Hi, I have a question about closed loop OPAMP frequency response.

Yunsik Chung

Intellectual 1150 points

Other Parts Discussed in Thread: OPA690

Hi,

If OPAMP's unity gain frequency is Fu.

In non-inverting opamp, Fu = Acl * Fcl(closed loop corner frequency)

In inverting opamp, Fu = (Acl+1) * Fcl.

I cannot understand what makes this difference between two topology.

Can you explain in more detail how come non-inverting topology has higher corner frequency than inverting topology?

over 6 years ago

0 Michael Steffes over 6 years ago

Guru 48395 points

Morning Yunsik

The ideal gainbandwidth model for a voltage feedback op applies for only a single pole loop gain. It is a very approximate model for any real device.

The gain you need for the model is the noise gain which is the inverse of the divide from the output back to the differential input voltage. In gain of +1 that is 1 in gain of -1V/V that is 2 if the source is zero ohm.

This confusion does come up regularly, especially with inverting attenuators, discussed it (and many other things) in this article,

https://www.planetanalog.com/are-inverting-attenuators-with-voltage-feedback-op-amps-inherently-unstable-and-why-might-you-want-to-apply-them-insight-11/

And then a more accurate way to estimated closed loop bandwidth is in this one,

https://www.planetanalog.com/why-is-amplifier-gbp-so-confusing-insight-12/

0 Former Member over 6 years ago

Hello Yunsik,

Actually, the bandwidth of the two circuits will be the same. You need to use the "noise gain" or the non-inverting gain when calculating the closed-loop bandwidth and not the signal gain.

Why is this the case? The short answer is that both circuits ultimately have the same negative feedback system. You can see a more in-depth explanation here: https://training.ti.com/system/files/docs/1213%20-%20Bandwidth%203%20-%20slides.pdf

Please let me know if you have any further questions.

Regards,

Daniel

0 Michael Steffes over 6 years ago in reply to Former Member

Guru 48395 points

Morning Daniel and Yunsik,

Certainly, a lot of the time (VFA ckts wise) the Noise gain is the non-inverting signal gain. But not always. More accurately, it is the inverse of the feedback divider to the "differential" input voltage doing a superposition analysis with the sources zeroed out. That detail becomes important in circuits like the ubiquitous MFB active filter. For an example NG analysis when you really need to focus in on the differential feedback voltage (since the MFB has "multiple" feedback paths) consider Section 2 in this detailed MFB design application note,

http://www.ti.com/lit/an/sboa114/sboa114.pdf

But the simple point is correct of course, most circuits what you care about is the non-inverting signal gain which is often the same as the NG. Of course this all goes out the window moving on to current feedback and FDA designs - in different ways.

0 Yunsik Chung over 6 years ago in reply to Former Member

TI__Intellectual 1150 points

Hi Daniel,

Sorry for the late reply.

I understood this concept:)

Thank you very much.

Yunsik

0 Yunsik Chung over 6 years ago in reply to Michael Steffes

TI__Intellectual 1150 points

Hi Michael,

Sorry for the late reply.

I really appreciate your detail explanation!

I understand this concept:)

Thank you very much.

Yunsik

0 Michael Steffes over 6 years ago in reply to Michael Steffes

Guru 48395 points

Morning Daniel and Yunsik,

The MFB example as a case where you need to do differential feedback voltage analysis is not applicable actually. Don't know what I was thinking, but here is a better example - I used to put tricks like this into a lot of datasheets, and they are valid, but not used that much I don't think, From the slew boosted unity gain stable OPA690 VFA datasheet,

http://www.ti.com/lit/ds/sbos223g/sbos223g.pdf

Amplifiers

Amplifiers forum

Hi, I have a question about closed loop OPAMP frequency response.