DIY Mutable Unsuccessful Builds

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

Moderators: lisa, luketeaford, Kent, Joe.

Post Reply
Still Sound
Learning to Wiggle
Posts: 8
Joined: Tue Jun 23, 2020 11:39 pm

Re: DIY Mutable Unsuccessful Builds

Post by Still Sound » Thu Jun 25, 2020 6:38 pm

Altitude909 wrote:
Thu Jun 25, 2020 6:25 pm
Yup. Dont forget the little header, I dont even bother soldering those in, just use friction
Boom!

Thanks alot.

Parts ordered.

soupspoon
Learning to Wiggle
Posts: 20
Joined: Thu May 28, 2020 3:45 pm

Re: DIY Mutable Unsuccessful Builds

Post by soupspoon » Fri Jun 26, 2020 12:49 pm

Sorry for the delay! Also, sorry the boards are filthy... I'm waiting on some flux remover, so hopefully I'll address the mess before I need to send more pictures.

Image
Image

User avatar
batchas
Super Deluxe Wiggler
Posts: 4654
Joined: Wed Nov 09, 2011 2:51 pm

Re: DIY Mutable Unsuccessful Builds

Post by batchas » Fri Jun 26, 2020 1:24 pm

soupspoon wrote:
Fri Jun 26, 2020 12:49 pm
Sorry for the delay! Also, sorry the boards are filthy... I'm waiting on some flux remover, so hopefully I'll address the mess before I need to send more pictures.
No time to go through the thread to know what is the issue you have on this, so most likely not related, but did you check the second pin top left (middle pin from 3 pins). Looks like a cold joint.
-----------------------------------
Batchas website
Bandcamp page

User avatar
cnicht
Veteran Wiggler
Posts: 699
Joined: Mon Sep 23, 2013 9:23 am

Re: DIY Mutable Unsuccessful Builds

Post by cnicht » Fri Jun 26, 2020 5:19 pm

It's difficult, with the resolution of the photograph, but some of the microcontroller pins appear to be bridged.

User avatar
Flounderguts
Wiggling with Experience
Posts: 447
Joined: Thu Dec 07, 2017 5:56 pm
Location: SLC

Re: DIY Mutable Unsuccessful Builds

Post by Flounderguts » Sat Jun 27, 2020 12:31 am

How hot are you running your iron?
----------------------

Flounderguts

soupspoon
Learning to Wiggle
Posts: 20
Joined: Thu May 28, 2020 3:45 pm

Re: DIY Mutable Unsuccessful Builds

Post by soupspoon » Sat Jun 27, 2020 9:52 am

cnicht wrote:
Fri Jun 26, 2020 5:19 pm
It's difficult, with the resolution of the photograph, but some of the microcontroller pins appear to be bridged.
I can reaffirm that the microcontroller pins are not bridged, I've been using a jeweler's loupe to verify that they aren't connected. My camera wouldn't focus properly for that region, so I guess my word will have to be good enough.
Flounderguts wrote:
Sat Jun 27, 2020 12:31 am
How hot are you running your iron?
I think it's set to 480F.

I'll try to fix the cold joint, since that particular pin is the input for the op-amp?

bawbag
Learning to Wiggle
Posts: 20
Joined: Thu Jul 18, 2019 1:29 am

Re: DIY Mutable Unsuccessful Builds

Post by bawbag » Sat Jun 27, 2020 9:58 am

soupspoon wrote:
Sat Jun 27, 2020 9:52 am
I think it's set to 480F.
Way too low.
Should be around 600-650F

User avatar
Flounderguts
Wiggling with Experience
Posts: 447
Joined: Thu Dec 07, 2017 5:56 pm
Location: SLC

Re: DIY Mutable Unsuccessful Builds

Post by Flounderguts » Sat Jun 27, 2020 11:25 am

bawbag wrote:
Sat Jun 27, 2020 9:58 am
soupspoon wrote:
Sat Jun 27, 2020 9:52 am
I think it's set to 480F.
Way too low.
Should be around 600-650F
Yes. The joints looked like they weren't quite melted.

The idea is to puddle the solder long enough for the surface tension to snap to the outside of the pads. By setting your iron to the minimum heating point, you have to wait while it barely melts...which means you soak *more* heat into the joint. Higher energy for less time won't damage things as much.

I have a very small tip on my iron, and I run it around 720 F, being careful not to dwell on a joint for more than about 2 seconds. Less, if it's a diode or something else sensitive.

I bet if you clean the flux off that board, and then go back in with a tiny bit of liquid flux and a hot iron, all the problems will disappear.
----------------------

Flounderguts

User avatar
OrganFixer
Common Wiggler
Posts: 167
Joined: Wed Feb 05, 2014 4:38 pm
Location: New York City

Re: DIY Mutable Unsuccessful Builds

Post by OrganFixer » Sun Jun 28, 2020 4:07 pm

I should start by saying I don't know what I'm doing. I thought I would give this mutable stuff a try. I bought a few boards and decided to start with the module tester. I've got it built but I don't think I have the right connector to program it. Earlier I did some quick poking around and purchased the ST-Link V2 and the olimex 20 to 10 adapter based on some stuff I read but now I'm thinking I need different tools for different modules. I guess my question is what hardware do I need to program the module tester V2?

User avatar
cnicht
Veteran Wiggler
Posts: 699
Joined: Mon Sep 23, 2013 9:23 am

Re: DIY Mutable Unsuccessful Builds

Post by cnicht » Sun Jun 28, 2020 5:09 pm

You need the ATMEL microcontroller programmer AVRISPMkII or equivalent

User avatar
AlanP
Ultra Wiggler
Posts: 953
Joined: Fri Mar 14, 2014 8:37 pm
Location: New Zealand

Re: DIY Mutable Unsuccessful Builds

Post by AlanP » Sun Jun 28, 2020 5:46 pm

The ST-Link is for programming STM32 based microprocessors (which most Mutable stuff is based on.)

The Module Tester runs in an Atmega processor, which needs a different programmer (like the AVRISPMkII mentioned above.)

User avatar
OrganFixer
Common Wiggler
Posts: 167
Joined: Wed Feb 05, 2014 4:38 pm
Location: New York City

Re: DIY Mutable Unsuccessful Builds

Post by OrganFixer » Sun Jun 28, 2020 6:29 pm

AlanP wrote:
Sun Jun 28, 2020 5:46 pm
The ST-Link is for programming STM32 based microprocessors (which most Mutable stuff is based on.)

The Module Tester runs in an Atmega processor, which needs a different programmer (like the AVRISPMkII mentioned above.)
Thank you for clarifying. That's what it was looking like but I didn't want to jump in and buy something else if I didn't need it. Ok, something new to figure out.

User avatar
Flounderguts
Wiggling with Experience
Posts: 447
Joined: Thu Dec 07, 2017 5:56 pm
Location: SLC

Re: DIY Mutable Unsuccessful Builds

Post by Flounderguts » Sun Jun 28, 2020 10:39 pm

I use AVRDude and a USBTiny ASP. There is a lot of good information on the Mutable Instruments forum as well.
----------------------

Flounderguts

User avatar
OrganFixer
Common Wiggler
Posts: 167
Joined: Wed Feb 05, 2014 4:38 pm
Location: New York City

Re: DIY Mutable Unsuccessful Builds

Post by OrganFixer » Sun Jun 28, 2020 11:06 pm

Do you mean the one on adafruit? I saw that but it was out of stock. The one everyone recommends (AVRISP Mk2) seems to be obsolete now. I considered buying the newer model but I've seen so much stuff about only certain models working I wanted to get the exact one. I ended up getting this one. I combined it with my Rings order so I'm in deeper now.

https://www.mouser.com/ProductDetail/Ol ... osVA%3D%3D

Edit: Worked fine using the steps in this video
Last edited by OrganFixer on Fri Jul 03, 2020 8:55 am, edited 1 time in total.

User avatar
Flounderguts
Wiggling with Experience
Posts: 447
Joined: Thu Dec 07, 2017 5:56 pm
Location: SLC

Re: DIY Mutable Unsuccessful Builds

Post by Flounderguts » Sun Jun 28, 2020 11:18 pm

Yeah...the adafruit one is what I got. Quick build, and in most cases I'm working with DIP Atmel chips, so I made one of those Synthroteck ZIF skiffs as well. I only use the ISP header for boards that have QFP microprocessors.

I think the reason so many mutable builders go with the AVRISP MK2 is because all of Emilie's documentation uses it. If you aren't into learning to compile, set fuses, and just want to get your damn module working, it's much easier if you can cut and paste her instructions. OTOH, if you actually want to learn wtf is going on, then translating the text strings and learning what all of it means can be fun. Building a Skookum Anookum (Anushri clone) is directly responsible for me learning about Arduinos...and then into that long, dark rabbit hole that (so far) has led to FPGA stuff.
----------------------

Flounderguts

Still Sound
Learning to Wiggle
Posts: 8
Joined: Tue Jun 23, 2020 11:39 pm

Re: DIY Mutable Unsuccessful Builds

Post by Still Sound » Sun Jun 28, 2020 11:31 pm

Altitude909 wrote:
Wed Jun 24, 2020 2:57 pm
well it sounds like it works, its powering up. The 4 lit LEDs are how its supposed to look like before you upload code to it, not sure why you would toss it at this stage. There is no option to uploading code, you're going to have to do that on any MI diy project unless you get a preflashed chip
Thanks for your help.

Got her up and running.

Most appreciated.

Still Sound
Learning to Wiggle
Posts: 8
Joined: Tue Jun 23, 2020 11:39 pm

Re: DIY Mutable Unsuccessful Builds

Post by Still Sound » Tue Jun 30, 2020 5:43 am

No output on the LEFT channel.

Also, I think the sounds might be "noisier" than they should be. Most noticeable when pluckin it in Resonestor.

Any advice?

Thanks

User avatar
hndmrsh
Learning to Wiggle
Posts: 10
Joined: Mon Feb 19, 2018 3:13 am
Location: Wellington, NZ

Re: DIY Mutable Unsuccessful Builds

Post by hndmrsh » Fri Jul 03, 2020 5:46 am

Hi again! I posted a few months about my not-quite-working Rings build. I've finally got a better multimeter, and so am better equipped to debug my issue!

My issue seems to be that the module is drawing too much current, and the microcontroller is overheating and shutting down after some time (I believe it's the microcontroller that's shutting down, as it behaves strangely for a few seconds before it switches off). The module draws 270-280mA on the +12V rail, and 5mA - the expected values are 110mA and 5mA respectively. Additionally, the module will not power up at all when running off a uZeus (which is only powering the Rings) - it only works in my main case with Trogotronic m15 power.
extralifedisco wrote:
Sat Mar 21, 2020 6:08 pm
That's pretty mystifying. It could be a leaky electrolytic capacitor - they should have basically infinite resistance, but if they're old or they've been overheated sometimes they don't. This can cause higher current draw than expected, overheat resistors, lower expected voltages. Have you measured the current draw on both rails and compared to spec?

If you want a quick and dirty test, power it up and wait until it shuts down. Then connect a lead to the PCB ground (alligator clip works nicely), and touch the other end to each capacitor's pads. If grounding one of them causes it to reset and turn back on, that cap is probably bad.
@extralifedisco, if you're still around, can you please clarify which capacitor(s) you're talking about in this message? I'm an electronics novice, but my guess is you're referring to C2, C3, and C8 (these are the 47u electrolytic caps before and after the 3.3V regulator on the Vcc line). Also, by "grounding a cap", does that mean shorting the Vcc/3V3 line to ground (i.e. the pads I've circled in blue below)?

Thanks!

Image

User avatar
extralifedisco
Learning to Wiggle
Posts: 30
Joined: Thu Apr 05, 2018 5:03 pm
Location: San Francisco, CA

Re: DIY Mutable Unsuccessful Builds

Post by extralifedisco » Fri Jul 03, 2020 7:26 pm

Interesting. It could be a capacitor, but that's an awful lot of current - that really is more likely to be a short circuit somewhere or a resistor value that's too small by a few orders of magnitude. A thermal camera would probably show you exactly where - or the touch test to see what heats up, but be careful not to burn your fingers. I would start by desoldering the LM1117 3.3v regulator. If removing that causes the current draw to drop down below ~50mA, then the short is probably somewhere along the 3.3v rail. If not, it's probably on the +12v rail.

It's possible the short is in the microcontroller itself, possibly due to an overvoltage on an input pin (have done this myself a few times). As a last resort be prepared to desolder it, test the current draw, and replace with a new MCU if needed.

User avatar
OrganFixer
Common Wiggler
Posts: 167
Joined: Wed Feb 05, 2014 4:38 pm
Location: New York City

Re: DIY Mutable Unsuccessful Builds

Post by OrganFixer » Fri Jul 03, 2020 9:41 pm

Here's the update on my module tester. Note I'm still in the unsuccessful builds thread but I think it was just due to my dumb mistake. When soldering the sockets I noticed that "all" the ICs were in the same direction. Well it turns out all except the Atmega. I didn't realize this until the hex file had problems uploading. I flipped it around, tried again, and the hex file installed but now I have issues.

All the LEDs are always lit the buttons to switch modes have no effect, Clicking the encoder switch works but twisting it does not. If I put the oscilloscope on the gate out, CV out, and audio out they all seem to be working. I confirmed the encoder and buttons are doing what they are supposed to do but it seems the ICs do not react to the incoming data. My question is this. By powering up and trying to load the hex file with the processor reversed is it possible that I just messed up a few things even though the program loads and a lot of the stuff seems to be working? Also, shouldn't IO_SCK have a clock signal? Right now it seems to be grounded all the time. My assumption is that this is the cause of most of my problems. My plan is to get a new processor and try again. This is my first AVR thing ever so I am not sure how to interpret the symptoms I'm seeing.

Does anyone want to confirm my assumptions? Or am I looking at this wrong?

User avatar
Datum
Learning to Wiggle
Posts: 48
Joined: Thu Jul 04, 2019 3:17 pm

Re: DIY Mutable Unsuccessful Builds

Post by Datum » Mon Jul 06, 2020 12:22 pm

Hey everyone, I'm programming my first module - Branches and... for the life of me I can't compile the .hex from the makefile. I have zero experience but have been following the guide here: https://gist.github.com/arirusso/88e5f4 ... 25cea74581 and running into nothing but issues on Windows 10.

I was able to upload the Twigs firmware using avrdudess because he included the .hex file in the download. So I know I am able to flash the firmware using by USB ASP, but I just need the Branches .hex file.

Does anyone have the Branches .hex handy that I could download?

User avatar
hndmrsh
Learning to Wiggle
Posts: 10
Joined: Mon Feb 19, 2018 3:13 am
Location: Wellington, NZ

Re: DIY Mutable Unsuccessful Builds

Post by hndmrsh » Mon Jul 06, 2020 7:50 pm

extralifedisco wrote:
Fri Jul 03, 2020 7:26 pm
Interesting. It could be a capacitor, but that's an awful lot of current - that really is more likely to be a short circuit somewhere or a resistor value that's too small by a few orders of magnitude. A thermal camera would probably show you exactly where - or the touch test to see what heats up, but be careful not to burn your fingers. I would start by desoldering the LM1117 3.3v regulator. If removing that causes the current draw to drop down below ~50mA, then the short is probably somewhere along the 3.3v rail. If not, it's probably on the +12v rail.

It's possible the short is in the microcontroller itself, possibly due to an overvoltage on an input pin (have done this myself a few times). As a last resort be prepared to desolder it, test the current draw, and replace with a new MCU if needed.
I removed the regulator, and it caused the current to drop to ~4mA (though the first time I tested it it remained at 260mA, so I spent a while hunting around the 12V rail for shorts... not sure why that would have happened, a capacitor holding charge maybe?).

I spent a few hours this weekend trying to find shorts or bad joints, with no luck - I did discover that the C49 wasn't soldered on correctly and so wasn't connected to ground, but fixing that didn't solve my issue unfortunately. I think I've checked just about every possible location for shorts and bad joints at this point, and I doubt I would have used the wrong resistor value anywhere as I only ordered the exact number of components I needed, so that would have caused me to not have the right parts later in the build... Not saying it's impossible of course, but seems unlikely!

The only source of heat seems to be the 3V3 regulator (which gets very hot) - the surrounding components get warm too, but suspect that's just from the heat put out by the regulator.

I'm at a bit of a loss for what to check now... Could it only be the MCU at this point? Could the module be operating normally (but with higher current draw) as it is now if there was a short in the MCU itself? Could it be the regulator?

Appreciate all the help - as frustrating as this is, I'm learning a lot!

User avatar
extralifedisco
Learning to Wiggle
Posts: 30
Joined: Thu Apr 05, 2018 5:03 pm
Location: San Francisco, CA

Re: DIY Mutable Unsuccessful Builds

Post by extralifedisco » Tue Jul 07, 2020 6:39 pm

hndmrsh wrote:
Mon Jul 06, 2020 7:50 pm

I'm at a bit of a loss for what to check now... Could it only be the MCU at this point? Could the module be operating normally (but with higher current draw) as it is now if there was a short in the MCU itself? Could it be the regulator?

Appreciate all the help - as frustrating as this is, I'm learning a lot!
That is quite possible. I once overvolted an Atmega328p, and afterwards the chip still worked but one of the ADC inputs was unresponsive and it did funny things with the clock speed. Eventually it overheated and cooked itself. Maybe the STM32 has better thermal self-protection and shuts itself off.

ICs have different failure modes due to the small scale of the silicon layers. They use 3.3v or lower to reduce the size of the parts they can get away with using, and the tolerance is quite fine. If you accidentally hit an 3.3v IC pin with 5v or more, it will often "break through" one of the insulating layers and create a conducting path from that silicon layer to the nearest one with lower potential, often to ground. This can also happen with plain old ESD, hence the static wrist straps recommended for installing chips. Here's a white paper from AD on IC failure modes:
https://www.analog.com/media/en/trainin ... pter11.pdf

At this point, if you're confident you've ruled out everything else, I'd go ahead and replace the chip. Even if it doesn't fix it, the component price is worth saving yourself an hour of frustration poking at a half-working board.

snowtires
Learning to Wiggle
Posts: 7
Joined: Fri Jun 03, 2016 12:58 pm

Re: DIY Mutable Unsuccessful Builds

Post by snowtires » Tue Jul 07, 2020 8:35 pm

I've got a Veils build with only the second and third channels working, LEDs work on channel 4 (and I'm assuming 1, if it got signal), as it flashes along with anything put into the working inputs. I feel like this means the OPA1654 chips are shot on those two channels, but if anyone else has seen this problem before, please let me know before I spend more money!

User avatar
OrganFixer
Common Wiggler
Posts: 167
Joined: Wed Feb 05, 2014 4:38 pm
Location: New York City

Re: DIY Mutable Unsuccessful Builds

Post by OrganFixer » Tue Jul 07, 2020 11:43 pm

Datum wrote:
Mon Jul 06, 2020 12:22 pm
Hey everyone, I'm programming my first module - Branches and... for the life of me I can't compile the .hex from the makefile. I have zero experience but have been following the guide here: https://gist.github.com/arirusso/88e5f4 ... 25cea74581 and running into nothing but issues on Windows 10.

I was able to upload the Twigs firmware using avrdudess because he included the .hex file in the download. So I know I am able to flash the firmware using by USB ASP, but I just need the Branches .hex file.

Does anyone have the Branches .hex handy that I could download?
I just went through my eight hour struggle of of figuring out how to compile hex files. It is actually all written here https://github.com/pichenettes/mutable-dev-environment if you download all the software and everything they list, step by step, it just works. Like I said it took me eight hours to get it to work but it was worth it. But if you just want me to compile it and send it to you I will. Let me know.

Post Reply

Return to “Music Tech DIY”