The blue line is from the Phase A lowside source pin to the .01 ohm shunt resistor compared to the gate signal in purple. So this is before amplifying it at the output of the SO1 pin. Is wanting to get down to less than an amp of control unrealistic for this setup? Amplfying the noise just gives us more noise that leaves a great deal of averaging for the ADC and will reduce accuracy.
Here the output of the SO1 pin after 10x gain. How can the ADC sample this on the lows with any accuracy? ( With PWM percentage higher by about 40% than previous )
And here it is zoomed out on the time scale to show the sine wave effect
