This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

OPA192: The output impedance is high, so is it suitable to as a unity-gain buffer?

Part Number: OPA192
Other Parts Discussed in Thread: OPA180,

As far as I know, the ideal open-loop output impedance of an op amp should be as low as possible.

But I see the open-loop output impedance of opa192 is really high especially at low frequency. Nevertheless, the OPA180 has a low output impedance. Let us call the output impedance "Zo".

Question 1: Can you please tell me what type op amps have lower output impedance and what type have high output impedance? And what advantages can be brought if the Zo is high? 

Question 2: In what situations is high Zo op amp more appropriate? And same question to low Zo op amp. 

Question 3: I know the Zo can be reduced significantly by the loop gain of the op amp. What will be affect by these two parameter respectively? the open-loop Zo and closed-loop Zo?

Question 4: As for stability of amp, the isolated resistor is related to the open-loop Zo right? I found some equations like Riso = Ro * Rg/Rf. If the Ro is high like 1k , the Riso will be large. But the value of Riso is small in most examples, even in the application notes of the amp with high Zo.

Looking forward to your reply.

Thanks.  

  • First, you need to differentiate between the open-loop output resistance, Ro, and close-loop output resistance, Rout.  What you really care about from the stability point of view is Rout at frequency where curves of AOL and 1/beta (close-loop gain) intersect - this defines an effective bandwidth of the configuration.

     

    You cannot tell from the Zo graphs below what is the Zo below 1Hz for OPA180 (it does not show) but regardless, Zo, gets devided by loop-gain so for example dc OPA192 Zo of 3kohm for buffer configuration becomes only 3mohm (~3,000ohm/1e6).  Having said that for stability only value of Zout at the effective bandwidth is important and NOT its dc value. For a buffer configuration, Zo's are shown below in the red circles.

    Below please see my answers to your other questions:

    Question 1: Can you please tell me what type op amps have lower output impedance and what type have high output impedance? And what advantages can be brought if the Zo is high? Zo is inversely proportional to current in the output stage, thus the higher the IQ, the higher bandwidth and lower Zo.

    Question 2: In what situations is high Zo op amp more appropriate? And same question to low Zo op amp. Ideally, you would want Zo to be zero so it does not form a pole (and cause a phase shift) with the capacitive load - Zo of zero would allow an op amp to drive any capacitve load.

    Question 3: I know the Zo can be reduced significantly by the loop gain of the op amp. What will be affect by these two parameter respectively? The open-loop Zo and closed-loop Zo?  An open-loop output impedance, Zo, is defined by the IC designer so you as a user cannot change it.  Only a close-loop output impedance, Zout, is a function of the circuit configuration or loop gain: Zout=Zo/(1+AOL*beta), which you have control over.

    Question 4: As for stability of amp, the isolated resistor is related to the open-loop Zo right? I found some equations like Riso = Ro * Rg/Rf. If the Ro is high like 1k , the Riso will be large. But the value of Riso is small in most examples, even in the application notes of the amp with high Zo.  Adding Riso equal to Ro creates a zero (Riso||CL) and immediately cancels the pole created by (Ro+Riso)||CL and absolutely guarantees the stability of the system - see below. 


    Riso of up to 10x smaller than Ro will cause a dip in the phase margin caused by the frequency separation between the pole and cancelling zero BUT it may be sufficient to guarantee stability of the system.  For more details, please see the link below:

    https://training.ti.com/ti-precision-labs-op-amps-stability-1

  • Hi Marek,

    Thanks for your explanation. It is really helpful. I will check the video later.

    I just want to confirm with you that this sentence "for stability only value of Zout at the effective bandwidth is important and NOT its dc value". should be "for stability only value of Zo at the effective bandwidth is important and NOT its dc value. ", right? I should care about the open-loop resistance not close-loop resistance at the effective bandwidth, right?

    And I also have some issues about stability of op amp. As for as I know, there are two method for stability. One is in-loop compensation and the other is out-of-loop compensation. What we talked about above is out-of-loop compensation.

    Q1: Can you please help me with the calculation of in-loop compensation? 

    Q2: What use cases are suitable for these two method respectively? I think in-loop compensation is suitable for reference buffer as this method do not have error. And out-of-loop is suitable for shield cable drive circuits. I don't know if it is correct.

    Q3: For in-loop compensation, Is the low value cap between output pin and negative pin of amp necessary? Because I saw some examples do not have that cap, only a Riso in the feedback loop.

    Thanks,

    Daniel  

  • At the effective bandwidth frequency AOL intersects 1/beta curve so the loop-gain, AOL*beta=1, which means Zout=Zo/2.

    I'm not sure what you mean by in-loop compensation vs out-of-loop compensation - first, please read the TI Precision Lab material and then explain what exactly you refer to.  A low value cap across the feedback resistor, Rf, is needed to cancel a zero created by a large input resistor, R1 and C1, ONLY if the zero occurs within the effective bandwidth of the circuit configuration - see below.  The Cf cap needs to be chosen so that R1*C1=Rf*Cf.

  • Hi Marek,

    I'm sorry I didn't make myself clear. 

    The out-of-loop compensation is identical to the method 1 in the TI Precision Lab video 5, known as the Riso method.

    The in-loop compensation is identical to the method 2 in the TI Precision Lab video 6, known as the "Riso + Dual Feedback" method.

    Below are the questions.

    Q1: Can you please help me with the calculation of "Riso + Dual Feedback"  compensation? 

    Q2: What use cases are suitable for these two method respectively? I think "Riso + Dual Feedback" compensation is suitable for reference buffer as this method do not have error. And Riso method is suitable for shield cable drive circuits. I don't know if it is correct.

    Q3: For "Riso + Dual Feedback" compensation, is the low value cap between output pin and negative pin of amp necessary? Because I saw some examples do not have that cap, only a Riso in the feedback loop.

    Thanks,

    Daniel  

  • Hi Daniel,

    have you viewed all TI's stability training videos?

    Kai

  • Hi Kai,

    Sorry, not yet. I was tied up on other things. If the answer is covered in the video. I will check it out later.

    Thanks,

    Kai