Miasma: Function Generator (DIY PCB based on Befaco Rampage)

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, Joe., analogdigital, infradead, lisa, parasitk, plord, sduck

Post Reply
User avatar
Sin_Phi
Common Wiggler
Posts: 115
Joined: Fri Feb 06, 2015 10:10 am
Location: L.A. (Palos Verdes)
Contact:

Post by Sin_Phi » Tue Jan 15, 2019 12:53 pm

@aragorn23, the first thing to check are things unique to the long range circuit, the timing capacitor and the rate switch. I would just swap the cap out and check continuity with the switch. After that I would do some basic things like seeing if there are knob settings that the long range works with or offsets on the attenuverters. In the original prototypes the cycling would stop when the attenuverter was full positive, but I haven't had that behavior with the latest iteration. Did you build yours with an SMD matched pair?

@aragorn23 & @MatrixModulator I don't see anything obvious that would indicate this behavior, but it isn't really limiting the function? I will look into it on my next build.

aragorn23
Common Wiggler
Posts: 174
Joined: Wed Jul 26, 2017 6:02 am
Location: Johannesburg, South Africa
Contact:

Post by aragorn23 » Wed Jan 16, 2019 1:41 am

Sin_Phi wrote:@aragorn23, the first thing to check are things unique to the long range circuit, the timing capacitor and the rate switch. I would just swap the cap out and check continuity with the switch. After that I would do some basic things like seeing if there are knob settings that the long range works with or offsets on the attenuverters. In the original prototypes the cycling would stop when the attenuverter was full positive, but I haven't had that behavior with the latest iteration. Did you build yours with an SMD matched pair?
Thanks Sin_Phi, I'll try those tips. Where should I have used a matched SMD pair?

aragorn23
Common Wiggler
Posts: 174
Joined: Wed Jul 26, 2017 6:02 am
Location: Johannesburg, South Africa
Contact:

Post by aragorn23 » Sat Jan 19, 2019 5:58 am

I switched my Miasma on today and it looks like the left channel is stuck on Fall, regardless of how many settings I tweak (range, attenuverters, etc.)

Anyone know where to start troubleshooting?

User avatar
pandereteiro
Learning to Wiggle
Posts: 7
Joined: Thu May 05, 2016 7:23 am
Location: Spain

Post by pandereteiro » Sun Jan 20, 2019 10:08 am

Built a a couple of units here and I noticed when comparing the behavior of Miasma and Rampage that the IN and TRIG inputs seem to have been swapped.

Lemmy
Common Wiggler
Posts: 146
Joined: Mon Oct 24, 2016 1:06 pm

Post by Lemmy » Sun Jan 20, 2019 1:12 pm

pandereteiro wrote:Built a a couple of units here and I noticed when comparing the behavior of Miasma and Rampage that the IN and TRIG inputs seem to have been swapped.
I think you're right... I always meant to work out why the ring mod wasn't working - using TRIG for IN seems to fix it. (I only get a ring mod effect if I feed both channels a pulse wave, nothing for saw or triangle - is that to be expected?)
Also triggering is working using the IN socket.

Good spot! I've been using mine mostly for free cycle-based CV output, this explains why my other tinkerings were frustrated...

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Mon Jan 21, 2019 9:13 am

Lemmy wrote:
pandereteiro wrote:Built a a couple of units here and I noticed when comparing the behavior of Miasma and Rampage that the IN and TRIG inputs seem to have been swapped.
I think you're right... I always meant to work out why the ring mod wasn't working - using TRIG for IN seems to fix it. (I only get a ring mod effect if I feed both channels a pulse wave, nothing for saw or triangle - is that to be expected?)
Also triggering is working using the IN socket.

Good spot! I've been using mine mostly for free cycle-based CV output, this explains why my other tinkerings were frustrated...
I looked at the I/O circuit and the trig and in are labelled the same as the panel so its not a panel labelling issue. I don't know enough about circuits to see that the trig and in labelling are swapped on the circuit. Can anyone else determine this from the circuit?


If it is a simple mislabelling, and I didn't want to buy a new panel, could I simply cut the traces and hand wire the signals? For example, cut the trace between in1 and r11, wire in1 to D2, cut the trace between trig 1 and d2, and wire trig1 to r11? Im not sure what that would mean for the normalization when no plug is present in the jack. Perhaps the normalled connections would have to be swapped as well. I haven't started my build yet, just trying to plan ahead.

pppier
Learning to Wiggle
Posts: 20
Joined: Sun Feb 26, 2017 3:06 pm
Location: Italy

Post by pppier » Mon Jan 21, 2019 10:09 am

I built two, had no issues on trigger / gate / signal input
no need to retrace, I mean

User avatar
Sin_Phi
Common Wiggler
Posts: 115
Joined: Fri Feb 06, 2015 10:10 am
Location: L.A. (Palos Verdes)
Contact:

Post by Sin_Phi » Mon Jan 21, 2019 1:14 pm

The trigger / input switching makes no sense. There is really no way for those signals to get confused, checked all my schematic and board file versions and there is nothing switched around. Check that they work as LPF when using the in input, leave the fall fast and control the cutoff with the rise and monitor out and the ring mod.

@aragorn23 Unfortunatly I had this happen on the right channel of one I built, never got it debugged, I will pull it out next week and see if I can revive it and give you pointers. I thought it was because I used some offbrand op amps

OIP
Common Wiggler
Posts: 151
Joined: Wed Oct 07, 2015 9:43 pm

Post by OIP » Mon Jan 21, 2019 5:37 pm

after reading the last batch of posts i checked mine yesterday and as far as i can tell the trigger and input work as labelled - have never had an issue with those. a trigger to the 'trig' input starts a cycle, and any signal to the 'input' can be filtered. also regularly cross patch from one of the pulse outputs of the left channel to the trig of the right channel and it works great. never had a rampage to see if there's any difference but i don't see how it could be mixed up.

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Mon Jan 21, 2019 6:18 pm

OIP wrote:after reading the last batch of posts i checked mine yesterday and as far as i can tell the trigger and input work as labelled - have never had an issue with those. a trigger to the 'trig' input starts a cycle, and any signal to the 'input' can be filtered. also regularly cross patch from one of the pulse outputs of the left channel to the trig of the right channel and it works great. never had a rampage to see if there's any difference but i don't see how it could be mixed up.
Okay good to know. Hopefully my build goes well.

Sin_Phi, is there somewhere I can read up on the charges between pcb revisions? My pcb is version 1.03. I ordered parts as per the current BOM on your website.
Last edited by squarewavesurfer on Wed Jan 23, 2019 8:58 pm, edited 1 time in total.

Lemmy
Common Wiggler
Posts: 146
Joined: Mon Oct 24, 2016 1:06 pm

Post by Lemmy » Tue Jan 22, 2019 12:47 pm

For reference, mine was made with IImyment's PCBs. See earlier in the thread. Not sure what version that was... IN is definitely functioning as TRIG though. No big deal here, I may redo the panel at some point, as I can do that easily enough.

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Thu Jan 24, 2019 8:44 am

squarewavesurfer wrote:
OIP wrote:after reading the last batch of posts i checked mine yesterday and as far as i can tell the trigger and input work as labelled - have never had an issue with those. a trigger to the 'trig' input starts a cycle, and any signal to the 'input' can be filtered. also regularly cross patch from one of the pulse outputs of the left channel to the trig of the right channel and it works great. never had a rampage to see if there's any difference but i don't see how it could be mixed up.
Okay good to know. Hopefully my build goes well.

Sin_Phi, is there somewhere I can read up on the charges between pcb revisions? My pcb is version 1.03. I ordered parts as per the current BOM on your website.
Can anyone confirm the current 2.0 BOM works for 1.03 pcbs? anyone know the changes from v1.03 to v2? I have all the parts ready to go but I don't want to start soldering until I know for sure.

Thanks

User avatar
Sin_Phi
Common Wiggler
Posts: 115
Joined: Fri Feb 06, 2015 10:10 am
Location: L.A. (Palos Verdes)
Contact:

Post by Sin_Phi » Thu Jan 24, 2019 1:05 pm

There is no v2? There is a v1.04 which just lets you use a through hole hand matched pair of transistors. The BOM is good for v1.02-4. Building with the included KiCAD files actually makes things easier if you are having trouble finding where things are on the board. I think ctrl+f in the board view is the way to do it, if you need to find "R76" just type that in and it will zoom to the component.

The difference from 1.02 to 1.03 was just the timing cap footprint change and some routing optimization iirc. If you somehow got a prototype, v1.01, board just throw that away and get the new one, not worth the hassle.

*edit, remembered that I changed the switch footprint to work with lugs or pins. If you have one of the older boards you can still use lugs, you just have to crush or trim one side of the lug.

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Thu Jan 24, 2019 2:35 pm

Sin_Phi wrote:There is no v2? There is a v1.04 which just lets you use a through hole hand matched pair of transistors. The BOM is good for v1.02-4. Building with the included KiCAD files actually makes things easier if you are having trouble finding where things are on the board. I think ctrl+f in the board view is the way to do it, if you need to find "R76" just type that in and it will zoom to the component.

The difference from 1.02 to 1.03 was just the timing cap footprint change and some routing optimization iirc. If you somehow got a prototype, v1.01, board just throw that away and get the new one, not worth the hassle.

*edit, remembered that I changed the switch footprint to work with lugs or pins. If you have one of the older boards you can still use lugs, you just have to crush or trim one side of the lug.
Okay great. Thanks for the quick reply. I definitely have board v1.03. The confusion of the v2.0 pcb came from a post in this thread from a couple pages back:
Sin_Phi wrote:Updated my project page with the version 2.0 boards. http://www.sinphi.com/synths/miasma/miasma.html.

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Fri Jan 25, 2019 8:13 am

How can I tell which way the ICs go? The PCs doesnt have markings. Neither does the placement guide (submitted by a user in this thread), or Kincaid files as far as I can tell.

User avatar
cnicht
Veteran Wiggler
Posts: 548
Joined: Mon Sep 23, 2013 9:23 am

Post by cnicht » Fri Jan 25, 2019 9:25 am

My version of the board had a white line which extended over Pin 1 of the IC package (likewise for the dual transistors)

Hope this helps.

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Fri Jan 25, 2019 1:45 pm

cnicht wrote:My version of the board had a white line which extended over Pin 1 of the IC package (likewise for the dual transistors)

Hope this helps.
Awesome! That helps tons. I was using the schematics to determine where pin 1 should go and that was a slow process.

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Fri Jan 25, 2019 11:29 pm

I calibrated the miasma's shape as per the instructions; however, with equal rise/fall time at full clockwise (logarithmic) on the shape pot, the rise/fall time at linear shape position is nowhere close to being equal. Is this to be expected?

The befaco rampage build manual says to do the above calibration but with the shape pot centred (linear). why the difference in calibration for these two modules?

Also, I didn't see any instructions on calibrating the rise trim or fall trim. Are these trimmers for the rise and fall attenuverters?

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Mon Jan 28, 2019 9:31 pm

Still unsure of the shape calibration as in my above post but I'll wait for sin phi to reply on the reasoning for different method as compared to befaco rampage.

On another topic, does anyone else have a noisy output on EOC2? EOC1 shows a constant voltage on the oscilloscope with spikes at EOC. EOC2 shows a kind of arched noisy waveform with the voltage spiking at EOC. It still outputs a trigger just fine but it just doesnt look right on the scope.

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Tue Feb 05, 2019 10:32 am

Trying to solve a couple minor issues of unequal rise/fall time throughout the travel of the shape potentiometer as well as the EOC2 output being noisy compared to EOC1. I followed the BOM during my build and I was able to catch the discrepancy of r56 and r57 being 110k in the BOM when they should be 100k as already stated in this thread.

One discrepancy between the BOM and the PDF part placement guide llmyment attached a few pages back is relating to Q1-4 on the IO board. The BOM lists all of the as being MMBT3904, wheras the placement guide says Q1,2 are BC847 and Q3,4 are BC547. Can anyone confirm which is correct?

Image

ImageImage[/img]

User avatar
Sin_Phi
Common Wiggler
Posts: 115
Joined: Fri Feb 06, 2015 10:10 am
Location: L.A. (Palos Verdes)
Contact:

Post by Sin_Phi » Tue Feb 05, 2019 3:14 pm

I will look at fixing R56 and R57 in the BOM. Q1-4 on the IO board absolutely are 3904 types. Probably just picked up the BC547 in the placement guide because that is the footprint I used because it had the right pin out and circuit symbol for the transistors (the prototype had to have rotated ones). On EOC 2 you could try without the 1M resistor, just leave that pad open.

User avatar
squarewavesurfer
Common Wiggler
Posts: 218
Joined: Tue Jan 27, 2015 2:43 pm

Post by squarewavesurfer » Tue Feb 05, 2019 3:26 pm

Sin_Phi wrote:I will look at fixing R56 and R57 in the BOM. Q1-4 on the IO board absolutely are 3904 types. Probably just picked up the BC547 in the placement guide because that is the footprint I used because it had the right pin out and circuit symbol for the transistors (the prototype had to have rotated ones). On EOC 2 you could try without the 1M resistor, just leave that pad open.
Excellent, thanks! I did some thorough inspection of my soldering and noticed a few spots i could clean up a bit. After that I tested the EOC2 output and it is perfect. No more noise. Also, coincidently it fixed my rise/fall time being unequal throughout the range of the shape pot after calibrating at full clockwise.

Happy to say it's operating at 100%! Thanks

User avatar
MatrixModulator
Wiggling with Experience
Posts: 492
Joined: Fri Nov 28, 2014 10:35 am
Location: Montreal
Contact:

Post by MatrixModulator » Thu Feb 14, 2019 8:51 pm

Sin_Phi wrote: @aragorn23 & @MatrixModulator I don't see anything obvious that would indicate this behavior, but it isn't really limiting the function? I will look into it on my next build.
Thanks, actualy i can't verify myself it was a build for someone, i will build another and test it myself but yeah it's works, he says it's fine but reversed, not limiting the function, just reversed. I thought you would know about it. I'll update here when i'm done with the new build.
ImageImageImage

KT
Learning to Wiggle
Posts: 10
Joined: Thu May 11, 2017 4:37 am
Location: Brussels

Post by KT » Tue Feb 26, 2019 9:55 am

does someone, preferably in Europe, has a spare pcb/panel? A silver pannel would be super

User avatar
patrickod
Learning to Wiggle
Posts: 20
Joined: Sun Dec 30, 2018 3:18 pm
Location: San Francisco
Contact:

Post by patrickod » Tue Mar 26, 2019 7:39 pm

hey folks. my own PCB and panel order arrived yesterday and as a result I have spare sets (a panel and each of the two PCBs) which I'm looking to sell or trade. 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 (higher res attached). Feel free to DM me if you're interested. Excited to get building on this as part of my first rig :)

Image
You do not have the required permissions to view the files attached to this post.

Post Reply

Return to “Music Tech DIY”