How is the differential mode noise suppressed by adding a capacitor across the Lines

Thread Starter


Joined Oct 26, 2017
I was reading about the common mode noise and differential mode noise.
I don't understand how a differential mode noise gets coupled and how adding a capacitor across the line will help.
I understand about common mode noise. Like, if we have a noisy source, it might equally couple on both lines and we can use a common mode choke to reflect back the common mode noise back to the upstream source.
But how is a differential noise generated on the lines? Is it generated by places the lines close to differential signals?

My questions:
  1. 1. In general, does the noise affect only power lines or signal lines as well? I am asking because, I have seen app notes and other illustration depicting these concepts only in the context of Power supplies.

  2. 2. Is differential mode noise, generated on the lines, by placing the differential mode signals near them? Or how else, will the line and neutral wires have noise voltage of opposite polarity? I assume that differential mode noise is obtained only by placing the circuit near differential mode signals. Am I correct in this understanding?

  3. 3.And I understand how a common mode choke attenuates the common mode signal. But how does adding a capacitor between the lines, attenuate the differential mode noise? Can someone provide some help on this.
Thank you


Joined Jan 8, 2017
A capacitor has a lower reactance at high frequencies than it does at low frequencies (It has infinite reactance at DC.) so it shorts out the higher frequencies.


Thread Starter


Joined Oct 26, 2017
I understand that. But differential noise, is something that occurs on both the right at opposite phase right? How does the capacitor handle this

And can someone provide answer to all my 3 questions


Joined Nov 5, 2019
Simplistically, the capacitor across the diff lines converts the diff noise (or part of it) into common mode noise, which then, in a perfect world, is rejected/ignored by the differential receiver. That said, I have never used a cap across a diff pair, except to slow down the rise and fall times (edge speeds) of the signals for signal integrity concerns.

Differential noise is difficult to couple into a properly designed differential interface. That's why they're used:). The main reason reason noise would couple differentially is if there was some imbalance between the two differential signals (such as them not being routed properly), or the noise source is much closer to one of the differential signals in the pair than in the other.