Influence of open-loop gain in opamp frequency response.

[Theo]

Hi all,

I used to think that the bandwidth of an opamp circuit was only determined
by the gain-bandwidth product. Now consider the inverting amplifier circuit
with an OP177. The datasheet gives an gain-bandwidth product of 400 kHz min.
So if R1=R2, you would expect a bandwidth of at least 400 kHz.

This is not what I measure and this is also the reason of this posting.
Instead of the bandwith I expect, I measure a bandwidth of about 40-50 kHz
of my circuit. (The circuit contains two inverting opamp configurations with
each an OP177; the first stage is used for impedance correction; the second
stage for driving a cable. Both configurations have a closed-loop gain of
1.)

I suspect that the cause is the influence of the open-loop gain.With the
open-loop gain included, the formula of the inverting amplifier circuit is:

R2
Acl = - ---------------------
R1 + (R1+R2)/Aol

R1 = resistor between the circuit input and the opamp inverting input
R2 = feedback resistor between inverting input and output
Acl = closed-loop gain
Aol = open-loop gain

If you take a look at the open-loop frequency response figure, you can find
that the gain at 100 kHz is only 10-15 dB. With two OP177 circuits cascaded,
the total gain is much smaller than 1 at 100 kHz. And this is also what I
measure!

Is this correct or am I totally wrong?

regards,
Theo.

 

[Jim Thompson]

Hi all,

I used to think that the bandwidth of an opamp circuit was only determined
by the gain-bandwidth product. Now consider the inverting amplifier circuit
with an OP177. The datasheet gives an gain-bandwidth product of 400 kHz min.
So if R1=R2, you would expect a bandwidth of at least 400 kHz.

[snip]

A common error with *inverting* amplifiers set for unity gain is to
assume you get a bandwidth equal to the GBW product. Not so, your
closed loop (internal) gain is TWO, so you get GBW/2.

Please post your whole schematic... something doesn't seem quite right
with your descriptions.

...Jim Thompson

 

[Fred Bloggs]

Hi all,

I used to think that the bandwidth of an opamp circuit was only determined
by the gain-bandwidth product. Now consider the inverting amplifier circuit
with an OP177. The datasheet gives an gain-bandwidth product of 400 kHz min.
So if R1=R2, you would expect a bandwidth of at least 400 kHz.

[snip]

A common error with *inverting* amplifiers set for unity gain is to
assume you get a bandwidth equal to the GBW product. Not so, your
closed loop (internal) gain is TWO, so you get GBW/2.

Please post your whole schematic... something doesn't seem quite right
with your descriptions.

Uh-huh....OP177 S.R.=0.3V/us typ

 

[Theo]

Hi all,

I used to think that the bandwidth of an opamp circuit was only determined
by the gain-bandwidth product. Now consider the inverting amplifier circuit
with an OP177. The datasheet gives an gain-bandwidth product of 400

kHz
min.

So if R1=R2, you would expect a bandwidth of at least 400 kHz.

[snip]

A common error with *inverting* amplifiers set for unity gain is to
assume you get a bandwidth equal to the GBW product. Not so, your
closed loop (internal) gain is TWO, so you get GBW/2.

Yes, of course :-(

In the meantime I replaced the OP177 by the AD711, which has a much higher
open-loop gain at 100 kHz. Now the bandwidth of the circuit is how I had
intended it.
My 'personal' problem still is that obviously the GBW product is not the
only factor that determines the bandwidth of a circuit. I had never realized
that.

Theo.

i hope you don't have a problem with the fact that the closed loop BW of
an amp is less than the GBW.

in other words, the rolloff of the open loop amp starts out at a low
freq and approaches unity gain at the GBW.


---- open loop gain
*
-----------* closed loop gain
*
*----- unity gain
| |
F3dB Ft

see, the BW of the top line is less than that of the lower line. BW and
GBW are not the same thing. i'm thinking you're confusing the two.

BW3dB = Ft/Acl where Ft is the GBW product.
Yes.

but if your Aol was 400kHz, your unity gain circuit should be 200kHz,
yet first you say you measure 40-50kHz and then you say 100kHz.

I ment that, according to the specs, the open-loop gain of the AD711 is much
higher than the open-loop gain of the OP177 at 100 kHz.

here's a new one on me. sniped from your post:
R2
Acl = - ---------------------
R1 + (R1+R2)/Aol

where'd that come from? (R1+R2)/Aol is negligible.

For an ideal opamp the open-loop gain Aol is infinite. The formula then is
simply Acl = -R2/R1. Solve following the equations:
. u- = R1/(R1+R2) * u_in + R2/(R1+R2) * u_out
. u+ = 0
. u_out = Aol * (u+ - u-) = - Aol * u-
Then you will get the formula i mentioned before.
I you also take a look at the open-loop gain vs. frequency figure in the
specs, then you see the influence of the open-loop gain at increasing
frequency.

Of course the slew rate also restrict the bandwith

regards,
Theo.