Pulsar #3

I have finished soldering two PSU in different configurations. On the left you have the 3 rail config that supports the use of the LTE/5G 3.8V modem, on the right the configuration with just 5V and 3.3V.

Soldering the boards was not too complicated except for the LTC4353IMF in the MSOP-16 package and the TPS6293x in the SOT-583 package, which might require a bit more patience. The MSOP-16 has pin very close together that might short and the SOT-583 has very tiny pins that are not exposed much from underneath the package, but it's definitely doable to solder everything with just a normal soldering iron.

Testing went well. The board works as expected.

The ideal diode that allows to switch between two power sources works perfectly and quickly. I tested it with a raspberry pi as the load on the 5V rail and it switched quickly enough for the raspberry to not reboot. I will redo the test with the STM32H733 but I expect the same result. This would confirm the ability to hot-swap the power sources during use.

When powering the Raspberry Pi I saw a ripple voltage, peak-to-peak, of less than 20mV on the 5V rail that, considering the extremely low switching frequency (460kHz), is very low and should guarantee very low noise levels in other components down the power path.

I'm planning a stress test as soon as I can simulate a 3A load for each rail. This is important for thermal testing and more data regarding the voltage stability at high loads.