[bug] Timing controls violate deadbands as mentioned in the AFE4400 datasheet. Result in noisy and therefore incorrect readings.

[yashmulgaonkar]

As mentioned, the timing scheme violates the 2x 4HMz clock cycle deadbands.
Snip from AFE4400 Datasheet:

Example violations:

protocentral-afe4490-arduino/src/protocentral_afe44xx.cpp

Line 145 in b3471f6

afe44xxWrite(ADCRSTCNT0, 0X000000);

protocentral-afe4490-arduino/src/protocentral_afe44xx.cpp

Line 146 in b3471f6

afe44xxWrite(ADCRSTENDCT0, 0X000000);

These need to be at least 2 clock cycles apart.

Resulting data:

Image from Protocentral blog shows cleaner data, so likely that the timing scheme needs updating on GitHub.

[protocentralashwin]

I am not able to reproduce the noise that you have mentioned with the current code (both on AFE4490 and AFE4400)., since we're sampling at a very low rate only. Usually we have seen this noise when either a) the probe is damaged and the current from the photodiode keeps fluctuating or b) when there is light leakage from ambient (mostly from flourescent lamps).

But I will still check the timing issue also in the meantime.