinouts.

I think the supplier brought up the wrong part of the nearest match when I searched for the opto, and I forgot to check before ordering.

Hopefully, I have a spare of the correct part lying around from previous builds.
TAROUG
Hey there,

I just finished building a performer and it seems to work fine, while jamming around.
When I enter the system to calibrate I can just calibrate CV1 and switching around with the T buttons is however not possible.

The other thing is, I have empty 64gb sd card sitting in the slot but it seems the performer can't see it. Saving is not possible and in the system menu I can't get to UTILS or UPDATE.

What's wrong? help
cnicht
Post some hi-res photos of the board and we can have a look to see if there are any constructional faults.

Have you run the test firmware?
TAROUG
I ran the tester.hex and there were one thing I didn't fix yet. CV2 IN is having 1V constantly.
I will give you some pics in a minute
TAROUG
Damn, I opened the panel and directly saw that I forgot to solder on D2 d'oh!

Let me fix this first and try again...
cnicht
Where are you measuring the voltage on CV2 IN?
TAROUG
D2 didn't solve the problem.




TAROUG
cnicht wrote:
Where are you measuring the voltage on CV2 IN?


Didn't measure at all, but it showed up in test mode.
cnicht
It’s difficult to tell from your photos if there’s a constructional problem - can you improve the image resolution?

What’s the voltage on pin 8 of U6 with nothing plugged in?
TAROUG
cnicht wrote:
It’s difficult to tell from your photos if there’s a constructional problem - can you improve the image resolution?

What’s the voltage on pin 8 of U6 with nothing plugged in?


Problem solved. I reflowed all SMD parts and found a short on the MCU. Now everything works fine It's peanut butter jelly time!

Thank you for helping!
cnicht
Great - happy sequencing! Rockin' Banana!
LeftyStrat
Does anyone know where I could purchase an assembled Performer. My eyesight make doing SMT a challenge.
Roni
selfdestroyer wrote:
This is the PCB & Panel from Pusherman's store.


Is that the anodized aluminium panel or just fr4? If aluminium, how did you fare with fitting it?

.
selfdestroyer
Roni wrote:
selfdestroyer wrote:
This is the PCB & Panel from Pusherman's store.


Is that the anodized aluminium panel or just fr4? If aluminium, how did you fare with fitting it?

.


It's the aluminum. It fit without issues. I did the usual, mount all jacks and place the panel then soldier all the jacks routine.
TAROUG
cnicht wrote:
Great - happy sequencing! Rockin' Banana!


Damn, I thought I am lucky but this think still is not working.
When i turn it on after a while it works perfectly, but after around 2-3min the screen is turning black and all LEDs are off.
I still measure 3.3V on the MCU. Nothing gets too hot.

Any ideas what this could be?
cnicht
It seems like something is heating up, expanding and opening a bad solder joint.

Do the buttons still work when this happens?
TAROUG
cnicht wrote:
It seems like something is heating up, expanding and opening a bad solder joint.

Do the buttons still work when this happens?


I think it's something with the MCU. I tried to connect again with the ST Link and now I just get an error message (Device under reset). I reflowed it several times and I can't find any short with my continuity meter.
Maybe I have fried it somehow?
cnicht
More likely a short or bad solder joint.

Inspect with microscope.
Lemmy
I found the interactive BOM really useful for checking the various power nets. That might help you narrow down the problem.
Beware though that every time you choose a different net to check (e.g. 3.3V) do a hard refresh in your web browser. The web app has a cache problem, it will just show you the components to check from the last net unless you do the refresh.
backfromearth
Do you need to assemble and connect the main power area in order to flash the MCU ?

I succeeded in detecting a MCU in a rings I built last week just by connecting the board with the ST link V2 and the 2*5 jtag connectors, so I know the problem is not the driver and the cable. Following the interactive guide, I just soldered the components around the MCU ( capacitors, and quartz) but do not achieve to detect it, even after a lot of reflows.
sad banana
Thanks!
toneburst
backfromearth wrote:
Do you need to assemble and connect the main power area in order to flash the MCU ?


I think you need to complete the power section, as well as the the MCU and associated components.

You can see from the photo in the Build Guide which components are required to flash the MCU.



I've not tried flashing without the module being powered. I assume that won't work, but I may be wrong.

I've also found it useful to fit the display header and display at this stage, as it's good to see if the firmware is properly installed.

If you do this, you might want to consider also soldering the components near the header, before fitting that, as it's fiddly soldering these after the header.
cnicht
The microcontroller needs the +3.3V and +3.3VA power rails and the associated components as outlined in the build guide (MCU Section)

It’s a good idea to flash the micro before the rest of the build and you should be able to get information on the display.

Use the tester.hex file for system testing.
Lemmy
This is annoying - just testing my midi in/out for the first time and I don't seem to be getting anything out of the midi OUT jack. USB midi out is working, and midi IN is working. I used the same TRS adapter which works with midi IN for midi OUT testing.

I have reflowed U2, R3, R4 and the jack socket.

If I put an oscilloscope on the midi OUT jack socket I get a constant 5V with periodic pulse drops to 0 when the sequencer is running. Is that how it should look? However it doesn't trigger any synth sounds when connected or flash the LED in my midi junction box if I connect to that.

N.B. this is Altitude's board for the Thonk jacks.
toneburst
There's not many components between the MCU and MIDI output.



So, check all those components.

MIDI TX is on pin 37 of the MCU.

Have you made sure to set the Vel. Source of the output notes to a constant value?

If you don't set that, it just puts out a load of 0 velocity notes, so most MIDI devices don't produce any notes.

You probably know that already, though.
Lemmy
toneburst wrote:


Have you made sure to set the Vel. Source of the output notes to a constant value?

If you don't set that, it just puts out a load of 0 velocity notes, so most MIDI devices don't produce any notes.

You probably know that already, though.


No, I didn't but I just tried it and it didn't change anything.

My theory is that my TRS adapter is the wrong type. It looks to me like Altitude added the mod for midi IN to accept either type of TRS adapter, but didn't do the same for midi OUT.

I could be wrong - I'll try putting together a cable with the connections swapped to see if it works.
toneburst
Can't check your theory, sadly, as my latest Per|former build (one of @Altitude909's v1.1 mod) has a broken MCU, and I'm waiting for a replacement.

I did have MIDI In and Out working, using a Make Noise TRS-DIN Midi adapter. That's the kind that works with the 1.0 board, too.
Altitude909
Lemmy wrote:
toneburst wrote:


Have you made sure to set the Vel. Source of the output notes to a constant value?

If you don't set that, it just puts out a load of 0 velocity notes, so most MIDI devices don't produce any notes.

You probably know that already, though.


No, I didn't but I just tried it and it didn't change anything.

My theory is that my TRS adapter is the wrong type. It looks to me like Altitude added the mod for midi IN to accept either type of TRS adapter, but didn't do the same for midi OUT.

I could be wrong - I'll try putting together a cable with the connections swapped to see if it works.


It only works for midi in, midi out is a complete different thing.

>If I put an oscilloscope on the midi OUT jack socket I get a constant 5V with periodic pulse drops to 0 when the sequencer is running. Is that how it should look? However it doesn't trigger any synth sounds when connected or flash the LED in my midi junction box if I connect to that.

Yes, that's exactly what it should look like. Sounds like a cable issue
Lemmy
Yep, it was the TRS adapter.

I can finally put the panel on and rack it now.
toneburst
Altitude909 wrote:
It only works for midi in, midi out is a complete different thing.


Will you implement a similar mechanism for MIDI Out in the future?
toneburst
toneburst wrote:
Can't check your theory, sadly, as my latest Per|former build (one of @Altitude909's v1.1 mod) has a broken MCU, and I'm waiting for a replacement.


Replaced the MCU, all now working again (including TRS MIDI In and Out)!
Bartelby
The final bits I need should arrive today or tomorrow. Hopefully on the weekend I can start to build this...
toneburst
Bartelby wrote:
The final bits I need should arrive today or tomorrow. Hopefully on the weekend I can start to build this...


It's peanut butter jelly time!
toneburst
Sorry, that should have been..

Rockin' Banana!
toneburst
Experimenting with different button colour options.









Yes, those are a couple of knives from the kitchen drawer, propping the panel up.

Ingenious, no..?
sduck
# 3
(having spent some quality time programming the thing in the last week, having a visual guide to find beats 1 5 9 13 would be nice - the other ones are too subtle)
Bartelby
toneburst wrote:
Sorry, that should have been..

Rockin' Banana!


You won't be saying that when I'm hassling you for build advice and then the board pops through your door for flashing!
toneburst
Bartelby wrote:
You won't be saying that when I'm hassling you for build advice and then the board pops through your door for flashing!


No, that’s all fine smile
zarar
Staticcharge wrote:
zarar wrote:
Any guides on how to Flash the STM the first time around after build? Do I need any equipment to do this or can it be done via USB/SD-Card? How do i do it?


I use this:
https://www.mouser.co.uk/ProductDetail/STMicroelectronics/ST-LINK-V2?q s=H4BOwPtf9MC1sDQ8j3cy4w==&vip=1&gclid=Cj0KCQjwjYHpBRC4ARIsAI-3GkGImcq BAIdTvURCENnXpYljMDHSIVC7RT1b19AvwJ3Trzd48ysD46caAmn1EALw_wcB
Software is downloadable from the ST website.
I found this video helpful:
https://www.youtube.com/watch?v=l300o8UsYOE


Thanks for the video, that was helpful. I've got one of these:



I includes 4 of those single-cables. Do i need to use a cable with all 10 cables for the header or are these 4 enough?
cnicht
You’ll find it difficult connecting to the header on the PCB with those cables because of the pin size and spacing.

Does anyone know of an adapter suitable?
TAROUG
I had some issues with my performer before so I replaced the STM. I connected via JTAG succesfully and uploaded the bootloader. After that I tried to upload the sequencer.hex and received this.

Now I can't connect anymore. What's going on???
Help is much appreciated smile

[/img]
Karma
wrong place sorry
gotama
zarar wrote:
Staticcharge wrote:
zarar wrote:
Any guides on how to Flash the STM the first time around after build? Do I need any equipment to do this or can it be done via USB/SD-Card? How do i do it?


I use this:
https://www.mouser.co.uk/ProductDetail/STMicroelectronics/ST-LINK-V2?q s=H4BOwPtf9MC1sDQ8j3cy4w==&vip=1&gclid=Cj0KCQjwjYHpBRC4ARIsAI-3GkGImcq BAIdTvURCENnXpYljMDHSIVC7RT1b19AvwJ3Trzd48ysD46caAmn1EALw_wcB
Software is downloadable from the ST website.
I found this video helpful:
https://www.youtube.com/watch?v=l300o8UsYOE


Thanks for the video, that was helpful. I've got one of these:



I includes 4 of those single-cables. Do i need to use a cable with all 10 cables for the header or are these 4 enough?


I toasted the MCU trying to use one of these mini STLink V2 clones. You need the real STLink V2 and the Olimex adapter.
toneburst
TAROUG wrote:
I had some issues with my performer before so I replaced the STM. I connected via JTAG succesfully and uploaded the bootloader. After that I tried to upload the sequencer.hex and received this.

Now I can't connect anymore. What's going on???
Help is much appreciated smile

[/img]


I'd check the MCU and JTAG header very carefully for shorts across the pins using a loupe or other illuminated magnifier.

If you use the drag-solder method (as I did on 3x Per|former builds), check very carefully for broken strands of desolder braid shorting across MCU pins.

One of those caused a short on the MCU of my last build that made the IC overheat and eventually fail.
Handmedown
I have been using Performer for many months and then when I went to do and a firmware update. It was not powered before I tried to do update. I help down encoder and powered and it would not go into update mode and nothing worked no lights or display. Tried to start with out holding down encoder and seemed dead. Did a reflash of bootloader and sequencer.hex and then it went into update mode. Installed update and then seemed ok but shortly after display shut off and then random light were on. Reflashed again and now seems better. Anybody else having issues like this? Is the precessor failing?

BTW all voltages were checked and ok as well as connects.

Thanks
TAROUG
toneburst wrote:
TAROUG wrote:
I had some issues with my performer before so I replaced the STM. I connected via JTAG succesfully and uploaded the bootloader. After that I tried to upload the sequencer.hex and received this.

Now I can't connect anymore. What's going on???
Help is much appreciated smile

[/img]


I'd check the MCU and JTAG header very carefully for shorts across the pins using a loupe or other illuminated magnifier.

If you use the drag-solder method (as I did on 3x Per|former builds), check very carefully for broken strands of desolder braid shorting across MCU pins.

One of those caused a short on the MCU of my last build that made the IC overheat and eventually fail.


I found another short on the MCU and solved this. Flashed bootloader and firmware. Next problem is, I start it up and I see start up on the screen, but after 3-4 seconds it completly shuts down. No screen, no LEDS, but still voltage reaching the MCU. I reflowed everything so damn often now, don't know what to do. Any ideas?
mongrol
toneburst wrote:


I've also found it useful to fit the display header and display at this stage, as it's good to see if the firmware is properly installed.

If you do this, you might want to consider also soldering the components near the header, before fitting that, as it's fiddly soldering these after the header.


Are any components required to actually make the display work other than the header? I'm at the MCU stage and would like to plug in the display as you said.
Warpus
Hey all, I am having a calibration issue. All 8 CV channels are offset by -5V so the calibration range is approximately -10V to 0V instead of -5V to +5V. I'm on the latest firmware. Anyone have an idea what I need to correct? Thanks!
cnicht
Are the output op amps getting VBIAS?
av500
gotama wrote:

I toasted the MCU trying to use one of these mini STLink V2 clones. You need the real STLink V2 and the Olimex adapter.


I programmed hundreds of STM32 board with these adaptors, I see no reason why they would not work for this one.
Warpus
cnicht wrote:
Are the output op amps getting VBIAS?


Yep. That was it. The LM4040 needed a little reflow. Thanks!
Lemmy
av500 wrote:
gotama wrote:

I toasted the MCU trying to use one of these mini STLink V2 clones. You need the real STLink V2 and the Olimex adapter.


I programmed hundreds of STM32 board with these adaptors, I see no reason why they would not work for this one.


I found it difficult to find an appropriate mini JTAG adaptor for mine. So I went with the 'official' STLink v2 + Olimex adaptor in the end. Did you wire your own adaptor, or were you connecting to boards with bigger JTAG adaptors?
av500
Lemmy wrote:
av500 wrote:
gotama wrote:

I toasted the MCU trying to use one of these mini STLink V2 clones. You need the real STLink V2 and the Olimex adapter.


I programmed hundreds of STM32 board with these adaptors, I see no reason why they would not work for this one.


I found it difficult to find an appropriate mini JTAG adaptor for mine. So I went with the 'official' STLink v2 + Olimex adaptor in the end. Did you wire your own adaptor, or were you connecting to boards with bigger JTAG adaptors?


both, though in case of these darn mini jtags I usually just solder 4 wires to them...

hopefully newer designs will go with a simple 4pin 0.1" header like the new mutable stuff
Altitude909
av500 wrote:


both, though in case of these darn mini jtags I usually just solder 4 wires to them...

hopefully newer designs will go with a simple 4pin 0.1" header like the new mutable stuff


What also works is just to use one header IN the STlink, not soldered into the board and just use friction. I've been doing this for a couple months and it's worked 100% of the time. I hate those things, they bend easily, are expensive for what they are, and snag on everything. Whole lot of PITA for something this is usually used once. I can see why she went to pogo pins/SWD header
TAROUG
I found another short on the MCU and solved this. Flashed bootloader and firmware. Next problem is, I start it up and I see start up on the screen, but after 3-4 seconds it completly shuts down. No screen, no LEDS, but still voltage reaching the MCU. Nothing gets hot. I reflowed everything so damn often now, don't know what to do. Any ideas what components to look at?
cnicht
Are all the power rails correct when this happens?
TAROUG
cnicht wrote:
Are all the power rails correct when this happens?


I just noticed I have around 1.7V on AREF_-10V. When I turn it on it's shaking
around 2.3V and is slowly loosing voltage. That's not correct I guess?
cnicht
Nope - although I doubt whether that is causing the problem.

The microcontroller needs both +3.3V and +3.3VA to work properly.
TAROUG
cnicht wrote:
Nope - although I doubt whether that is causing the problem.

The microcontroller needs both +3.3V and +3.3VA to work properly.


3.3V and 3.3VA is coming in on the appropriate pads of the microcontroller. I checked for shorts between all pads with continuity meter and I reflowed several times. I replaced the chip before and lost 2 pads during that, but these weren't connect to anything so that shouldn't be the problem.

Sometimes a few LED's blinking up randomly. What would you suggest? Reflow again and again and again? grin
cnicht
Reflowing may reconnect unsoldered pads but inspection with a microscope might give you a less all or nothing approach.

A technique I use, to see if pins of a microcontroller are soldered, is very carefully with the tip of some tweezers see if the pin will move to the side if you apply pressure. Disclaimer!!!!! You must do this VERY VERY carefully and at your own risk!!!! If the pin moves then it’s not soldered. It can then be pushed back to be soldered.

I suggest practice on some scrap boards/chips first.

Good luck!
TAROUG
cnicht wrote:

A technique I use, to see if pins of a microcontroller are soldered, is very carefully with the tip of some tweezers see if the pin will move to the side if you apply pressure. Disclaimer!!!!! You must do this VERY VERY carefully and at your own risk!!!! If the pin moves then it’s not soldered. It can then be pushed back to be soldered.

I suggest practice on some scrap boards/chips first.

Good luck!


Did exactly that and all pins are soldered.
cnicht
Post some high resolution zoomed-in photos of your board
rithma
Hi Everybody! So I built 3 performers, two working great, and one boots up, and rotating the encoder works, but none of the buttons (encoder or matrix) are working. Is this a "reflow the STM" situation? Anybody else experience this?

thanks!
/Rithers
TAROUG
cnicht wrote:
Post some high resolution zoomed-in photos of your board


My phone can't get better than this :(
And yeah, this MCU looks ugly but I checked connectivity between all pins and appropriate destination.



cnicht
I can’t see anything obvious but it’s difficult to tell from the photographs.

I’ll do some more staring!

Edit: Is there a short between pins 34 and 35?
TAROUG
cnicht wrote:
I can’t see anything obvious but it’s difficult to tell from the photographs.

I’ll do some more staring!

Edit: Is there a short between pins 34 and 35?


No continuity between 34 and 35...
cnicht
The only thing I can now suggest is that something is getting just warm enough to expand a break a connection.

Sometimes a freezer spray is useful to diagnose these problems.
windspirit
Hey so which firmware do I need to upload to get the oled to light up? I tried "
tester.hex" and "sequencer.hex" and while they both uploaded successfully I see no activity on the oled. I tested the voltage regulators and voltage input on the oled and everything looks correct.

Also did anyone else think that the orientation markers of the ICs wasn't super clear? I've never built another project where pin one was marked with just a white line.

EDIT: I got it, did not realize I had to upload "bootloader" before "sequencer". New to STM programming
iheartmodular
has anyone made a nice acrylic screen insert protector for these yet?
cnicht
Ask Gbiz
gbiz
My batch sold out a long while ago.

If you can't find anyone selling them (i thought Pusherman was ?), if you don't mind getting your own made up, you can use my dxf at Ponoko/Razorlab.

Details in this post from earlier in the thread https://www.muffwiggler.com/forum/viewtopic.php?p=3033244#3033244
toneburst
iheartmodular wrote:
has anyone made a nice acrylic screen insert protector for these yet?


I have an order in for 14 units at the moment.

I don't plan to make 12 more Per|formers, so I will have some spare.

I can sent you a couple, for a nominal fee + postage, when they arrive.

It might be just as easy (and quicker) for you to have some of your own made, though.

There are instructions in the .zip file gBiz linked to above.
toneburst
gbiz wrote:
If you can't find anyone selling them (i thought Pusherman was ?)


He said he was going to, but they haven't appeared on his site, yet.
steviet
Looks to me like there are some at Pusherman right now. I just put in an order for two PCB + Panel combos.
toneburst
steviet wrote:
Looks to me like there are some at Pusherman right now. I just put in an order for two PCB + Panel combos.


As far as I know, his Per|former panels (which are lovely, it has to be said) don't come with display windows.

Phil, Mr Pusherman, said he would sell them separately, but they're not listed on his site, yet.

Happy to be proved wrong.
steviet
toneburst wrote:


As far as I know, his Per|former panels (which are lovely, it has to be said) don't come with display windows.

Phil, Mr Pusherman, said he would sell them separately, but they're not listed on his site, yet.

Happy to be proved wrong.


Ahh my apologies, I read up to Gbiz's post and assumed everyone was talking about PCBs, they had been out of stock at Pusherman for a while as well.

Not sure on the acrylic panel! I'll post back here when I receive the order, but I'd assume you're correct in that it doesn't come with a protector. Ill throw some saran wrap on it for the time being if that's the case. thumbs up
iheartmodular
nice thank you all very much thumbs up

I will pm!!
Warpus
On my build all 4 CV inputs are measuring a fluctuating -0.05V to -0.04V without any external voltage source connected. Can this be calibrated? Is it normal behavior?

Thanks!
rithma
hi!!! I've attached a clip of the noise floor of my system when the performer is plugged in, (1st half of clip) and whenm its not plugged in (2nd half of clip). Can anybody tell me what the hell I am doing wrong here? It is pretty much unbearable.

[s]https://soundcloud.com/rithma/7-audio-0002-2019-08-17-223401[/s]

https://soundcloud.com/rithma/7-audio-0002-2019-08-17-223401
forestcaver
Anyone added a mic803 supervisor to their performer? Just thinking about adding one as I always have power-on issues (needs turning on-off-on again in order to boot most of the time on most of my power supplies). I am pretty confident the build is fine.
It doesnt look too hard to add it - just wondering about physically placing it (maybe hot glue :-) )
I was just curious if anyone else had done it.... (I’ve got one spare so will try it soon)
cnicht
Is the issue that the microcontroller is not being held in reset state long enough for the power rails to reach operating voltage?

I wonder if a capacitor across SW8 might help? This would extend the reset period.
forestcaver
cnicht wrote:
Is the issue that the microcontroller is not being held in reset state long enough for the power rails to reach operating voltage?

I wonder if a capacitor across SW8 might help? This would extend the reset period.


I think that’s the issue. I’ve tried playing with the brown out reset and other power management in the fw with no luck. Simon also added some code to try to improve this.
If it is the problem you descrie then a mic803 is probably the right solution and is not much bigger than a cap anyway - I didnt solder the soft touch buttons so I will probably use those pads.... but I’ll need to triple check the schematic and board to make sure it is simple to do - not keen to start cutting traces on an otherwise working build ! (But dont mind jumpering with thin wire)
cnicht
Or perhaps a long delay in the firmware on power-on?

An experiment might be to hold down SW8 on power-on then release it to see if the microcontroller boots correctly. Not practical to do this every time but might point to it being a power-on reset problem?
forestcaver
cnicht wrote:
Or perhaps a long delay in the firmware on power-on?

An experiment might be to hold down SW8 on power-on then release it to see if the microcontroller boots correctly. Not practical to do this every time but might point to it being a power-on reset problem?


firmware delay wont help. I’ve got a spare supervisor anyway. I’ll need to investigate and experiment properly. I just lose motivation when I think of the number of nuts I’ll need to remove :-)

The mic803 should be able to be soldered to one of the pads of sw8 (which exposes the lines I need) and then a couple of jumper wires and a 100n cap should do it fine. Should be straight-forward. I’ll try it when I have some motivation - hopefully his week.
sduck
I have a similar problem in my large euro cabinet. I need to power on twice fairly quickly to get performer to boot. Oddly enough this isn't an issue in the make Noise Shared System case I sometimes use. What I got to fix it is a WMD Soft Start module - https://wmdevices.com/collections/eurorack-modules/products/soft-start -module
but that cap over SW8 sounds like a better idea.
forestcaver
sduck wrote:
I have a similar problem in my large euro cabinet. I need to power on twice fairly quickly to get performer to boot. Oddly enough this isn't an issue in the make Noise Shared System case I sometimes use. What I got to fix it is a WMD Soft Start module - https://wmdevices.com/collections/eurorack-modules/products/soft-start -module
but that cap over SW8 sounds like a better idea.


I think it’s not an uncommon issue - I think a number of people have had the same problem (I raised an issue a while ago).

The trouble with using an RC filter on the reset is in matching the power on slew rate for each psu... it would be a bit hit and miss but very cheap and effective when it works. But it’s not a universal solution (I dont think). I think the right answer is a reset supervisor and soldering one on sw8 is no more hassle than a cap. I’ll report back when I get my arse in gear!
forestcaver
Ok - tried the mic803 and it works really well on several different psus - there’s a noticeable delay in startup (as expected) and then it boots normally. Not had a single failed boot since fitting it. Tried it about 10 times on three different (previously problematic) PSUs. Previously it failed to boot immediately every time.
It’s trivial to add on as well.
I soldered an 0805 100nF cap (as it fit beautifully) across the supervisor gnd and nrst pins. I soldered the supervisor across the GND and RESET pads of SW8. I soldered vcc to the +3v3 terminal of R82. I hope the pics help as well as some scribbled notes that may or may not help!


shiny saw
Hi you all,

I've finished my Performer, and I have very woggling CV out. Even in Calibration, when the voltage should be steady, it wobbles a lot. Any idea about the possible cause of this issue?
cnicht
Post some oscilloscope photos of the voltages
shiny saw
Wow.. I've just switched on the system to provide those oscilloscope captures, and now it's just not happening! Same setup, right the same one I left yesterday when I got this very different behaviour:
cnicht
You may have a bad solder joint making intermittent contact or a component breaking down.

Are all the CV outputs exhibiting this behaviour? If so is VBIAS present and stable on all the CV output op amps?

Have you run tester.hex?
shiny saw
cnicht wrote:
You may have a bad solder joint making intermittent contact or a component breaking down.

Are all the CV outputs exhibiting this behaviour? If so is VBIAS present and stable on all the CV output op amps?

Have you run tester.hex?


All CV outputs were, yes. But not now, so I can't do any significant test I guess. I didn't run tester.hex. I flashed with a friend's programmer I don't have right now with me.

It's now behaving pretty well about CV so far. But I'm not able to get any MIDI IN in MIDI/CV track nor record it. Monitor doesn't show any MIDI activity. Tried MIDI and USB.
cnicht
Try tracing the MIDI signal through the optoisolator and then to the microcontroller and signal trace the USB data
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 22
Powered by phpBB © phpBB Group