How to test the bandwidth of a real opamp

[jiangyanjin]

Hi,

I want to know how to test the GBW of a real opamp,

and how to setup the circuit for this test?

Thanks.

 

[jiangyanjin]

Hi,

I want to know how to test the GBW of a real opamp,

and how to setup the circuit for this test?

Thanks.

Can anybody help me?
Thank you!

 

[LvW]

Hi,
I want to know how to test the GBW of a real opamp,
and how to setup the circuit for this test?
Thanks.

It´s rather simple:
When the opamp has a single pole response (20 dB/dec drop until gain is 0 dB), I would propose to use a suitable amount of feedback (resulting in a noninverting gain of app. 20...40 dB). Then, increase the frequency until the gain has dropped to 0 dB.

 

[frankliner]

I would do something similar, but simplify the calculations.

Set the gain to something reasonable = G. Input a signal which produces 1Vp-p. Set a marker for .707V. Increase the frequency = BW until the peak is at the marker. now you have G*BW.

Make sure the input signal is the same for both measurements. My function generator output impedance sometimes affected results.

 

[jiangyanjin]

I would do something similar, but simplify the calculations.

Set the gain to something reasonable = G. Input a signal which produces 1Vp-p. Set a marker for .707V. Increase the frequency = BW until the peak is at the marker. now you have G*BW.

Make sure the input signal is the same for both measurements. My function generator output impedance sometimes affected results.

In fact, I want to plot the curve that Open-loop Gain VS. Frequency, I want to know how to setup up a test circuit to plot this curve with Network Analyzer.

 

[LvW]

In fact, I want to plot the curve that Open-loop Gain VS. Frequency, I want to know how to setup up a test circuit to plot this curve with Network Analyzer.

It´s not easy. There are several items to be considered.
1.) DC offset.
a) Compensation of the dc offset error which otherwise would drive the output into saturation.
b) Provide a reasonable amount of feedback which keeps the output within the linear range. However, in this case, you cannot record the open loop gain at low frequencies, but mostly it is not so important.
c) Provide 100% feedback only for dc (and very low frequencies), for example with two large resistors in series (2*100 kohms) and a large high quality capacitor between ground and the node between both resistors.
2.) Be aware of the slew rate which influences measurements at very high frequencies in the vicinity of the transit frequency. Thus, try to measure with signal amplitudes as low as possible.