125 kHz ground noise

ajras

We're looking to clean up some ground noise in our lab, specifically a 125 kHz signal with a long train of harmonics (out to 10s of MHz). Unplugging the kasli crate, the 125 kHz (and harmonics) go away. Has anyone else measured something like this? After reviewing our hardware specs, we're struggling to identify what is 125 kHz in the kasli, why there are so many sidebands, and how to mitigate it.

Thanks in advance!
AJ

claytonho

it could be the power brick, we had a similar issue, though i don’t remember exactly what frequencies we found spurs at. you could try powering artiq with a reliable dc power supply that can go 12V 5A and see if it goes away

sb10q

What power brick is it?

ajras

We're also suspicious it's the power brick. Thanks for the suggestion! We see this with the XP Power AKM65US12 and AKM65US12C2

mschoeler

Did you ever figure out the source of the problem or even a solution?

We're having the same problem with a 125kHz noise coming from the ground of our TTL BNCs (other grounds aren't affected).

ajras

We are still suspicious it's the switching power supply, but we currently don't have a linear supply that can go to 12V 5A to verify.

sb10q

If you don't have many cards, the current could be substantially less than 5A. Do you have 12V 3A perhaps?

evilmav

So we've been hunting a _112kHz noise source either that looked suspiciously like a switching regulator and turns out, disconnecting DIOs did got rid of it, connecting DIOs of the same bank didn't do much, but connecting DIOs of different banks (of the same EEM) simultanously seemed like adding uncorrelated noise sources. The NXE2S1205MC used to provide isolated voltages according to the current version's schematics is specified at f=100kHz switching typ. ... Fits @mschoeler 's description...