ADS1262EVM-PDK: ADS1262EVM-PDK

Hi Julien,

I can help answer these questions...

1. Conversion Start Delay is a programmable delay that you can set to force the ADS1262 to wait a certain amount of time before starting a new conversion (See 9.4.3 Programmable Time Delay on page 64 of the ADS1262 datasheet). Typically this is used when multiplexing between channels to allow additional time between conversions for external signals to settle (for example, if there is an external RC settling time).
   
   
   
2. Calibration allow you to compensate for systematic offsets and gain errors in the measurement to improve the accuracy of your data acquisition system. The ADC contributes a small amount of offset and gain error that can be greatly reduce by calibration. However, external circuitry such buffers and RC filter may also contribute to a system's total offset and gain error. Calibrating an ADC is a process in which you measure the offset when a 0V input signal is applied (typically by shorting the inputs to the system) and the gain error of a near full-scale signal (after compensating off the offset error).
   
   When you provide calibration coefficients to the ADS1262, it will program the digital logic to calculate a scaled output result which compensates for these errors, so that you don't have to use up any of your processing power/bandwidth to perform these operations.
   
   Calibration is optional. If the accuracy of the ADC is sufficient without performing calibration then you may not need to calibrate at all.
   
   
   
3. Reducing the digital filter's output data rate changes the digital filter's noise bandwidth and is the most effective way of reducing measurement noise. However, you may have a bandwidth or data throughput requirement which prevents you from going below a certain data rate.
   
   Depending on the type and source of the noise, sometimes noise performance can be improved by analog filtering or by improving the quality of the reference source (using a lower-noise reference source).
   
   If you were operating at 7.2 kSPS or lower, then you would have the option of selecting one of the SINC1-4 filters, which trade off time-domain latency with noise bandwidth.
   
   You can sometimes also collect data at the required data rate, but then post-process the data (perform additional averaging, for example) so that you get faster step response detection and a moving average that contains a lower noise figure for a stable (slow changing) signal.
   
   
   1. Enabling CHOP mode on the ADS1262 is a type of averaging that would delay your first conversion result, but would provide you with a moving average of the last two conversion results. It updates after each consecutive conversion, so you'd still get continuous conversions at about the same output data rate if you using either the SINC1 or FIR filter. For all of the other filters types the output data rate will be roughly the nominal data rate divided by the filter order.