How to avoid Bipolar OPAMP oscillation ?

[satheeshvelu]

Hi all,
I simulated a Bipolar OPAMP Voltage Follower. The Frequency Response don't have any Oscillation.

But when connected a Resistor (45k) in the feedback, there's an Oscillation.

As this OPAMP is one of the many OPAMPs inside the IC, I could not add a capacitive load, which affects the Phase Margin.

Is there any other way to avoid the Oscillation..?
Attached the PDF.

Thanks,
Satheesh

 

[laktronics]

Re: OPAMP Oscillation

Hi,
Try putting a small capacitor across the feedback resistor, to compensate for the phase shift introcuced by opamp input capacitance.

Regards,
Laktronics

 

[satheeshvelu]

Re: OPAMP Oscillation

Hi

Thanks for your suggestion.
It works fine.
But I would like get it cleared, what does the input capacitance refer to..?
Inside the OPAMP we have only one Miller Capacitor, which is set to decide the Pole and Zero.

So in another way, instead of adding the capacitor across the feedback resistor, we can change this Miller Capacitance to set the Pole.

Am I right..?

Thanks,
Satheesh

 

[FvM]

Re: OPAMP Oscillation

Inside the OPAMP we have only one Miller Capacitor

Would be very comfortable, but actually you have more capacitory and poles inside an OP.

instead of adding the capacitor across the feedback resistor, we can change this Miller Capacitance to set the Pole

Yes, you can slow it down, so the input capacitance pole wouldn't be seen. Generally, OP external components must match it's speed class. A 45k feedback resistor with a 100 MHz OP isn't suitable.

 

[jeepx5]

OPAMP Oscillation

I'm not a bipolar guy, but I am quite confused why the feedback cap can remove or push the zero and pole pair to higher frequency?
And by introducing this cross coupled cap will the corner freq be lower in the second bode? if it does decrease the dominant pole, then the corss-coupled cap should not be too small cuz the feedback resistance is supposed to be much larger than the input one.

 

[laktronics]

Re: OPAMP Oscillation

Hi,
Actually it adds a zero and cancels the pole due to the feed back resistor and the input cap of the opamp, that is how I see it.

Regards,
Laktronics

 

[jeepx5]

OPAMP Oscillation

According to the second bode, the rise of the magnitude must be caused by a zero instead of a input pole.

Added after 2 minutes:

And it's amazing that in the bipolar world the GBW can be increased without any other cost just by introducing a resistor.

 

[FvM]

Re: OPAMP Oscillation

it's amazing that in the bipolar world the GBW can be increased without any other cost just by introducing a resistor

I don't think that this is the case, you just get a certain gain increase right of the resonance frequency. But the sketched bode diagrams are most likely not drawn to scale. Generally, it's easy to increase OP gain by "bad" feedback, but it won't be usable in regular OP operation.

There may be other effects, too. You have to check the circuit in detail, to know about. But I don't think that they introduce any new insights to wideband OP design.

 

[laktronics]

Re: OPAMP Oscillation

Hi,
I also did not notice the zero in the bode plot due to the feed back resistor. For that matter, I only went by his problem introduced by the feed back resistor for which I suggested a solution.

Regards,
Laktronics.

Added after 19 minutes:

Hi,
Is this the effect due to a feedforward design in the amplifier?
Regards,
Laktronics