# How to simulate PSRR- and PSRR+

Status
Not open for further replies.

#### AMSA84

I have a question:

My OTA has a power supply from 0V to 3.3V. I want to measure the PSRR+ and PSRR-. I know that to measure those one must put in series with the VDD voltage and GND a vsin with 1V in the VAC field. However there are some things that is confusing me.

1st: We have to measure the PSRR+ and PSRR- with open loop, for example, with 1pF capacitor in the output, or with a, for example, 1pF capacitor in parallel with a 10M resistor, or just a 10M resistor in the output of the amplifier?

2nd: I have read around some books and throught the web that this is to be measured in closed loop. I have seen people doing that with a buffer connection and other (baker's book) using a large resistor making the feedback and a very large capacitor connected between the output and ground. For the baker's example, he uses something like a feedback resistor of 100M and a capacitor around 10uF. Well, this is correct? We're not cheating our amplifier composing the testbench like that?

3rd. I have been told that one should use the feedback network that it will be used in that amplifier. However, at this stage it is supposed for me don't know anything about what will be inserted in the feedback, you understand what I mean?

So what to do?

More, in baker's book he defines the PSRR as the (open loop gain) / (vout/VDD) or (open loop gain) / (vout/VSS) -> in my case GND.

How this is translated into that thing, independently on the configuration one use (above mentioned), of using one VSIN source in series with VDD or GND (1 VAC) and ask for the output of the amplifier?

Kind regards and thank you in advance!

#### erikl

##### Super Moderator
Staff member
My OTA has a power supply from 0V to 3.3V. I want to measure the PSRR+ and PSRR-. I know that to measure those one must put in series with the VDD voltage and GND a vsin with 1V in the VAC field. However there are some things that is confusing me.

1st: We have to measure the PSRR+ and PSRR- with open loop, for example, with 1pF capacitor in the output, or with a, for example, 1pF capacitor in parallel with a 10M resistor, or just a 10M resistor in the output of the amplifier?
There's no standard, the used output load just should be stated. If you know the application, use its load; if you don't know it, use a most probable load. OTAs (without buffer) usually are thought to drive capacitive loads only (i.e. CMOS inputs).

2nd: I have read around some books and throught the web that this is to be measured in closed loop. I have seen people doing that with a buffer connection and other (baker's book) using a large resistor making the feedback and a very large capacitor connected between the output and ground. For the baker's example, he uses something like a feedback resistor of 100M and a capacitor around 10uF. Well, this is correct? We're not cheating our amplifier composing the testbench like that?
Just a very little bit (the output usually doesn't drive such a load). However, this is the most simple - and most often used - configuration to achieve full DC and no ac feedBack (at the left and right side limits of the frequency axis). Obviously, this creates a closed loop configuration for DC, and an open loop configuration measurement ("no ac feedBack") far above the 1/2πRC frequency defined by the feedBack network.

3rd. I have been told that one should use the feedback network that it will be used in that amplifier. However, at this stage it is supposed for me don't know anything about what will be inserted in the feedback, you understand what I mean?

So what to do?
PSRR measurement is in no way standardized - s. this tutorial from AD, e.g.: View attachment AD__OpAmp-PSRR-tutorial_MT-043.pdf.
You can measure it in open loop, in closed loop, or in your special application configuration - if you know it. It doesn't matter, just state your measurement conditions clearly.

If no measurement configuration is stated, in most cases the above mentioned method is used - s. Fig. 1 of the tutorial given above, or many opAmp dataSheets.

More, in baker's book he defines the PSRR as the (open loop gain) / (vout/VDD) or (open loop gain) / (vout/VSS) -> in my case GND.

How this is translated into that thing, independently on the configuration one use (above mentioned), of using one VSIN source in series with VDD or GND (1 VAC) and ask for the output of the amplifier?

To my understanding PSRR+ = vout/VDDac and PSRR- = vout/GNDac , with actually used measurement configuration stated.

AMSA84

### AMSA84

Points: 2

#### AMSA84

I see. Ok, I'll do it with my load. Thanks.

Status
Not open for further replies.

Replies
13
Views
9K
Replies
3
Views
2K
Replies
1
Views
1K
Replies
5
Views
2K
Replies
0
Views
2K