I remember reading somewhere that the longer the S/H cap has to charge the better effective resolution the A2D is capable of achieving up to its resolution limit.
For example, I have a DSP (TMS320F2812) with 12 bit A2D that can sample at
12.5MS/s. It charges the S/H cap for 80nS.I only need to sample every 50uS so if I set the S/H to charge the cap for
10uS, would you suspect that the LSB dither to be significantly reduced?In this application, there is a 5" length of ribbon cable beween the high speed opamp buffers (AD8030) and the A2D pins that cant be avoided and I'm sure is part of the problem. At the 80nS acquisition time, there is about
+/-200 counts (+/-50mV) of noise on a 3V input which I think is caused by the impedance if the wire.If I increased the S/H time window would that reduce the noise?