DAC updates may cause transient spikes at specific code boundaries

sinara-hw / Stabilizer

2-channel microcontroller servo with EEM and Ethernet based on STM32 CPU

Other

22 stars 5 forks source link

DAC updates may cause transient spikes at specific code boundaries #83

Closed ryan-summers closed 2 years ago

ryan-summers commented 3 years ago

When the output DAC on stabilizer crosses the 10-bit code boundary, there is a ~20mV spike on the output waveform. For example, the following output was captured when the DAC code was repeatedly transitioned from 0x83FF to 0x8400 at a rate of 50KHz. The directionality of the spike appears to be a function of the transition direction (e.g. transitioning to a lower code will cause a positive spike): DS4_QuickPrint10

ryan-summers commented 3 years ago

Reference https://github.com/quartiq/stabilizer/issues/98 for further investigation and background information

ryan-summers commented 3 years ago

Note that this appears to be in-spec for the AD5542A and is specified in Figure 20: datasheet

jordens commented 3 years ago

Let's

Check whether fastino or stabilizer v1.0 are affected
Copy/harmonize the afe with fastino
Determine if this is dac batch dependent (w.r.t fastino and v1.0)
Block v1.2 on this.

jordens commented 3 years ago

This is assuming the datasheet axes are µs and not ns which is likely. The glitch is 1.1 nVs typ = 11 mV*0.1 µs consistent with that figure.

gkasprow commented 2 years ago

Do we do anything here?

jordens commented 2 years ago

There havent been any suggestions to resolve it or complaints about it. Seems inherent to the DAC architecture.