Which Op-Amps can be used as integrators?

[ballimo]

Hello All,

Some Op-Amp data sheet mention a "minimum gain compensation" value per example +2V/V, +10V/V..., other Op-Amp data sheet don't mention any value.

if I want to use an Op-Amp as an "integrator"; for a "integrator" there is no "gain" (in other word the amplification is not a gain it is a integration); My question is: can I use a op-Amp with a minimum gain compensation of +10V/V as an integrator? or it should be a unity-gain stable? or is there a specific type of Op-Amp to be use as integrator?

Many Thanks,

Ballimo

 

[WimRFP]

OP-Amp unity gain

Hello,

When an opamp is not unity gain stable, you cannot use it for a standard integrator (series resistor and feedback capacitor) as it will oscillate.

So an opamp that is stable for gain > 5, cannot be used. Adding some additional components may enable use of non-unity gain stable opamps, but this increases complexity.

Note that you need some other form of feedback to avoid that the output of the opamp runs to positive or negative supply voltge (because of offset issues).

 

[ballimo]

Re: OP-Amp unity gain

Thanks,

I have used a resistor in parallel with the feedback capacitor, is this what you mean with "other form of feedback to avoid that the output of the op-amp runs to positive or negative supply voltage"? Please let me know what do you mean by by "other form of feedback".

Ballimo

 

[WimRFP]

OP-Amp unity gain

Hello,

Yes, this is one of the "other form of feedback". It could also be that your integrator is part of a larger system where DC feedback is provided. In that case a feedback resistor is not required.

 

[LvW]

Re: OP-Amp unity gain

Hello,

Yes, this is one of the "other form of feedback". It could also be that your integrator is part of a larger system where DC feedback is provided. In that case a feedback resistor is not required.

I think, here two different effects are mixed:
(1) A parallel resistor (or overall dc feedback) is needed because of dc stability, which means: a stable dc operating point.
(2) Unity gain stability concerns rf stability which means: stabilty against parasitic oscillations. Here, a parallel resistor cannot help at all.
(3) Unity gain stability is needed also for integrating stages because there is always a frequency (in the noise floor) which otherwise would bring the circuit close to the stability limit. As for 100% resistive feedback, also for the integrator the loop gain is unity at the opamps "unity gain frequency" (practically, 100% feedback).

 

[ballimo]

Re: OP-Amp unity gain

Thank you LvW for the information,

I have 2 questions, if you have some time:
1) OK
2) "stability against parasitic oscillations. Here, a parallel resistor cannot help at all"; if a parallel resistor cannot help, how can I modify the circuit to improve stability?
3)I have not understand it well, but do you mean that for the integrator circuit we have to use a unity gain Op-Amp?

Ballimo

 

[LvW]

Re: OP-Amp unity gain

2) "stability against parasitic oscillations. Here, a parallel resistor cannot help at all"; if a parallel resistor cannot help, how can I modify the circuit to improve stability?
3)I have not understand it well, but do you mean that for the integrator circuit we have to use a unity gain Op-Amp?

2) If you cannot use a full compensated opamp (unity gain stable) you can stabilize the circuit (inverting MILLER integrator) using an additional resistor in series to the feedback capacitor. This modification prevents the feedback factor going as high as 100% for large frequencies. However, I would not recommend this method as it sacrifies the integrating function too much.
3) Yes - but not a "unity gain opamp" but a "unity gain stabilized opamp".
An opamp with a minimum recommended gain of 3 or 5 may not oscillate or become unstable, but its stability margin is probably too small (ringing, oscillatory effects).