Announce Thread: Magpie Modular - uBraids - Braids in 8 HP.

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

Moderators: Kent, luketeaford, Joe.

Post Reply
User avatar
mwvm
Ultra Wiggler
Posts: 890
Joined: Sat Aug 09, 2014 3:19 am

Post by mwvm » Thu Oct 27, 2016 1:29 am

ZZ Ardoz wrote:Does anyone have a regular Braids they could check this against? I'm hearing a noise floor in Wavetable modes (particularly loud in WTx4)

If you tune course all the way down, you'll (maybe) hear it. I'm getting it on 3 builds here, identical, mostly on those wavetable modes. All using the latest runs of pcb from Magpie.

Is anyone else getting this, and if you have a regular Braids, is that also doing it?

Is it possibly a software issue, as most of the modes are without this noise?

Would appreciate any input - here's a sample of the noise above a very low wavetable sound

[soundcloud url="https://api.soundcloud.com/tracks/290093428" params="auto_play=false&hide_related=false&show_comments=true&show_user=true&show_reposts=false&visual=true" width="100%" height="450" iframe="true" /]
Yes. I get this noise.

User avatar
mwvm
Ultra Wiggler
Posts: 890
Joined: Sat Aug 09, 2014 3:19 am

Post by mwvm » Thu Oct 27, 2016 1:42 am

igowen wrote:Two issues i've encountered with both of my builds:
1) the amount the display is lit affects the output pitch (maybe due to power draw in the display?)
2) there's a lot of noise in meta mode

anyone know if there's any way to address these?
I get the same on both. Weird.

User avatar
Southfork
Veteran Wiggler
Posts: 667
Joined: Mon Oct 06, 2014 12:55 pm

Post by Southfork » Thu Oct 27, 2016 3:16 am

GryphonP3 wrote:
ZZ Ardoz wrote:
You can either edit it terminal or in a text editor. There's a ton of references to the file here and on the mutable site
ah, got it. Hmm... Still timing out, taking it all the way down to 41K in 10k increments. Driving me crazy! It almost got there a few tries ago.. through the second set of numbers are letters in the READ phase. Oh well. I am well over 100 tried now in the past 3 hours. Maybe I will just get lucky eventually :) Seems like I am doing everything right, my el capitan machine just doesn't want to work.
I take it you're using OS X? There's a problem from El Capitan that causes Ftdi to randomly drop connection. Some people claimed to have solved it with drivers but never worked for me. Just sit tight, keep plugging and eventually it works.

User avatar
mwvm
Ultra Wiggler
Posts: 890
Joined: Sat Aug 09, 2014 3:19 am

Post by mwvm » Thu Oct 27, 2016 3:58 am




At least I have couple of merzbow braids!!!

User avatar
Southfork
Veteran Wiggler
Posts: 667
Joined: Mon Oct 06, 2014 12:55 pm

Post by Southfork » Thu Oct 27, 2016 7:42 am

Can anyone confirm they have meta mode working perfectly on any of there builds? My one does not. Otherwise fully functioning.

User avatar
ZZ Ardoz
Wiggling with Experience
Posts: 496
Joined: Wed Sep 02, 2015 4:12 pm
Location: Los Angeles

Post by ZZ Ardoz » Thu Oct 27, 2016 8:18 am

Southfork wrote:Can anyone confirm they have meta mode working perfectly on any of there builds? My one does not. Otherwise fully functioning.
Works fine as far as switching via fm on 4 builds

User avatar
Southfork
Veteran Wiggler
Posts: 667
Joined: Mon Oct 06, 2014 12:55 pm

Post by Southfork » Thu Oct 27, 2016 8:28 am

Flips fine via FM input but Audio has signs of corruption compared to my factory braids, will do more tests.

User avatar
mwvm
Ultra Wiggler
Posts: 890
Joined: Sat Aug 09, 2014 3:19 am

Post by mwvm » Thu Oct 27, 2016 8:43 am

ZZ Ardoz wrote:
Southfork wrote:Can anyone confirm they have meta mode working perfectly on any of there builds? My one does not. Otherwise fully functioning.
Works fine as far as switching via fm on 4 builds
Hmmmm when were these built?

Both mine produce noise... Obviously the merzbow thing is sarcasm.

User avatar
ZZ Ardoz
Wiggling with Experience
Posts: 496
Joined: Wed Sep 02, 2015 4:12 pm
Location: Los Angeles

Post by ZZ Ardoz » Thu Oct 27, 2016 9:49 am

mwvm wrote:
ZZ Ardoz wrote:
Southfork wrote:Can anyone confirm they have meta mode working perfectly on any of there builds? My one does not. Otherwise fully functioning.
Works fine as far as switching via fm on 4 builds
Hmmmm when were these built?

Both mine produce noise... Obviously the merzbow thing is sarcasm.
One from the first run, 3 from the second. Other than the noise on the wavetable settings I described before, functions seem to be as normal

User avatar
mwvm
Ultra Wiggler
Posts: 890
Joined: Sat Aug 09, 2014 3:19 am

Post by mwvm » Thu Oct 27, 2016 11:05 am

ZZ Ardoz wrote:
mwvm wrote:
ZZ Ardoz wrote:
Southfork wrote:Can anyone confirm they have meta mode working perfectly on any of there builds? My one does not. Otherwise fully functioning.
Works fine as far as switching via fm on 4 builds
Hmmmm when were these built?

Both mine produce noise... Obviously the merzbow thing is sarcasm.
One from the first run, 3 from the second. Other than the noise on the wavetable settings I described before, functions seem to be as normal

Bees in trees or standard firmware?

User avatar
ZZ Ardoz
Wiggling with Experience
Posts: 496
Joined: Wed Sep 02, 2015 4:12 pm
Location: Los Angeles

Post by ZZ Ardoz » Thu Oct 27, 2016 11:44 am

Standard

Breakinge
Learning to Wiggle
Posts: 3
Joined: Sun Feb 23, 2014 1:26 pm
Location: Rennes - France

Post by Breakinge » Thu Nov 03, 2016 6:51 am

Built 2.
On the first one I've been able to upload the display code on the Atmega once, and after I tried to upload igowen's code (the one with the display that looks like the original Braids).
It failed, and now I can't get the communication between Arduino IDE and the chip (sck problem, and my programmer is not seen by IDE anymore).
I will check for a bad solder joint. I just hope I didn't fucked the atmega trying to upload multiple times the code, different bootloaders etc...

NOTE : I've got the same noise issue in META mode. The noise is on every preset, and even if no FM is applied.

On the second one, no problem to upload the display code, but I have a glitchy horizontal segment. That segment is never where it should be (ie displays on another letter, not on the letter it should be displayed)

On the 2 units, I couldn't upload the firmware via FTDI until the end, it always failed in the middle of the process.

But I could finish the upload via audio bootloader with no issue.

User avatar
Faustgeist
Wiggling with Experience
Posts: 500
Joined: Thu Apr 16, 2015 1:48 pm
Location: Columbia, SC

Post by Faustgeist » Thu Nov 03, 2016 8:31 am

Greetings all,

I received my PCB and panel recently for the MicroBRAIDS.

I am now searching for a Mouser BOM, is there a link floating around anywhere?

My sincere thanks!
~Robin
Seeking the esoteric and funky in 5U/COTK/MOTM and MODCAN B, send me a message if you are selling gear and thanks!

User avatar
ZZ Ardoz
Wiggling with Experience
Posts: 496
Joined: Wed Sep 02, 2015 4:12 pm
Location: Los Angeles

Post by ZZ Ardoz » Thu Nov 03, 2016 8:36 am

Faustgeist wrote:Greetings all,

I received my PCB and panel recently for the MicroBRAIDS.

I am now searching for a Mouser BOM, is there a link floating around anywhere?

My sincere thanks!
~Robin
On the magpie product page

User avatar
ZZ Ardoz
Wiggling with Experience
Posts: 496
Joined: Wed Sep 02, 2015 4:12 pm
Location: Los Angeles

Post by ZZ Ardoz » Thu Nov 03, 2016 8:44 am

Breakinge wrote:Built 2.
On the first one I've been able to upload the display code on the Atmega once, and after I tried to upload igowen's code (the one with the display that looks like the original Braids).
It failed, and now I can't get the communication between Arduino IDE and the chip (sck problem, and my programmer is not seen by IDE anymore).
I will check for a bad solder joint. I just hope I didn't fucked the atmega trying to upload multiple times the code, different bootloaders etc...

NOTE : I've got the same noise issue in META mode. The noise is on every preset, and even if no FM is applied.

On the second one, no problem to upload the display code, but I have a glitchy horizontal segment. That segment is never where it should be (ie displays on another letter, not on the letter it should be displayed)

On the 2 units, I couldn't upload the firmware via FTDI until the end, it always failed in the middle of the process.

But I could finish the upload via audio bootloader with no issue.
Kris is working on a solution to the noise issue. He believes it is the OLED drawing too much power.

As far as I know, you can't break the chip by loading it multiple times. The programmer not being seen looks like a good place to start debugging. What system and programmer are you using?

The glitch is most likely related to a solder issue with one of the trio of chips that sort the display out. I had a segment issue on one, and it was fixed by reflowing and cleaning.

Breakinge
Learning to Wiggle
Posts: 3
Joined: Sun Feb 23, 2014 1:26 pm
Location: Rennes - France

Post by Breakinge » Thu Nov 03, 2016 9:23 am

ZZ Ardoz wrote:
Breakinge wrote:Built 2.
On the first one I've been able to upload the display code on the Atmega once, and after I tried to upload igowen's code (the one with the display that looks like the original Braids).
It failed, and now I can't get the communication between Arduino IDE and the chip (sck problem, and my programmer is not seen by IDE anymore).
I will check for a bad solder joint. I just hope I didn't fucked the atmega trying to upload multiple times the code, different bootloaders etc...

NOTE : I've got the same noise issue in META mode. The noise is on every preset, and even if no FM is applied.

On the second one, no problem to upload the display code, but I have a glitchy horizontal segment. That segment is never where it should be (ie displays on another letter, not on the letter it should be displayed)

On the 2 units, I couldn't upload the firmware via FTDI until the end, it always failed in the middle of the process.

But I could finish the upload via audio bootloader with no issue.
Kris is working on a solution to the noise issue. He believes it is the OLED drawing too much power.

As far as I know, you can't break the chip by loading it multiple times. The programmer not being seen looks like a good place to start debugging. What system and programmer are you using?

The glitch is most likely related to a solder issue with one of the trio of chips that sort the display out. I had a segment issue on one, and it was fixed by reflowing and cleaning.
Thanks, I will check the 3 chips, reflow and clean. I wouldn't be surprised if I find a bad solder joint on the middle chip ;)

I use a USBasp with Arduino IDE on Mac os x (Sierra).

On the 2 modules I had lots of errors with the programmer not being seen, or sck / sync issues, or impossible to upload the bootloader, before I decided to make a better 10 to 6 pin cable.
It solved the issues for one module, my DIY cable was faulty.

But for the other module... always the same errors. So I think it's a bad solder joint somewhere (tried to make a continuity test between programmer and chip, and all looks good, I should give it another try).

The weird fact is that I've been able to upload the display code once on that module, but now I can't anymore.
Maybe a bad reflow, I will check and let you know

adam
Super Deluxe Wiggler
Posts: 2950
Joined: Wed Dec 25, 2013 6:24 am

Post by adam » Thu Nov 03, 2016 11:53 am

if the wrong fuse settings are used you can brick an avr chip for all intents and purposes, there's even a "do not allow further writes" fuse

User avatar
igowen
Wiggling with Experience
Posts: 282
Joined: Wed Jul 01, 2015 10:07 pm
Location: pittsburgh

Post by igowen » Thu Nov 03, 2016 5:36 pm

ZZ Ardoz wrote: Kris is working on a solution to the noise issue. He believes it is the OLED drawing too much power.
this is good to hear, looking forward to a fix.
As far as I know, you can't break the chip by loading it multiple times.
i certainly hope this is the case, i probably flashed mine 20 times trying to get the segments to look right :P

User avatar
ZZ Ardoz
Wiggling with Experience
Posts: 496
Joined: Wed Sep 02, 2015 4:12 pm
Location: Los Angeles

Post by ZZ Ardoz » Thu Nov 03, 2016 5:40 pm

igowen wrote:
ZZ Ardoz wrote: Kris is working on a solution to the noise issue. He believes it is the OLED drawing too much power.
this is good to hear, looking forward to a fix.
As far as I know, you can't break the chip by loading it multiple times.
i certainly hope this is the case, i probably flashed mine 20 times trying to get the segments to look right :P
If the segments are off, it's most likely the hardware making that happen

User avatar
igowen
Wiggling with Experience
Posts: 282
Joined: Wed Jul 01, 2015 10:07 pm
Location: pittsburgh

Post by igowen » Thu Nov 03, 2016 6:41 pm

ZZ Ardoz wrote: If the segments are off, it's most likely the hardware making that happen
nah, i meant the code i wrote to make the segments shaped like a real segmented display, the display works fine from a hardware perspective

User avatar
phidelity
Common Wiggler
Posts: 131
Joined: Mon Sep 01, 2014 1:12 pm

Post by phidelity » Fri Nov 04, 2016 11:43 am

Altitude909 wrote:Is it me or is the hole for the display way off? Using the Adafruit display and its not even close, the left there is a good 1mm gap between the opening and the display edge
I've had problems with the OLEDs I've gotten from adafruit as it seems they are manually positioned with a piece of not too sticky double sided tape. The hole on the panel is made to the spec of where the OLED is *supposed* to be.

User avatar
phidelity
Common Wiggler
Posts: 131
Joined: Mon Sep 01, 2014 1:12 pm

Post by phidelity » Fri Nov 04, 2016 11:46 am

ZZ Ardoz wrote:Anyone have an idea why the resistor/cap pads on the new run are just small circles instead of the squares of the first run? Kind of makes it a bit more fussy to solder.
This isn't normal. Im not sure what happened with my vendor but I used the same gerbers as the first time. I've stopped selling these in the meantime and if anybody got these feel free to write to brandy@magpiemodular.com and let her know I can send replacement boards when we get the new batch.

User avatar
phidelity
Common Wiggler
Posts: 131
Joined: Mon Sep 01, 2014 1:12 pm

Post by phidelity » Fri Nov 04, 2016 11:50 am

igowen wrote:Image

:sb: :goo: :sb: :goo: :sb:

this is by far the most challenging build i've attempted. i had to reflow/resolder some of the ICs multiple times to get rid of solder bridges.

here's the display code to make it look like a fake segmented display, if anyone's interested:
Looks awesome!

User avatar
phidelity
Common Wiggler
Posts: 131
Joined: Mon Sep 01, 2014 1:12 pm

Post by phidelity » Fri Nov 04, 2016 11:58 am

igowen wrote:Two issues i've encountered with both of my builds:
1) the amount the display is lit affects the output pitch (maybe due to power draw in the display?)
2) there's a lot of noise in meta mode

anyone know if there's any way to address these?
We are looking into this: I have the same issue and here are my observations
The noise is worse the more segments are lit up on the OLED. Its weird that it only seems to affect META mode. We changed c17 to 1u and it smooths out the noise but apparently at the expense of fm at upper audio rates.

We are redesigning the board right now to attempt to isolate this issue, add designators for parts, Add a JTAG header and beef up the power and ground rails.

User avatar
phidelity
Common Wiggler
Posts: 131
Joined: Mon Sep 01, 2014 1:12 pm

Post by phidelity » Fri Nov 04, 2016 12:03 pm

igowen wrote: i certainly hope this is the case, i probably flashed mine 20 times trying to get the segments to look right :P
One thing I've had issue with in the past is trying to flash via a USB hub. it would drop communication at random intervals. Hope this helps someone.

Post Reply

Return to “Music Tech DIY”