mastodon.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
The original server operated by the Mastodon gGmbH non-profit

Administered by:

Server stats:

373K
active users

minute

> Contrary to popular belief, differential signalling does not affect noise cancellation

> There exists great confusion as to what constitutes a balanced interface

@mntmn@mastodon.social One gets noise cancellation just by terminating V+ and V- to the same impedance, instead of terminating one side to circuit ground. External noise induces roughly the same noise voltages on both lines, so noise is rejected. It works regardless of whether the signal is symmetric or not. This is called "balanced" signaling. "Differential" signaling is a special case of balanced signaling in which the voltages on both lines are also equal and opposite. My understanding, if I got it right, is that balanced signaling is sufficient for low-noise audio, so it's wrong to say this noise suppression comes from differential "signaling" - it does not, it comes from the termination impedance. But for high-speed digital interfaces, I think differential signaling is still needed to suppress radiated EMI.

@niconiconi yes, this is also how i understand it. so i'm wondering if intra-pair skew has a bigger bad effect on PCIe 3.0 signal integrity than the diff pair impedance. it's somehow hard to find some good material on the effect of different impedance values on diff signalling integrity.