I think I do not properly understand how to predict the time that it takes from the moment a conversion is initiated to the moment when the first sample is ready, and I would like some clarification. A little background first.
I am using the AD7124-4 in full power mode with the sync4 filter. I am also using single conversion mode with three differential channels enabled.
What I expected to see was the time it took from when a single conversion was initiated to when the first channel data was ready would be:
filter settling time + channel conversion time.
What I am actually seeing is that the time from the initiation of the conversion to the first channel being ready is:
~350uS + Channel conversion time.
I have tried a handful of different FS values and always get similar results. It seems that the filter settling time is fixed at ~350uS, but I know that this cannot be true since the filters are clearly working.
To get my results I am measuring the time from when the last SPI bit is clocked to the AD7124 (starting a conversion) to the time when DOUT goes low for the first time. I then subract the channel conversion time from this total time to arrive at ~350uS.
Can you explain to me what I am seeing? Is the filter settling time only applicable in single conversion mode if you switch between filters? If so, then what is causing that ~350uS delay?