Earlier this month Power Electronics received a question about "How to measure PSRR of a linear regulator circuit that powers an oscillator for an ADC. How can I make this measurement using my oscilloscope?" We asked our expert in PSRR, Steve Sandler to weigh in on the subject (view his response here) and also encouraged readers to submit other questions they had about the space.
Below are the responses to the questions that Steve had time to answer.
I've seen PSRR called power supply rejection ratio and also power supply ripple rejection. Which do you prefer? - Brian
Dear Brian,
I believe either is acceptable, but I personally prefer rejection ratio, since we're not only concerned with ripple, but the attenuation of all signals. Another aside is that some like to see this is a positive graph, and others like to see it as a negative graph. This is confusing to some, but I find either ok. It depends on whether we refer to it as rejection (a positive number) or as a gain which would be a negative number (hopefully).
Hi Steve,
Does PSRR affect other analog ICs other than LDOs? - Matt
Dear Matt,
Yes, certainly. The PSRR issues I see most other than LDO's are opamps, voltage references and also ADC clocks.
Steve,
What is the definition of PSRR? I am having a hard time understand the application - Regards, Tito
Dear Tito,
This is Power Supply Rejection ratio, which is a measurement of the attenuation of signals from the power supply input to the device output.
Hello Steve,
How does power supply rejection ratio impact power supply performance? - Bill
Dear Bill,
Poor PSRR generally results in noise not inside the power supply but the load. For example it increase opamp noise, voltage reference and ADC noise and circuit noise in general
Continue reading Steve's responses pg. 2
Dear Steve,
How does frequency impact PSRR? - Joe
Dear Joe,
In general the PSRR performance looks a lot like the open loop gain plot, so high at low frequency with a pole and then the PSRR degrades with increasing frequency. Different devices appear differently. The example below is for an opamp. There are better and worse performing devices than this and in general the higher the PSRR the better. A better metric might be the PSRR multiplied by frequency, much like the gain bandwidth of an opamp. On reason we are so concerned with stability is that poor stability degrades PSRR. An example of this is shown below. The "dip" in PSRR is the impact of poor stability at the frequency of the dip