Waiting to trim the LEDS to the panel…
There’s an excellent article from After Later Audio on the differences between the various versions of MI’s Clouds here.
I couldn’t get the firmware to build (issue with the ARM toolchain), so I just flashed it with the Supercell hex file from Patrick Dowling’s Github.
Flashing via ST-Link
As that worked fine, I wasn’t expecting any problems, but on finishing the build and firing it up, it was completely dead.
First thing I checked after the power rails and pins was the 8mhz clock signal on the STM on a scope – nothing.
The crystals for the STM and the audio codec are a smaller form factor than the ones MI use, and Mouser were out of stock of the 8mhz one. I thought I had some of these left over from some other Jak Plugg builds – turns out they weren’t, and were completely wrong, having four pins rather than two – two of which are shorted. So no clock.
I had a few of the larger “stock” Clouds crystals, so to test, I pulled the plastic base off and bent the legs to jerry rig soldering it to the pads…and everything works…I just have to find another supplier for this or wait a couple of months for Mouser to have it in stock.
I guess the crystal isn’t needed to flash STMs.
Two things I’d suggest doing if you’re building this: one, get the free copy of DipTrace and have the board file from the Jak Plugg github open – very useful for component placement and orientation, troubleshooting, etc. On drag soldering the STM, I thought there was a stubborn tiny solder bridge between two pins – turns out they’re bridged on the board.
The second thing is: make sure you’ve got the two screws to fit the panel to the slide pots, there’s nothing else holding the panel to the top of the PCB, and you can’t solder the LEDS in place without that. Datasheet for the pots says M2.
This is a real “instant gratification” module and the extra controls really add something. I’m tempted to build another one.