I am trying to get a design I inherited to work, and am seeing a large amount of variation noise in LTC2386-18 ADC readings with a constant level input the the ADA4500-2 used for ADC driver. Variation is about 700 ADC counts p-p, or about 23mV p-p. With ADC input shorted, noise decreases in to microvolt range as expected. I suspect the variation is caused by blowback from the ADC (it is running at 10Msps) into the ADC driver and slow settling time of the driver? The input to the ADC has 180pF to ground on each input, and 24.9 ohms series resistance to the ADC driver output.
Looking at ADC input I am seeing about an 80mV p-p variation at 10MHz. See scope photo.
Please let me know what you think. Can send more info if needed.
(1) Values read from ADC with a constant input to ADC driver.
(2) Schematic snippet.
(3) Scope photo of the plus input to the ADC. 600MHz scope with 500MHz probe.