DIY Mutable Unsuccessful Builds

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

Moderators: lisa, Kent, Joe.

Post Reply
User avatar
cnicht
Veteran Wiggler
Posts: 569
Joined: Mon Sep 23, 2013 9:23 am

Re: DIY Mutable Unsuccessful Builds

Post by cnicht » Wed Jan 15, 2020 7:34 am

forestcaver wrote:
Wed Jan 15, 2020 5:59 am
Is that solder paste or flux flooding all the areas between passives eg r39/r40 etc etc (and likely between pins of ic1 - the voltage regulator) ???? Ps looks also like you have dry joints as well as bridges on the stm32 - but always hard to tell from a photo....
What fault finding have you done? Do you get the correct reference voltages as well as the 3.3v rails and +/-12v rails?
(The problem is unlikely to be the firmware....)
Agreed!

oldrobot
Learning to Wiggle
Posts: 8
Joined: Sun Dec 01, 2019 12:46 pm
Location: Poland

Re: DIY Mutable Unsuccessful Builds

Post by oldrobot » Wed Jan 15, 2020 7:57 am

forestcaver wrote:
Wed Jan 15, 2020 5:59 am
Is that solder paste or flux flooding all the areas between passives eg r39/r40 etc etc (and likely between pins of ic1 - the voltage regulator) ???? Ps looks also like you have dry joints as well as bridges on the stm32 - but always hard to tell from a photo....
What fault finding have you done? Do you get the correct reference voltages as well as the 3.3v rails and +/-12v rails?
(The problem is unlikely to be the firmware....)
Yep, it was the solder paste. I've tried to clean it up as much as I can, still not perfect but here is the new pic.
IMG_1669.jpg
Yes, I have +/- 12V and 3.3V everywhere it should be - 19,32,48,64 pins of stm32 have 3.3v; pin 8 of IC8 receives 12v and pin 4 receives -12V; pin 4 of IC 7 receives 3.3v; as well as pin 8 of IC5 and pins 1,8,14,27 of IC6. Pin 16 of IC4 also receives 3.3v.

So I believe everything is ok with power. And I don't know where to dig next :despair: Maybe one of the chips isn't oriented properly?
You do not have the required permissions to view the files attached to this post.

User avatar
forestcaver
Veteran Wiggler
Posts: 537
Joined: Tue Mar 14, 2017 4:41 pm
Location: UK
Contact:

Re: DIY Mutable Unsuccessful Builds

Post by forestcaver » Wed Jan 15, 2020 11:10 am

oldrobot wrote:
Wed Jan 15, 2020 7:57 am
forestcaver wrote:
Wed Jan 15, 2020 5:59 am
Is that solder paste or flux flooding all the areas between passives eg r39/r40 etc etc (and likely between pins of ic1 - the voltage regulator) ???? Ps looks also like you have dry joints as well as bridges on the stm32 - but always hard to tell from a photo....
What fault finding have you done? Do you get the correct reference voltages as well as the 3.3v rails and +/-12v rails?
(The problem is unlikely to be the firmware....)
Yep, it was the solder paste. I've tried to clean it up as much as I can, still not perfect but here is the new pic.
IMG_1669.jpg

Yes, I have +/- 12V and 3.3V everywhere it should be - 19,32,48,64 pins of stm32 have 3.3v; pin 8 of IC8 receives 12v and pin 4 receives -12V; pin 4 of IC 7 receives 3.3v; as well as pin 8 of IC5 and pins 1,8,14,27 of IC6. Pin 16 of IC4 also receives 3.3v.

So I believe everything is ok with power. And I don't know where to dig next :despair: Maybe one of the chips isn't oriented properly?

Looks like you still have solder paste briding *under* caps and resistors eg c39,c45 to name just two !!!
Seriously - you need to go over that board under high magnification and get rid of all the excess solder paste. You also need to check for resistance across all of your caps and resistors - even after visually cleaning it you may well have bridges underneath. If it was me, I’d take them all off and clean them, then replace them - it would be really quick with hot air, flux, an iron, desolder braid and a microscope. I probably woulndt bother doing anything else until I’d done that..... (tbh I’d also take all the chips off and check for bridges underneath if it was me - doing all of that would take less than an hour....)

Also - your leds are in backwards.... (but they should still light up - proving you used the right parts)

oldrobot
Learning to Wiggle
Posts: 8
Joined: Sun Dec 01, 2019 12:46 pm
Location: Poland

Re: DIY Mutable Unsuccessful Builds

Post by oldrobot » Wed Jan 15, 2020 11:57 am

forestcaver wrote:
Wed Jan 15, 2020 11:10 am
Looks like you still have solder paste briding *under* caps and resistors eg c39,c45 to name just two !!!
Seriously - you need to go over that board under high magnification and get rid of all the excess solder paste. You also need to check for resistance across all of your caps and resistors - even after visually cleaning it you may well have bridges underneath. If it was me, I’d take them all off and clean them, then replace them - it would be really quick with hot air, flux, an iron, desolder braid and a microscope. I probably woulndt bother doing anything else until I’d done that..... (tbh I’d also take all the chips off and check for bridges underneath if it was me - doing all of that would take less than an hour....)

Also - your leds are in backwards.... (but they should still light up - proving you used the right parts)
Checking the resistance across all caps and resistors as well as every single pin of the chip was the first thing I did. But I'll do it one more time :)
Why do you think the LEDs are in backwards? The green is on top and red is below (according to this datasheet and the PCB mark)

User avatar
cnicht
Veteran Wiggler
Posts: 569
Joined: Mon Sep 23, 2013 9:23 am

Re: DIY Mutable Unsuccessful Builds

Post by cnicht » Wed Jan 15, 2020 12:25 pm

The flat on the LED package isn’t facing the same way as the legend on the PCB

oldrobot
Learning to Wiggle
Posts: 8
Joined: Sun Dec 01, 2019 12:46 pm
Location: Poland

Re: DIY Mutable Unsuccessful Builds

Post by oldrobot » Wed Jan 15, 2020 4:32 pm

cnicht wrote:
Wed Jan 15, 2020 12:25 pm
The flat on the LED package isn’t facing the same way as the legend on the PCB
Oh, you're right! And it looks that I found an issue. Looks like IC6 codec is dead. I've soldered it out and now LEDs work as they should. So I'll wait for the new one and hope everything will work.

Thanks for your help :hail:

oldrobot
Learning to Wiggle
Posts: 8
Joined: Sun Dec 01, 2019 12:46 pm
Location: Poland

Re: DIY Mutable Unsuccessful Builds

Post by oldrobot » Wed Jan 15, 2020 5:24 pm

Actually an interesting story..
While my testing I've accidentally torn off the 8th pin of IC6 (don't ask how :hihi: ) So I've decided to uninstall this chip and replace it with the new one. Then I've found out that LEDs work without this chip, so I've decided that it was the reason for the issue.
And then I've decided to experiment a little and soldered it back without 8th leg. And guess what? Now everything works :yay:
The 8th pin is responsible for 3V3_A along with pin #14, so it seems that IC6 can work without 8th pin. :despair:

User avatar
cnicht
Veteran Wiggler
Posts: 569
Joined: Mon Sep 23, 2013 9:23 am

Re: DIY Mutable Unsuccessful Builds

Post by cnicht » Wed Jan 15, 2020 5:34 pm

You were lucky - it's the supply for the headphone output channels which aren't used. :guinness:

User avatar
forestcaver
Veteran Wiggler
Posts: 537
Joined: Tue Mar 14, 2017 4:41 pm
Location: UK
Contact:

Re: DIY Mutable Unsuccessful Builds

Post by forestcaver » Wed Jan 15, 2020 5:39 pm

oldrobot wrote:
Wed Jan 15, 2020 5:24 pm
Actually an interesting story..
While my testing I've accidentally torn off the 8th pin of IC6 (don't ask how :hihi: ) So I've decided to uninstall this chip and replace it with the new one. Then I've found out that LEDs work without this chip, so I've decided that it was the reason for the issue.
And then I've decided to experiment a little and soldered it back without 8th leg. And guess what? Now everything works :yay:
The 8th pin is responsible for 3V3_A along with pin #14, so it seems that IC6 can work without 8th pin. :despair:
Yes - pin 8 is just for the headphone amp part of the wm8731. Not needed in rings. You do need pin 14 though !!!

oldrobot
Learning to Wiggle
Posts: 8
Joined: Sun Dec 01, 2019 12:46 pm
Location: Poland

Re: DIY Mutable Unsuccessful Builds

Post by oldrobot » Wed Jan 15, 2020 6:51 pm

cnicht wrote:
Wed Jan 15, 2020 5:34 pm
You were lucky - it's the supply for the headphone output channels which aren't used. :guinness:
Sometimes break the module is the only way to fix it :lol:

User avatar
sduck
experimental use of gravity
Posts: 13331
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Re: DIY Mutable Unsuccessful Builds

Post by sduck » Thu Jan 16, 2020 10:51 pm

Argghh - I built a mutable edges. Used a complete kit from synthcube, the kind with the preflashed chip (I like these, nice little respite type projects between the larger stuff I do). Everything looked good, no build problems. Finished it up this evening, did an initial test without the panel on, everything checked out, every thing worked. Attached the panel, no worky. First power up, had only one led on, no sound. Next power up, no leds, no sound, no nothing. Yuck. Tried different power cords, no dice. Took panel off again, still no worky. I've given it a bit of stern eyeballing, but haven't found anything obvious yet. I did find some somewhat shady slack soldering on the jacks and pots, I will revisit those tomorrow.
flickr cloud of sound touyube NOT A MODERATOR ANYMORE

User avatar
cnicht
Veteran Wiggler
Posts: 569
Joined: Mon Sep 23, 2013 9:23 am

Re: DIY Mutable Unsuccessful Builds

Post by cnicht » Fri Jan 17, 2020 8:35 am

It's probable that fitting the front panel has flexed the PCB and broken a solder joint or track.

A fun-packed evening of inspection ahead?

User avatar
forestcaver
Veteran Wiggler
Posts: 537
Joined: Tue Mar 14, 2017 4:41 pm
Location: UK
Contact:

Re: DIY Mutable Unsuccessful Builds

Post by forestcaver » Fri Jan 17, 2020 9:14 am

If it is an aluminium panel - could a dodgy led have shorted something eg the atmega to gnd?

User avatar
sduck
experimental use of gravity
Posts: 13331
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Re: DIY Mutable Unsuccessful Builds

Post by sduck » Fri Jan 17, 2020 11:01 am

cnicht wrote:
Fri Jan 17, 2020 8:35 am
It's probable that fitting the front panel has flexed the PCB and broken a solder joint or track.

A fun-packed evening of inspection ahead?
This is my best guess also. I'm not terribly concerned about it though - this isn't a high priority fix. Too much other stuff on my plate.
flickr cloud of sound touyube NOT A MODERATOR ANYMORE

User avatar
frequenzteiler
Learning to Wiggle
Posts: 29
Joined: Wed Nov 14, 2018 12:46 am
Location: Berlin

Re: DIY Mutable Unsuccessful Builds

Post by frequenzteiler » Fri Jan 17, 2020 8:25 pm

im still stuck with marbles.
while i was trying to calibrate following the calibr. pdf i found on forests github i realize x3 does not carry the same voltages like x1 and x2 . in fact it does not change at all when moving the bias knob. tho the led from x3 follows the given voltage . which can be seen when power cycling the module x3 might have a different voltage then before switching the power.
the voltage that comes from the dac for x3 ( going to the opa) is constantly 0.72v - i exchanged the dac thinking it might be damaged , but still the same .
this basically means the mcu is not instructing to change the output for x3 -so softwarewise the value is not changing .
am i missing something ? do i need to force the x outputs to follow the bias using the settings.cc ?

User avatar
forestcaver
Veteran Wiggler
Posts: 537
Joined: Tue Mar 14, 2017 4:41 pm
Location: UK
Contact:

Re: DIY Mutable Unsuccessful Builds

Post by forestcaver » Sat Jan 18, 2020 4:25 am

frequenzteiler wrote:
Fri Jan 17, 2020 8:25 pm
im still stuck with marbles.
while i was trying to calibrate following the calibr. pdf i found on forests github i realize x3 does not carry the same voltages like x1 and x2 . in fact it does not change at all when moving the bias knob. tho the led from x3 follows the given voltage . which can be seen when power cycling the module x3 might have a different voltage then before switching the power.
the voltage that comes from the dac for x3 ( going to the opa) is constantly 0.72v - i exchanged the dac thinking it might be damaged , but still the same .
this basically means the mcu is not instructing to change the output for x3 -so softwarewise the value is not changing .
am i missing something ? do i need to force the x outputs to follow the bias using the settings.cc ?
Dunno! I suspect a soldering error downstream of the dac, but I really dont know.
As a sanity check, if it was me, I’d use the on-chip debugger and see what dac code was sent to the dac for x3 in response to pressing buttons, turning knobs etc. First, I’d *erase* and reflash the chip - flashing it with an unedited fw. Then if that didnt help you, I’d use the on chip debugger as above.
I’d also consider taking off the dac and checking what voltages I got without the dac and see what offsets you had at the dac pins without the chip. You could have a dodgy opamp downstream.... but I really dont know tbh

User avatar
frequenzteiler
Learning to Wiggle
Posts: 29
Joined: Wed Nov 14, 2018 12:46 am
Location: Berlin

Re: DIY Mutable Unsuccessful Builds

Post by frequenzteiler » Sat Jan 18, 2020 4:13 pm

thank you for all the suggestions - i will check further - and try to isolate the issue .
just to make sure ... x1, x2 and x3 should carry the same voltage , right ? ( ..value depending on the bias pot )

i really need a better tool for magnification to check those soldering joints - im using a jewelery hand magnifier of X10 which is ok to quickly check but for longer investigations its horrible .

User avatar
forestcaver
Veteran Wiggler
Posts: 537
Joined: Tue Mar 14, 2017 4:41 pm
Location: UK
Contact:

Re: DIY Mutable Unsuccessful Builds

Post by forestcaver » Sat Jan 18, 2020 4:17 pm

frequenzteiler wrote:
Sat Jan 18, 2020 4:13 pm
just to make sure ... x1, x2 and x3 should carry the same voltage , right ? ( ..value depending on the bias pot )
Yes - providing you set everything as I suggested in the calibration notes !

User avatar
nickajeglin
Learning to Wiggle
Posts: 18
Joined: Thu Mar 07, 2019 7:41 pm
Location: Omaha

Re: DIY Mutable Unsuccessful Builds

Post by nickajeglin » Sat Jan 18, 2020 8:09 pm

Hello everyone,

Just finished a clouds build. Everything seems in order except that when I turn the texture knob up past a certain point, there is a loud squeal of feedback and the module reboots. I noticed this in the MI clouds manual: "I. Grain TEXTURE...Past 2 o’clock, activates a diffuser which smears transients."

It seems like past 2 o'clock is when the issue happens. Does anyone know what exactly "diffuser which smears transients" means? I'm having trouble translating that to a hardware level so I can figure out what to check. I'm currently checking everything along the mc's reset line to look for shorts, but haven't found anything yet.

Does anyone have test points for voltage references? I have the schematic, but still looking for a nudge in the right direction.

User avatar
nickajeglin
Learning to Wiggle
Posts: 18
Joined: Thu Mar 07, 2019 7:41 pm
Location: Omaha

Re: DIY Mutable Unsuccessful Builds

Post by nickajeglin » Sat Jan 18, 2020 8:14 pm

nickajeglin wrote:
Sat Jan 18, 2020 8:09 pm
Hello everyone,

Just finished a clouds build. Everything seems in order except that when I turn the texture knob up past a certain point, there is a loud squeal of feedback and the module reboots. I noticed this in the MI clouds manual: "I. Grain TEXTURE...Past 2 o’clock, activates a diffuser which smears transients."

It seems like past 2 o'clock is when the issue happens. Does anyone know what exactly "diffuser which smears transients" means? I'm having trouble translating that to a hardware level so I can figure out what to check. I'm currently checking everything along the mc's reset line to look for shorts, but haven't found anything yet.

Does anyone have test points for voltage references? I have the schematic, but still looking for a nudge in the right direction.
Correction, it is the size knob that causes the issue. Not the texture knob. I just fired it up to double check.

User avatar
nickajeglin
Learning to Wiggle
Posts: 18
Joined: Thu Mar 07, 2019 7:41 pm
Location: Omaha

Re: DIY Mutable Unsuccessful Builds

Post by nickajeglin » Sun Jan 19, 2020 2:23 am

nickajeglin wrote:
Sat Jan 18, 2020 8:14 pm
nickajeglin wrote:
Sat Jan 18, 2020 8:09 pm
Hello everyone,

Just finished a clouds build. Everything seems in order except that when I turn the texture knob up past a certain point, there is a loud squeal of feedback and the module reboots. I noticed this in the MI clouds manual: "I. Grain TEXTURE...Past 2 o’clock, activates a diffuser which smears transients."

It seems like past 2 o'clock is when the issue happens. Does anyone know what exactly "diffuser which smears transients" means? I'm having trouble translating that to a hardware level so I can figure out what to check. I'm currently checking everything along the mc's reset line to look for shorts, but haven't found anything yet.

Does anyone have test points for voltage references? I have the schematic, but still looking for a nudge in the right direction.
Correction, it is the size knob that causes the issue. Not the texture knob. I just fired it up to double check.
Disregard, I gottem. I was sure it had to be a short on the mc somewhere, especially because the size pot input pin and reset pin are right next to each other (7 and 8). Eventually I flipped the board over and noticed there was a tiny little short between 2 pot legs. It wasn't resetting because I was pulling down the reset pin, but because I was shorting 3v3 to gnd through the pot.

This also explains why I had such a hard time getting it going in the first place. 3v3 was shorted to ground, thought it was a short on the mc pins, reflowed, still couldn't find it, then it just started working for some reason. I must've had the size pin turned up enough to short it, then accidentally lowered it at some point.

collumn2016
Learning to Wiggle
Posts: 43
Joined: Mon Aug 29, 2016 4:42 am
Location: in orbit

Cannot connect to tides 2 to flash firmware

Post by collumn2016 » Sun Jan 19, 2020 11:34 am

delete
Last edited by collumn2016 on Wed Jan 22, 2020 12:57 pm, edited 1 time in total.

StrangeCaptain
Common Wiggler
Posts: 75
Joined: Sun Sep 15, 2019 8:19 am
Location: Detroit

Re:

Post by StrangeCaptain » Sun Jan 19, 2020 6:48 pm

batchas wrote:
Sun Nov 24, 2019 2:12 pm
StrangeCaptain wrote:when I look at the ARM now I can see pins that don;t look soldered but something is connected because the program is definitely running.
Looking at HP photo: as mentionned in my previous post it looks like there's a bridge between pin 35-36 on UC1.
Also some caps do not look like soldered when seen from the top.
C17 also looks weird.

You need a DVM with continuity check. This is what I'd pers. do before moving to the scope.
Checking that the components are soldered, I mean connecting where they should. You look where the trace goes and check between the components that are supposed to be connected, because the soldering does not look ok.

Trying to explain with the 3 pictures I'm posting here. I only take this as an example, but I'd do the same in most places where the solder is critical, starting with UC1 pins.
If you don't see where the traces goes, you can use an application (diptrace, kicad, eagle and so on) to open the pcb and see clearly on your computer.

Image

You can check both sides, in case both components don't look soldered.
Image

Image

OK
I have finished a few more through hole builds and am now back to trying to fix this module before moving to something else for an oscillator.

I now have NO display in the first LED position, CSAW shows SAW.

Have checked all the IC pins and they are connected to the Caps they are supposed to be. Got a Multimeter with Continuity
I have reflowed a few caps and still no sound.

Trying to run out all the possibilities before scrapping this project
I will check all the ARM pins tonight
any thoughts?

c0ntr4d1ct10n
Common Wiggler
Posts: 214
Joined: Mon Aug 13, 2018 9:02 am
Location: Australia

Re: DIY Mutable Unsuccessful Builds

Post by c0ntr4d1ct10n » Mon Jan 20, 2020 11:51 am

I just built an Ears module and the 5000mcd LEDs (as recommended by the bom) don't seem to dim in brightness when I reduce the volume when I turn down the B50k pot. I wonder if this is something wrong with my connections or is it to do with the LEDs I'm using.

User avatar
cnicht
Veteran Wiggler
Posts: 569
Joined: Mon Sep 23, 2013 9:23 am

Re: DIY Mutable Unsuccessful Builds

Post by cnicht » Mon Jan 20, 2020 12:42 pm

LED1 and LED3 are driven from comparators so they won’t vary much in brightness.

LED2 brightness should vary with the amplitude of the signal amplitude.

Post Reply

Return to “Music Tech DIY”