The Power Distribution Board (PDB) is an SRAD power management solution that combines reliable power delivery with wireless arming capabilities. Each PDB is a complete battery and multiple voltage output power system for a flight computer and pyro. Additionally, its pyro power output is toggled by a web interface hosted on its WiFi network, allowing remote arming of the rocket’s deployment system. Each PDB is also a battery charging and management solution. The ease of charging combined with state-of-charge information available on the PDB’s webpage ensures that batteries will be appropriately charged before flight.

It is recommended that 2 PDBs be used for all flights, one for the primary, and one for the backup. Also, in case of problems with a PDB, it is recommended that at least one backup be available at the launch for replacement. Each PDB is meant to be always connected to its two batteries (battery replacement requires replacement of the battery management ICs). To facilitate this, the PDB will be housed with its batteries in a housing. This entire assembly will then be loaded into whatever AV bay contains it. This modular approach also makes replacement and charging easier.

Prototype Notes

Assembly of the first board was started by stenciling and placing the bottom components and reflowing in an oven. Next, we stenciled the front, and the components were soldered manually with hot air. The ESP32 was not heated enough, so it was removed and we tried resoldering. However, without being able to lay the paste using the stencil, we were essentially eyeballing the amount of paste. The result was that we had to reattempt soldering many times, never getting all of the pins to solder properly. In conclusion, the ESP32 must be soldered using the stencil and an extra-long heating cycle with the hot air must be used to ensure that all of the paste flows.

When tested WiFi did not function on the prototype. This is thought to be either due to accidental shifting of the RF shield on the module (and thus components underneath) during soldering or from initial testing without the WiFi antenna attached causing signal reflections. The conclusion is that care should be taken to only move the base of the module when heated and to always have the antenna connected during the operation of the board.

The batteries were initially not charging, and several code errors were found that caused the chips not to be configured properly. We must disable write protection before writing to NVS registers.

The arming system works about as expected, and it takes about 7 seconds for the pyro output to turn on after the board is armed and the alarm sounds. However, the same delay circuit causes the output to take a while to turn off as well, at least 10 seconds. This means that even when the alarm turns off, we must wait at least 20 seconds before considering the system disarmed.

Update:

After receiving the new ESP32s, the prototype’s chip was replaced. This was done by using the stencil to apply paste directly onto one of the ESP32s and then placing the chip onto the board and reflowing normally. This method, although difficult and time-consuming, finally allowed the prototype board to be fully functioning.