Lab · Active Projects
Halfstep Rev B Bring-up
Active
Encoder reads clean at rest and garbage above 900 RPM.
Motion Control / STM32G4 / CAN-FDBoardHalfstep Rev B
EncoderAS5047P
DriverTMC5160
StartedAug 2026
Progress48%
The question
Rev A proved the control loop. Rev B has to prove it survives at speed on a production-shaped board.
Where it stands
Position reads to spec while the rotor is stationary and degrades sharply past roughly 900 RPM. The pattern points at SPI timing colliding with the driver’s chopper frequency rather than at the encoder — the corruption is periodic, not random.
03Bench log
22 Aug 2026
Rev B boards received. Power rails and CAN-FD verified.
02 Sep 2026
Encoder corruption reproduced. Onset is consistently 880–920 RPM.
14 Sep 2026
Rev A kept alive on the bench as a known-good reference.
04Next step
Next
Scope SPI clock against the chopper output. If they beat, move SPI to a frequency with no common factor.