how to test the open loop gain?

[cetc1525]

measure open loop gain

I have designed a error opamp.the negative input is inner.only the positive input and output is outside.It's open loop gain can go to 80dB,Can anyone help me?

 

[gevy]

open loop gain hspice test

It is necessary temporarily to open loop of a feedback on AC in any place, having kept connection and modes of the circuit on DC.
After that to make the AC analysis of the changed circuit.
In result you receive magnitude to frequency and phase to frequency characteristics.

 

[sinatra]

Theorically you should open the loop to be able to find the open loop gain.
But I believe there is another problem in the case of this operational amplifier.
Maybe if you open the loop the 2 parts of the circuit where you opened (the one that became now the input of the open loop and the one that became the output) need to see the same bias and impedance conditions it had before the loop was opened otherwise the open loop gain results will be wrong.
It maybe difficult to realize this conditions.
This maybe possible with Spice ideal controlled sources.....
Anyone has other hints to add???
S.

 

[sutapanaki]

Do you need to physically measure the open loop gain in the lab, or you need to simulate the open loop gain? Obviously, simulating is easier in spice. In any case, though, you need to keep the bias conditions of the opamp intact in order for your measurement to make sense. There are techniques allowing you to measure/simulate the open loop gain without breaking the loop, thus keeping the biasing and the loading in place. This is the so called Middlebrook method. Although the original paper is probably hard to find, I'm sure that if you do a search in google you'll find a lot of references to this method. Here is one, which also points to other papers concerning this matter:

**broken link removed**

 

[hebu]

you should make sure the bias condition is the same as closed loop.
So that the performance of amplifier will in the real condition.
And, you need to make the same impedance looked from input
and output, keep the same environement to the amp,i.e give a signal
source with the internal impedance as the Zout of amplifier. But this
is diffcult to model accurately, but do it as possible, also the load of
the open loop should be the same as Zin of amp.
If you would like to check stability of this loop, I also suggest you
utilize step response to check the closed loop, and look at the response and
its ringing.

 

[Ipanema]

Hi,

I have this problem when measuring the phase margin of 4 stages op-amp connected as close loop, which consists of 1 stage diff amp cascaded with 3 stages of common source plus a big 5pF capacitor connected at the 2nd stage common source output to ground. I break two loops to verify whether there is any different in phase margin at different loop breaking point, one at the 1st stage common source output (seeing 5pF cap) and two at the 3rd stage common source output. At loop break point 1, I get phase margin of 95degree, but at loop break point 2, I only get 34degree of phase margin. Both points produce gain of 78dB. The input bias is accurate to 6 decimal points. May I know is it because of the 5pF capacitor that make my measurement inaccurate or other issue?

Thanks.