Op-Amp - Power Supply Rejection Ratio

[Nikos A.]

Hi everyone, I am trying to select the best-suited op-amp for my application and I am looking the MCP6V61T-E/OT from Microchip

https://eu.mouser.com/datasheet/2/268/20005367B-1099368.pdf

I am a bit confused regarding the PSRR referring to the datasheet. On pages 1 and 3 says that the minimum PSRR is 117dB and the typical is 134dB. So far so good...

From figure 2-13 PSRR vs Ambient Temperature, it's shown that the PSRR varies from 129-138db for temperatures -50 to +125 (far away from the min. PSRR of 117dB that mentioned before)

But, if you look at figure 2-26 PSRR vs Frequency then the highest PSRR is around 100dB for DC, going down to less than 20dB for frequencies above 50kHz

I am really confused... Can someone explain to me all these values as well as the importance of each one? How is it possible the highest PSRR on figure 2-26 be less than the typical value (134dB) on specification table (table 1-1).

[magic]

117dB is the worst-case room temperature DC PSRR of the lousiest unit that their QA wouldn't throw to trash.

129-138dB is the variation of an average unit over temperature. A lousy unit may be worse than 117dB at temperature extremes, unless there is a guaranteed "PSRR at any temperature" spec somewhere.

The 100dB figure is for 10Hz, not for DC. I suspect that open loop gain still continues to rise as you gown down to 1Hz or even a hair below and that's why PSRR improves too.

[macboy]

...
From figure 2-13 PSRR vs Ambient Temperature, it's shown that the PSRR varies from 129-138db for temperatures -50 to +125 (far away from the min. PSRR of 117dB that mentioned before)

But, if you look at figure 2-26 PSRR vs Frequency then the highest PSRR is around 100dB for DC, going down to less than 20dB for frequencies above 50kHz

I am really confused... Can someone explain to me all these values as well as the importance of each one? How is it possible the highest PSRR on figure 2-26 be less than the typical value (134dB) on specification table (table 1-1).

Note that Figure 2-26 does not go to DC.  If you were to extend the X axis of Fig 2-26 two more decades left (down to 0.1 Hz) then you can see that the PSRR lines will extend to the 134 dB mark around that point.  The typical PSRR or 134 dB is listed under the "High DC Precision" features, so it is specified at DC. This makes sense, because zero drift chopper type amplifiers are intended for use at "DC" (whatever that means), and are highly optimized for that.

An op-amp like this might be used for high precision DC uses, such as a DC voltage reference (i.e. scaling a ~7 V zener voltage to 10.0000 V), buffering the output of a high impedance source such as a Kelvin-Varley divider, etc.

[Nikos A.]

117dB is the worst-case room temperature DC PSRR of the lousiest unit that their QA wouldn't throw to trash.

129-138dB is the variation of an average unit over temperature. A lousy unit may be worse than 117dB at temperature extremes, unless there is a guaranteed "PSRR at any temperature" spec somewhere.

The 100dB figure is for 10Hz, not for DC. I suspect that open loop gain still continues to rise as you gown down to 1Hz or even a hair below and that's why PSRR improves too.

Aha, so the lousiest value (117dB), as well as the 129-138dB variation, are referred to DC values.

In conclusion, this op-amp is not recommended for high-frequency noise since its PSRR is really low, isn't it?

I want to supply the op-amp from a 2MHz switching converter, that's why I am concerning about the PSRR.

[Wimberleytech]

117dB is the worst-case room temperature DC PSRR of the lousiest unit that their QA wouldn't throw to trash.

129-138dB is the variation of an average unit over temperature. A lousy unit may be worse than 117dB at temperature extremes, unless there is a guaranteed "PSRR at any temperature" spec somewhere.

The 100dB figure is for 10Hz, not for DC. I suspect that open loop gain still continues to rise as you gown down to 1Hz or even a hair below and that's why PSRR improves too.

Aha, so the lousiest value (117dB), as well as the 129-138dB variation, are referred to DC values.

In conclusion, this op-amp is not recommended for high-frequency noise since its PSRR is really low, isn't it?

I want to supply the op-amp from a 2MHz switching converter, that's why I am concerning about the PSRR.

Maybe you should not rely only on the opamp's PSRR to clean up an ugly SMPS power rail.  Clean it up before powering the amplifier.
Essentially, you are asking for an amp with considerable open-loop gain at 2MHz otherwise.

[Nikos A.]

Thanks for the advice guys!! As Wimberleytech said I will focus on minimizing the ripples on the power rail