compensating overshooting

[Keruskerfuerst]

What is the best way to compensate the overshooting of an audio op amp (like BB OPA 627 or 134PA)?

 

[Hardwire]

There isn't much overshoot in an op amp unless you are driving capacitive loads. Is that what you mean?

 

[Optikon]

You have a stability problem that needs to be analyzed first before any sensible method of compensation can be determined. There is no single way that works in all cases.

 

[Optikon]

There isn't much overshoot in an op amp unless you are driving capacitive loads. Is that what you mean?

Not true. Many cases can cause overshoot. Capacitive loads working on output impedance of amplifier to create a feedback pole are only ONE example.

 

[Hardwire]

Not true. Many cases can cause overshoot. Capacitive loads working on output impedance of amplifier to create a feedback pole are only ONE example.

I guess what I mean is that there isn't much inherent overshoot to most op amps (right?) Step a voltage into a non-inverting opamp and there won't really be any overshoot at all, right?

 

[Keruskerfuerst]

In case of inverting and non-inverting opamp, the datasheet tells me a overshoot factor of 2%-25% when driving a 100pF load.

I have an audio application (2 way active crossover network).
Input: 2x BB OPA 627AP, input frequency limited to 100kHz.
Differential stage: BB OPA 134PA.
24dB/oct, low/high pass.
BB DRV 134 output stage.

All resistors are Vishay non-magnetical. Impedance in this network is below 2kOhm.

I don´t have any stability problem in my network.

Do I have to limit the frequency range in this network directly before the output to prevent overshoot?

 

[Hardwire]

Well, you're probably not going to be driving a purely capacitive load in a crossover circuit, so you probably don't have to worry about it. Check out some crossover designs on the web.

 

[gootee]

In case of inverting and non-inverting opamp, the datasheet tells me a overshoot factor of 2%-25% when driving a 100pF load.

I have an audio application (2 way active crossover network).
Input: 2x BB OPA 627AP, input frequency limited to 100kHz.
Differential stage: BB OPA 134PA.
24dB/oct, low/high pass.
BB DRV 134 output stage.

All resistors are Vishay non-magnetical. Impedance in this network is below 2kOhm.

I don´t have any stability problem in my network.

Do I have to limit the frequency range in this network directly before the output to prevent overshoot?

That might work. But I think more information is needed.

Can you post a schematic?

What are your signal levels, etc? What are the loads?

Are you using a step input, when you see overshoot? If so, what size step, with what rise-time? And how much overshoot, and with what frequency and damping time?

Do you have Spice models for your opamps, etc, and loads, and have you simulated the circuit?

- Tom Gootee

**broken link removed**

-

 

[mneary]

How do you observe this overshoot? Are you using a scope with enough bandwidth and properly compensated probes?

 

[Keruskerfuerst]

I am using the square wave output of the ozilloscope as an input signal.
The probe is adjusted with this square wave output signal.
Oszilloscope bandwith is 200MHz.
Overshoot at output is 10% with a "duration" of 0,8us.

 

[Nigel Goodwin]

Why do you consider this a problem?.