I'm trying to understand the reason for setting the HPF filter cap from OFST to CNTR rather than to ground. If Cntr is being driven by a low impedance Vref it shouldn't matter. Since they are both DC.
But your data sheet gives other options, like leaving CNTR floating and lightly bypassed. (eg 0.1uF in the eval board schematic, pg 29 of Rev H datasheet)
Now in the Offset Compensation section (pg 20, near eq 12) You suggest a 3.3uF Chp to get a 100 Hz corner. In that case the filter cap is 33x the bypass cap value, so it's clearly going to alter the performance.
Do you intend to feed a signal from OFST back into the CNTR pin in these cases? If so, you need to take the bias resistors and bypass caps on CNTR into account. Also, how would this work if CNTR was drive by a DC reference?
If you do NOT intend for OFST to feedback into and affect the CNTR voltage, why bypass to there instead of to ground? Also, how can your 3.3uF 100Hz cap possibly work?
In my application I'm targeting Vocm of ~ 1V and it seems I can hit that with a 2.67k R from CNTR to GND, with a suitable bypass cap in parallel to it. (0.1uF? 1uF?)
But I also want a HPF corner of 145 Hz, so eq 12 is telling me to use ~ 2.2uF But Eq 2 is only taking into account the internal 480 ohm Rint on OFST. In this case, it seems that OFST would be modulating the CNTR voltage and that the effective impedance of CNTR to GND would dramatically alter the corner frequency.
Am I thinking about this wrong, or did you just mess up the suggestions for low freq HPF in the data sheet?
