Ground

I wonder if the noise you measure and the grounding effect is an artifact of your scope measurement? Common-mode ground noise currents can cause that.

Ahh... OK. With my DMM in AC voltmeter I read 0.9mV (rms) no matter how I ground it.

The DMM only has a BW of 300 kHz ( I think) still it should show some change.

(Well not quite it changes from 0.885 to 0.865 mV when I ground it. I'll have to dig up a faster AC voltmeter....

Thanks Win

George H.

Use a grounded wall transformer?

I tried floating the 'scope but it did nothing different. The CUI things could be spraying magnetic stuff all over. It's got a place to go when I ground it.

George H.

We use some Triad Wall switchers, THey tend to let in hash from the AC line. Not sure if the CUI do the same. Maybe plug one into an ISOBAR strip and test again.

Cheers

It's hard to "float the scope" capacitance-wise, at high frequencies.

Ahhh .... "failure to communicate"? I meant instead of your wall wart, use a wall transformer that has a ground output.

Leave everything as is but unplug the wall wart. Wait for caps to discharge then probe again. What do you see?

Mark L. Fergerson

Yeah I was thinking of that... It's all fast edges... With the "floating 'scope" it's still less noisy if I make a hard connection between 'scope ground and the circuit ground.

With the circuit floating doesn't the differential noise turn into common mode? Those high frequency current have to go somewhere.

George H.

Oh OK, We use these (universal) Phihong wall warts with different (plug) adapters for different countries. I guess I could try something else.

George H.

~5mV of crud.. If I ground the pcb (with no power) it turns into ~10 mV of crud.

George H.

is the wall wart a switcher or an older iron transformer?

Nope a new Phihong. (24V) It's all moot now. I put it in a box with a connector going to the circuit it will be powering and the noise is fine. I'm guessing it was the long cable length I had... I don't know.

Ground can cause problems, sometimes I can draw pictures, follow the currents and figure it out. Other times I cut and hack till it works.

Thanks all, George H.

Circuits spread out over a board can make good antennas even if you've done well with local grounding. I've had to add a high CMRR instrumentation amp or diff opamp to some moderately fast low noise circuits that couldn't be probed by meters or scopes due to pickup from cables, power wires or missing covers. Won't get you much past 100 MHz but you can get to uV for stuff in the 10 MHz range.