Hi Rick,
Thanks for getting back to me so quickly.
I checked the DIR pin on the scope, and saw no significant noise. It was switching only when the motor changed directions, as it should.
The nFAULT pin was high. The fault register showed only that a stall was detected, despite internal stall detection being disabled. Resetting the register did not clear the stall flag.
The sense resistors are 0.04 Ohm. Those waveforms were taken with ISGAIN set to 40x and torque set to 255.
From section 7.3.4 in the datasheet, I_FS = (2.75V * Torque) / (256 * ISGAIN * R_I_SNS). So I_FS should be 1.71A.
The current with no microstepping is 1.08A, which is close enough to the theoretical value of 1.71 * 71% = 1.21A.
Here's 1/2 microstepping with both phases and the STEP pin included:
The current seems to be 0%, 71%, and 50%, like you said.
Increased microstepping seems to follow the same pattern where it peaks at 71%, drops to 50%, then peaks at 71% again.
Register settings for T_Blank, T_Off, T_Decay, and Decay Mode have varied all over the place during testing. As long as the waveform is anywhere close to a sine, it has double peaks. Adaptive blanking has no effect either.
This is 1/64th microstepping again:
I haven't been able to test with another board. This is the only one I have available at the moment.