OP amp stability gain < 1

[Jester]

I don't have a good understanding of stability with gain near or below 1

Will this circuit be stable?



The op amp is rail to rail input and output and is unity-gain stable.

https://www.ti.com/lit/ds/symlink/opa2348.pdf

If I changed R9 to 150k would the circuit suddenly become unstable?

Thank you for your help

 

[juan_zo]

Dear Jester,
If you want to make an 100% stable amplifier, you must set Opamp's gain to 1, how you can do it? It is easy, only connect the output to inv input directly.
About you say of R9 to 150K, yes you can do it without problem, R9 and R11 are a voltage divisor, if you increase r9's value, the level voltage in the non inverting input will be lower, only this.
About "Unconditionally Stable Amplifiers" they are only a study item than we all studied, only this. You can design an amplifier with upper than 1 gain and it could be stable too, however, if you set its gain to 1, it will be stable always, in the theory
... You know, electronics could be rebel.

I hope I can helped you with this,

Have a nice day.
Best regards,
Juan

 

[Audioguru]

R9 has nothing to do with the gain of the opamp. R9 and R11 are simply an attenuator that reduces the input level.
R10 can be replaced by a piece of wire since the opamp is Cmos with almost no input bias current. R10 and the stray capacitance to ground at pin 2 probably reduces its stability.

The gain of an opamp affects its stability. The stability is the worst when the negative feedback is the highest which results in a gain of 1. But this and most other opamps are designed to be stable with a gain of 1.

 

[crutschow]

Juan's description is not quite correct. Actually, an op amp gain of 1 (100% feedback) is the worst-case for op amp stability. If it's stable at a gain of 1 it will be more stable at higher gains (less feedback).
Some high frequency op amps are not even compensated for a gain of 1, only for gains greater than 1 (such as ≧5).

R9 (or R11) has no effect on the stability. It's the feedback to the inverting (-) input that determines the stability. Even though those resistors reduces the overall circuit gain to <1 by attenuating the input signal, the op amp itself is still operating with a closed loop gain of 1.

To complete the discussion, the only effect R10 has on the stability is that high values can lead to peaking of the response due to stray capacitance at the inverting node.

 

[Surendhar M]

R9 & R11 is just voltage divider. if you change the R9 value the output value will change. So if you want to make it stable then connect a capacitors parallel between R11 & R12. So that you will get stable output.

 

[crutschow]

R9 & R11 is just voltage divider. if you change the R9 value the output value will change. So if you want to make it stable then connect a capacitors parallel between R11 & R12. So that you will get stable output.

Where is R12?

 

[Surendhar M]

Where is R12?

Sorry its R10

 

[juan_zo]

Hello Jester!

Have you solved it?

 

[Jester]

The circuit is not unstable, I'm simply trying to understand the concept of gain less than 1 can be unstable.

Is it correct to state that with a non-inverting topology where gain = 1+ feedback ratio that a unity gain stable op amp will be inherently stable, and that only with inverting configurations with Rf/Rin < 1 that the circuit might be unstable?

 

[tggzzz]

The circuit is not unstable, I'm simply trying to understand the concept of gain less than 1 can be unstable.

The stability depends on the loop gain and phase margin, not on the complete circuit's gain.

 

[D.A.(Tony)Stewart]

The circuit is not unstable, I'm simply trying to understand the concept of gain less than 1 can be unstable.

Is it correct to state that with a non-inverting topology where gain = 1+ feedback ratio that a unity gain stable op amp will be inherently stable, and that only with inverting configurations with Rf/Rin < 1 that the circuit might be unstable?

NO, Not quite.

If it is guaranteed to be unity gain stable with (-) input then any higher gain with linear R feedback ought to be same or better for stability.

Some Video Amps only guaranteed stable minimum gain of 10

Adding an attentuator before (+) input lowers overall gain but does not affect feedback and hence does not affect stability.

 

[crutschow]

Re: OP amp stability gain &lt; 1

Sorry its R10

You don't want to add a capacitor between R9 and R10. That will cause peaking in the circuit response.

- - - Updated - - -

..................................
Is it correct to state that with a non-inverting topology where gain = 1+ feedback ratio that a unity gain stable op amp will be inherently stable, and that only with inverting configurations with Rf/Rin < 1 that the circuit might be unstable?

No. If the op amp is non-inverting unity-gain stable (as most are), then any combination of Rf/Rin will also be stable, since the equivalent non-inverting gain is always ≧1.
A non-inverting unity-gain stable op amp is stable for all reasonable configurations of resistive negative feedback. Only certain forms of reactive feedback will cause such an amp to become unstable (obviously you can use an op amp to make an oscillator, for example).

 

[Jester]

crutschow,

Thanks, I think we both stated the same thing.

 

[crutschow]

...............
Thanks, I think we both stated the same thing.

Well, you stated that "with Rf/Rin < 1 that the circuit might be unstable", and that's not the case. Even with Rf/Rin < 1 the op amp will still be as stable (or more stable) as compared to a non-inverting gain of 1 (which is the worst-case for stability).

 

[juan_zo]

Dear Jester,

I am going to explain how an Opamp works in an easy way, I know some users could criticize my way, I need to say I will glad if someone could explain us a better way. I hope you can solve your doubt and learn of our experience.

An Operational Amplifier (OA in next) has 2 signal input ( -: inverting input and +: non inverting input) and 1 signal output.

for DC and with gain1:
If on the + input you put a positive signal, the output is going to have a positive signal too. On the contrary,
if on the - input you put a positive signal, the output is going to have a negative signal. I attached a picture showing this.
Well, the OAs in general have very high gain, then if you do not set the gain, it will not to be stable, why? Because with their very high gain, they will saturate the output and it will be near to +Vcc or -Vcc.

How to control or set the OA's gain?
The best way is with negative feedback.

Which is Negative Feedback?
Negative Feedback is put on the input an output sample but with inverted level.
How you can do it?
It is not necesary to invert the output signal or output level "you have the Inverting Input" then if you put
a sample of the output, you are feedingback the OA.

When an OA is not stable?
A lot of answer you can hear about it, in my opinion it is when OA's gain fluctuates or when in a frecuency range
output's level variates more than 3dB.
Some people could say "When output's signal is distortioned or its signal is not in fase, it could be, however, I think those are problems of distortion not of stability.

How to set the gain with Negative feedback? It depends on your circuit, but you need to know that to less Rf(resistor between Output and Inverting Input) less gain. I worked with OA's in Audio for 20 years and I learned that the easiest gain 1 topology is connecting the output directly to inverting feedback. It topology has a lot of advantages in my opinion: Not extra component, not formulas, not frequency dependency, high input impedance, low output impedance and the best, not temperature problems, temperature vary a lot the gain, well, in it topology temperature is not a problem, maybe for this you can see it in a lot of audio schematics, electromedicine and instrumental equipment.

I hope I could explain my ideas clearly to you.

It is my pleasure to share my experience here.

Have a nice day.

Best regards,
Juan

 

[LvW]

Hi Juan,
I have, in general, nothing against your intuitive explanations of the opamp`s working principle.
However, everything connected with "stability" deserves some additional comments/corrections:

Well, the OAs in general have very high gain, then if you do not set the gain, it will not to be stable, why? Because with their very high gain, they will saturate the output and it will be near to +Vcc or -Vcc.

This has nothing to do with "stability" - it is just an effect of output saturation and/or input offset voltage.


When an OA is not stable? A lot of answer you can hear about it, in my opinion it is when OA's gain fluctuates or when in a frecuency range
output's level variates more than 3dB..

No - again, such observations have nothing to do with "stability".
At first, we have to discriminate between
(a) DC stability (of the operating point), which is ensured with sufficient dc feedback;
(b) Dynamic stability (against oscillations). As an unwanted consequence of dc feedback (see above) the dynamic stability is always degraded . That means the margin against stability gets smaller for increasing feedback. A quantitative analysis of this effect requires anaylyzing the loop gain (applying the Nyquist stability criterion).

 

[Jester]

Thanks to everyone for their comments.

It seems R10 has no purpose in the original diagram?


Focusing on the inverting topology, would this circuit be considered unstable?

 

[FvM]

It seems R10 has no purpose in the original diagram?

R10 in the original circuit might be intended as compensation for input current generated offset, but it's completely useless for a CMOS amplifier like OPAx348. The method is only appropriate for bipolar OP without internal input current compensation, in other words for |Ib| >> |Ios|.

Focusing on the inverting topology, would this circuit be considered unstable?

Did you learn anything from the previous discussion? The circuit is stable with any passive source impedance connected to J2.

 

[Audioguru]

An unstable opamp oscillates when negative feedback is added to reduce its gain, because the phase shift inside the opamp causes the high frequency negative feedback to become positive feedback.
A stable opamp does not oscillate when any amount of negative feedback is added, but with plenty ofd negative feedback the output probably has overshoots and ringing. It has its high frequency response reduced with an internal capacitor so that at a frequency where the phase shift would cause oscillation, the gain is less than 1.
Some opamps are stable only if their gain is 10 or more, allowing a higher frequency response and higher slew rate than if it was compensated to be stable when the gain is 1.

 

[juan_zo]

Dear LvW,
Thank you very much, for your good explanation!
And yes, I forgot the auto-oscillations, a very critical point for stability.
About the plane gain or frequency response in AC (less than 3dB) I think it is very important because if you need an unit gain amplifier and it does not have linear gain along the work frequency spectrum, output level could fluctuate between a half or twice input level and it must to be once. Maybe I have a concept problem with this, I am sorry in advance if it is so.

About Jester schematic, it is stable and its gain is less than once [-(0.3325 times)]. It is an inverter amplifier.

Have a nice day you all.
Rgds,
Juan