Many factors could affect the self-oscillation of op-amp.
For example, Looking at Non-inverting Small-Signal Frequency Response figure in the top left of p.4 of the OPA847 datasheet, there is already about 4dB of peaking in the frequency response with a gain of 12V/V. Frequency response peaking translates to the possibility of ringing or oscillation in the time domain and to inadequate phase margin in terms of stability analysis. That is to say, when there is no input signal or the input signal is weak, a self-oscillation could exist.
However, the third figure on p. 4 of the datasheet, Non-inverting Large-Signal Frequency response shows that the amplifier has a decreasing bandwidth, which will reduce the peaking, as the output voltage swing gets larger. It means this oscillation can be eliminate by using larger input signal.
List possible facts here:
- Power supply, parallel ~pF capacitance
- Input signal level
- GBW and op-amp
- Circuit fabrication, like breadboard (high frequency)
- Not too much gain
Check it one by one.
Update 12 July:
Parallel ~pF capacitor with the feedback resistor
The capacitor value depends on your working frequency, simply to say, the bigger capa, the smaller bandwidth.
serie ~ohm resistor at the output of op-amp
like this :
Note: The stronger inverse feedback, the easier self-oscillation: