RCA - Arduino Tracking Antenna

Update the title following the format: Jan 01, 2023 - RCA Name

Incident overview |  Executive summary |  Incident timeline |  Postmortem report |  Recommendations for future

Incident overview

 Executive summary

During Mar 29th, @Yuchen Lin is following the pre-flight check list to test the fully functionality of the tracking antenna. When the battery is plugged in, the buck convert caught on fire and result in the damage of a buck converter, GPS, Arduino and possibly a IMU.

 Incident timeline

• check for all necessary components

• bring all components outside of the bay

• plug in battery with tracking antenna plugged in

• observe the buck caught on fire

• immediately unplug the battery

• Test the new buck converter without tracking antenna plugged in

• plug in to test the tracking antenna but find out the arduino will kill the computer

Incident report:

The main reason of the incident is the GPS not being properly powered.

Open

A picture of the GPS

This GPS model only supports 3.3v for power. The prototyped version of the tracking antenna used a LDO to convert 5V from arduino to 3.3v to power the GPS module. When the incident happened, there are two wires connected to the two 3.3v pin on the GPS module. One wire is 5v from arduino to LDO and then output 3.3v to the GPS module. Another wire is directly connecting the the system power(5v) to the GPS. The reason causing the buck converter (12v to 5v) on fire is because the Vin and Vout of the LDO are being connected to the together through the two 3.3v pin and shorted on the GPS module and cause the LDO being fried and eventually leads to short between VIN and GND on tracking antenna prototyping board input screw terminal.

This also result in the Arduino on the prototyping board will draw more than the spec of the USB port on a normal laptop and trigger the over current protection on the laptop. (warning: the LED indicator might make an illustion that the Arduino is still working)

recap

Issue 1:

• after the buck convert caught on fire should immediately bring the whole system back to the bay and test under a monitored enviornment instead of with a battery. Since the battery does not have over-current protection. This might cause further issue to the rest of the system.

Issue 2:

• @Yuchen Lin assumes the system is working as expected. (Note: the system is fully powered tested a week before the incident happened)

• @Yuchen Lin propose a solution of flying 5v to 3.3v LDO solution instead of planning it out carefully with the EE team, which this may leads to people not knowing the GPS module powering detail.

 Postmortem report

 Recommendations for future