MAST Cell for Wildlife Research

Commit 7d753bf

Minor component & silkscreen changes.

What worked:

- SPI flash

What didn't work:

- Solar charging

- 1.8V buck converter

SPI flash orientation was fixed and the chip was immediately detected by my programmer. No problems there.

Adding a 43K resistor (top left, 0603, not soldered in image) to the MPPC pin didn't seem to do much. I'm probably going to figure out the solar charger in a separate PCB so I don't have to order 6-layer boards every time.

It turns out that I also had the wrong footprint for the 1.8V buck converter IC (top center, 8-ball BGA, TPS62748). The IC is 0.4mm pitch, and the footprint in the KiCAD default library was 0.5mm, I didn't notice.

For some reason, the same chip in the last batch was placed so perfectly in the center that it didn't raise a problem on both boards (I have 2 boards assembled each revision).

This time, none of them worked.

One was drawing >500mA idle and heating up like crazy, the other one just had nothing on the 1.8V rail.

Hotspot on the top is the 3.3V LDO, the one on the bottom is the 1.8V buck. (alignment’s pretty bad)

No big deal though, I just removed the 1.8V buck converter and soldered a wire to it. I'll feed 1.8V manually while testing.

More updates on the RF section coming soon.

Also, if anyone has any idea of how to make the solar charger IC (LTC3105) work properly, please leave a comment.

Commit 13a81ae

First prototype. Based on the reference design provided by Arribada. RF traces are calculated with JLC2313 stackup. PCB manufactured by JLCPCB.

Assembled board weighs less than 2g.

What worked:

- Microcontroller

- Power regulators / switch

What didn't work:

- Solar charging

- SPI flash

Apparantly, the "optional" MPPC function on the solar charger IC was not so optional after all.

I also messed up the symbol for the flash chip. The datasheet listed the BGA balls bottom-up, I assumed it was top-down.