I'm using the AD9625-2.5-EBZ as my reference design which is set up for AC coupling and a differential input (single ended transformed by a balun).
My end application involves capturing pulses that are not periodic and occur relatively infrequently. I also need to take advantage of the full voltage range of the ADC as my input signal can swing from 0V to +1V. Pulse are in the tens of nanoseconds in length with rise times in the single nanoseconds.
Given the above criteria, it would seem that DC coupling a single ended input is the best path. The AD9625 data sheet indicates that this can be done by tying the negative differential input to Vcm. Is this a reasonable path to pursue? Is there an app note that discusses DC coupling to this ADC in more detail?
Would it also be beneficial to put an op-amp between the input and the ADC?
Yes you can always match the CM voltage on the positive input of the ADC to DC couple a single ended signal. The negative input of the ADC needs a 0.1uF to ground. You might also need termination on the positive input to match the plus production stage in front of the ADC. Like you mentioned, performance would be degraded. We don't have this set up to quantify the degradation though.
The amp would allow gain in the front end to amplify pulse to maximize dynamic range if pulse is not full scale of the ADC.
ADL5566/ADL5565 is the released alternative to the ADL5567, but for DC coupling, the ADA4937/8/9 works best.