# CMRR and PSRR impact on performance

#### Choon Kwang

Joined Jun 6, 2022
9
Hi, I am trying to analyse a replacement op-amp for an existing one that will be EOL-ed. The op-amp is configured as differential configuration. The input voltage will range from 0Vdc to 3.3Vdc. The replacement op-amp has slightly different CMRR and PSRR values. How do I calculate the impact to the same circuit design with the new op-amp?

#### Attachments

• 1.4 MB Views: 12

#### Ian0

Joined Aug 7, 2020
7,522
If not they will have a bigger impact on CMRR than the op-amp.
The effect on performance of the PSRR will depend on how much power supply noise you have that needs rejecting.

#### Choon Kwang

Joined Jun 6, 2022
9
If not they will have a bigger impact on CMRR than the op-amp.
The effect on performance of the PSRR will depend on how much power supply noise you have that needs rejecting.
They are all 1%. Do you mean 1% or 0.1%? But how can I do the calculation to show the impact?

#### Ian0

Joined Aug 7, 2020
7,522
They are all 1%. Do you mean 1% or 0.1%? But how can I do the calculation to show the impact?
I do mean 0.1%.
1% resistors give you 40dB CMRR. Any op- amp will be so much better that it makes no difference.

#### prairiemystic

Joined Jun 5, 2018
297
We don't know the old op-amp's part number, the proposed new op-amp's part number, or the application. Hard to help.
OP's schematic are all high value resistors so I'm assuming it's a FET input unless DC offset drift due to input bias currents, is not a priority. What frequency response do you need? Speed?

#### ag-123

Joined Apr 28, 2017
273
there are more things than CMRR to consider, one of which is your supply voltages to the op amp.
https://www.ti.com/lit/pdf/sloa030
quoting from the document

This is a best graph I've (ever) seen about OpAmps. It goes well beyond CMRR if you are constraining your power supply options e.g. to between 0-3.3v. In this graph alone, none of the op amp (except the rail-to-rail op amp) will work with 0-3.3v unless you are happy to make do with only a portion of the viable input/output voltage which that is 'linear'.

The other thing about CMRR is, normally there is some max voltage ranges where they specify that the CMRR is valid. Above that voltage, it won't achieve the CMRR. This *locks out* a whole range of *generic* op amps for high side current measurements.
And it leaves the whole world the only option to use specialized purposed designed op amps for high side current measurement. Or make your own solution. high side current measurements opamps are competitive e.g. INA219 https://www.ti.com/lit/gpn/ina219
so much so that there is now a shortage
i.e. for those using INA219 (and all other 'high side current measurements') it is likely all that op amps are in short supply.
and it isn't limited to only high side current measurement op amps.
It seemed even mostfets are short
https://community.infineon.com/t5/Power-MOSFET/shortage-of-IPN80R900P7-MOSFET/td-p/354034

low side current measurements is where it is possible to *cheat* and still measure currents to some extent, say with a LM358 (not a good choice).
that *cheating* has some implications, it is disastrous if there is another common ground circuit flowing somewhere else, so it should only have only one single gnd outlet and the only one pipe to measure currents.

Last edited: