I wanted a low drift ADC driver in a precision thermometer circuit so I used an AD8629 (please refer to attached pdf). The op-amp +ve input was fine (i.e. a dc level proportional to the thermistor resistance) but the output was oscillating at several hundred mV, at about 800kHz. The same thing happened using an LTC2052, but problem was fixed when I used an LMP7702. What was I doing wrong?
I would suspect the LMP will drive a little more C. Try changing the 10 nF to1 nF and see what happens.
Or with the 10 nF, change the 20 ohm to 100 ohm. What frequency is the oscillation?
I was using a 1nF cap originally, but the AD8629 still oscillated. The frequency is about 800kHz. Subsequently, after I got the circuit working with the LMP7702, I increased the cap to 10nF. I'm averaging 4096 successive ADC samples at a rate of 100k samples/s, and I needed the increased capacitance to keep the ADC input stable.
I've found an article in EDN, October 16, 2008 that describes this kind of op amp stability problem - I guess I need to learn a bit more op amp theory.
By the way, there's an error on the schematic - pin 5 of the op amp is connnected to a +2.5V reference, not +5V(A) as shown.