Today I finally received the PCBs (DHL took them for a trip to the other side of town for a few days). I knew it would be tiny but I was really impressed with how small it was. The text indicating the SOICbite pinout was even unreadable.

In fact, I had to test if it was correct to call it "thumbmouse". It seems it is.

While I was still waiting for the batch of passives I ordered, since I had some resistors and capacitors of acceptable values I decided to go ahead and solder this. I still don't have the BLE module (it is sleeping in some post office storage building for some days now), I figured it would be best to first try and test this via an external MCU, since I would not want to solder the BLE module to a useless board since it was a bit expensive.

I decided to use solder paste for this, since it would be good practice and also the FPC connector had a really small pitch and not so accessible pins. The solder paste I bough seems to be a bit dry (it may be since it is a bit cold here and it may require higher ambient temperature, or it is also possible it was not stored correctly). It was quite tough to apply the paste since it would not stick to the pads, so this was a bit messy. I placed all components and used hot air gun to melt the paste. This worked more or less OK, although I melted the plastic base of the microscope and the PCB got stuck. When I tried to take the PCB off the base I messed up and moved a lot of components. So, I had to manually fix everything with the iron and fix a few shorts. In the end it seems to work: 2v8 and 1v8 voltages were output correctly and the level-shifting worked reasonably (which about 300/400mv drop.

I decided to use a STM32F407 discovery board and an Arduino sketch to test. I had some difficulties at first which I thought was due to the low voltage of my USB port (it outputs 4.5v, which for the 3v regulator ends up being around 2.9v, not enough for the 2.8v regulator. I ended using the 3.3v output of the USB-serial converter which was  correct. However, it was still not working. In the end it was what I suspected: the Arduino STM32duino libraries chose a very wierd combination of SDA/SCL pins and I had to go into the headers to find this (:facepalm:).

With the communication working, I used the Arduino sketch I had written a while ago for the protoboard version and I added Mouse emulation support. Finally, I had thumbmouse working as a mouse for my PC.

So now, the next step will be to solder the BLE module when it arrives and try to have this working over Bluetooth. It will not be fully wireless actually unless I use a coin-cell, which means I will probably power it from the SOICbite connection. I will test the power draw of the sensor to see if it is not too power hungry for the coin-cell. Anyways, next version based on 8520 trackpad will definitely use a rechargeable li-po battery.