DIY Mutable Unsuccessful Builds

[cnicht]

Are the files you’re programming it with too large to fit in the flash memory?

I had a few issues with this.

[Azzoz]

No errors when flashing, i've tried few hex files including original yarn_bootloader and yarn_1.3, seems to finish the proccess successfully but still nothing on the screen when i reconnect it again :(

[rafyc]

Hi again, ist it normal on Rings moving the freq and brightness cv trimmer will change sound without cables plugged in ?

MI module tester can tell me if all working well ?

still have "position" knob not working and cv position is working (but dont know if it properly workin)

[Azzoz]

Hey guys ,

so I finally managed to make mbrane ( Yarns DIY ) works except only the 4th cv output is working , when i try to calibrate it, the outputs gives fixed voltage ( 1 = 2.88v , 2 = 2.00v . 3= 4.00v ) and only output 4 changing the voltages .

the gates on 4 outputs seems to work perfectly .

what could be the problem ? what should i look for here?

thanks

[rafyc]

How to solder ABMM2 quartz ? pads are exactly the size of the component.

[steviet]

How to solder ABMM2 quartz ? pads are exactly the size of the component.

I posted my method on my blog, hopefully you find this little excerpt helpful!

First, my method involves painting a little solder paste on to the back of the crystal with a toothpick. For this particular crystal, an 815-ABMM2-8-E2T, I put a little on every pad except the first pin/pad on the crystal. For this pad (circled below) I left it clean, and presoldered the PCB where it would sit.

Diagram showing pin 1 on crystal
It doesn't really matter which pad you leave clean, but leaving pin 1 clean just makes it easy to remember

Next, plop the crystal down, and heat the solder that is already on the PCB. Once it is sitting properly, let the solder cool, and then apply heat to every other pad while holding the crystal in place. This is where our little Bob Ross painting adventure from earlier pays off. The solder paste will reflow from the heat of your iron, and you will have one well soldered little crystal, with no reflow oven or SMD rework station!

[steviet]

Sure! Here you go. I tested the connection of the jacks - the outer ring of all jacks shorts to ground of the board - I assume thats how it should be?

The grounding on your jacks looks proper. But the two capacitors at the bottom near all the jacks look dangerously close to being shorted. I'd double check that those two aren't shorted and do some reflowing on the chips.

[rafyc]

How to solder ABMM2 quartz ? pads are exactly the size of the component.

I posted my method on my blog, hopefully you find this little excerpt helpful!

First, my method involves painting a little solder paste on to the back of the crystal with a toothpick. For this particular crystal, an 815-ABMM2-8-E2T, I put a little on every pad except the first pin/pad on the crystal. For this pad (circled below) I left it clean, and presoldered the PCB where it would sit.

Diagram showing pin 1 on crystal
It doesn't really matter which pad you leave clean, but leaving pin 1 clean just makes it easy to remember

Next, plop the crystal down, and heat the solder that is already on the PCB. Once it is sitting properly, let the solder cool, and then apply heat to every other pad while holding the crystal in place. This is where our little Bob Ross painting adventure from earlier pays off. The solder paste will reflow from the heat of your iron, and you will have one well soldered little crystal, with no reflow oven or SMD rework station!

thanks !!

[funkytransport]

Ive just built a plaits.. but having a couple of issues.
the LED's are switching between two shades of orange, instead of red/green. it also won't calibrate properly. I get flashing red after the process.

other than that it passes sound fine and everything seems to work.

anyone have any ideas? ive tried reflowing the LED driver and the caps next to it. and ive also re-soldered the LED's.

[cnicht]

Do all the potentiometers and inputs work correctly?

Are the op amps, which condition the input signals, producing the correct range of voltages?

[funkytransport]

Do all the potentiometers and inputs work correctly?

Are the op amps, which condition the input signals, producing the correct range of voltages?

I just gave all the pots, attenuverters and CV inputs a test and all seems to be responding as expected.

I also changed out the LED driver chip, and checked it was getting 3.3v on pin 1. grounds also seem OK.

I've reflowed pins 2,3 and 4 from MCU to LED driver..

I'm stumped.

[KittenVillage]

I'm stumped.

It may be time for some unsolicited DAC pics then.

[funkytransport]

https://www.dropbox.com/s/vo3g90kiovtxi ... 7.jpg?dl=0 picture of DAC here

[KittenVillage]

I do not know for sure on plaits, but I know that for the clouds firmware, you'll get flashing lights when the stm can't connect to the dac. Your soldering looks good on the dac... you may want to check everything between the two. It's best after a couple of rounds to take some pictures of everything and post it here so others can take a look.

[funkytransport]

i know what you mean, but i have flashing lights after calibration, not on start up.

[funkytransport]

Ive just finished the second one I am building and it has exactly the same issue.. could it be crap LEDs that just dont blend the colours well? I tested them before i put them in and green and red both worked separately? I got them on Aliexpress or Ebay i think..

[pjbulls]

I was thinking that at first, these dual colour LEDs are quite sensitive to the currents going into each colour to make them look sufficiently distinct. But, Plaits only ever lights one colour at a time during normal operation.

You did get common anode ones, right? Pretty much all other modules run common cathodes, I missed that myself before.

[funkytransport]

I was thinking that at first, these dual colour LEDs are quite sensitive to the currents going into each colour to make them look sufficiently distinct. But, Plaits only ever lights one colour at a time during normal operation.

You did get common anode ones, right? Pretty much all other modules run common cathodes, I missed that myself before.

Haha.. yeah , they wouldn't light up at all if they were the wrong type

I tried changing R3 to see if it made any difference but it never. I will order some later and see if it works!

[the_laconic]

Tried Branches as my first SMD build, which is a heck of an undertaking when you have tremors. I used the hotplate/sand method, and it seemed to work, mostly. However, when I plug it in, the LM1117 immediately gets scorching hot.

My understanding is that there is a short somewhere. I re-mounted a few bits that looked dodgy, but to no avail. I have been over the whole thing with a 10x loupe, but I don't see any solder bridges. I thought the (single) diode might be backwards, flipped it around, got no power then, and realized I had it right in the first place.

I have tested every possible point of contact for continuity with ground, and everything is as it should be. This is not a very complicated circuit, so what am I missing? I could understand if it just didn't work -- it's not impossible that I fried a chip with the hotplate, or when removing bridges -- but at least I should be able to get the power working, no?

Oh yeah, I tested the resistance between ground the the +5V out on the regulator, and it's around 800-1000 ohms. Please don't tell me the short is inside the ATMega...

[Mu Meson]

I have a weird problem with the Mutable Dev Environment. I have built this on Windows 10 using the latest versions of VirtualBox etc and cloning the Mutable dev environment using GIT. I then successfully used it to flash Grids from inside the Vagrant environment.

I returned to the same environment a few days later to flash Braids. Executing VAGRANT UP produced a timeout after the Private Key message. I waited and waited, numerous timeouts occurred and eventually the session shit down, with various hints, one of which was to look in the VirtualBox GUI for error messages. There were none - the VM was up and running without obvious errors. I messed and messed around obsessive and sweaty-palmed into the night. Nothing worked except VAGRANT DESTROY followed by VAGRANT UP, after which I can VAGRANT SSH once and once only before I get the timeout again.

Then I flashed Braids and it worked first time!!

I guess I should be happy; I have a workaround but I'd love to know if anybody knows how to fix this? I am an engineer with a good grasp of PCs but I am not a Unix hack so please be gentle and go easy on me ...

[Mu Meson]

I meant shut down sorry!

[mutronic]

I know it's sort of a pain but if it were me I'd just re-install the vagrant environment.

[Altitude909]

I meant shut down sorry!

Its currently broken on windows10 (i just tried it on a clean machine) and there is no reverting to Vbox 5 since windows now blocks that as a security hazard. The current Vagrant and virtualbox combination do not work out of the box. I managed to fix it somehow and havent touched it since but I cant tell you how i did it. I was some permission thing, if you boot the VM through virtualbox and watch where is craps out, that should point you in the right direction.

[mossa]

So I managed to get my elements build up and running around a week ago then 4 days after, I decided to update V1.2 which worked just fine for 2 days. The next day I turned on my rack PSU and elements was still operational, after being away for 10mins I came back to elements not working with non of the LEDs being illuminated.

I checked the voltage across the power pins at the power header, both 3.3Vregs and finally the voltages across the caps connected to the STM32’s 3.3V power pins. All seemed to be fine so my guess was that the STM32 is the issue, so I connected my st-link via jtag to find that the device can only be connected in hotplug mode when using the st-link utility. I believe this is because the read out protection is level 1 which limits the ability to read, write or erase the STM32.

However, I am unable to change the read out protection from level 1 back to level 0 as the program says “Could not set Option bytes! Please reset the target and retry”.
I am unsure how that option was even selected in the first place! So my question is how can I get rid of the read out protection so I can then erase and reprogram my elements build?

[forestcaver]

(I’ve managed to lock STM32s when messing around with eg the serial ports or doing silly things with them in hardware). You can use openocd try something like this:

init
reset halt
stm32f4x unlock 0
reset halt
exit