All, Latest Copter, MP, H743 firmware, dead-cat 450 quad.
I observed what I initially thought was an issue with the H743 voltage/current sensor: It would not consistently report the calibrated voltage on the HUD or via telemetry. The display was often 0.5V low, and would then slowly (over 5-8 seconds) return to within 0.03V of the calibrated voltage. This ‘drop-then-return correctly’ cycle repeats every 15-20 seconds.
The cycling and voltage drop are a result of the IMU heater requesting heat.
To prove this, I set BRD_HEAT_TARG to -1, which kills the heating element. After rebooting, the HUD reported voltage and actual voltage correctly remain within 0.03V of each other. I determined that the voltage drop is ocurring within the FC itself, not caused by a voltage drop in the external wiring due to increased current demand.
In my case, the false low voltage resulted in an unanticipated premature RTL Failsafe event, and a minor crash.
Test setup consists of a variable voltage lab power supply and several DVMs of known accuracy.
Representative test result:
- Calibrated voltage at 11.90V HUD says 11.93V, system wide current is 340 mA
- When IMU heater fires, HUD says 11.41V, current is 410 mA
- After ~20 seconds, displayed voltage slowly increases to 11.92V, current 340 mA
- Above cycle repeats every 20-30 seconds
So an internal-to-FC ~70mA current bump is causing the voltage sag - I’m not sure of a work around yet.
No .bin file, this was all bench work.
I’m thinking that the power bus (or internal traces/wires) in the H743 might need a revisit…or power the heating element from a different source?
