driving amplifier VOCM

Hi jn_elbit:

The AD7768 specs a differential input range of +/-Vref, and an absolute input range of AVSS to AVDD1.  So I would be inclined to set the LTC1992 Vocm at 2.5V so that even at full +/-4.096 output swing, it is not close to clipping.  If you were the set Vocm at 2.048V, then the lower output would be close to clipping at full +/-4.096V output swing.

Regarding the R's and C's, these are best determined by experimentation.  You typically want the sampling glitch to be settled within the acquisition time window.  A large output capacitor can swamp the glitch charge, reducing its amplitude.   But of course this means the tau will be long (RC product) or with too small and R, the load will appear capacitive to the amplifier which could cause ringing.  A small capacitor means that the glitch will get presented to the amplifier output which will then have to settle it adequately within the acquisition window.

The LTC1992 data sheet shows frequency response curves driving capacitors directly up to 10nF.  Some series R will reduce the peaking drastically.  If your signals are 2kHz and you wanted, say, 10kHz of bandwidth, the preliminary R would be 1/(2pi*10kHz*10nF) = 1.6k Ohms. 

hi Glen

appreciate your insight and quick response. 

if i choose to stay with VOCM of 2.048V will i loose dynamic Range or bits ? 

thanks much

Eli