Unexplained interference in PSU test lab

[Phaedrus]

EU lab has a Chroma AC source, and they test at both 230VAC and 115VAC.

[Phaedrus]

The testing spec calls for both a 10uF and a 100nF capacitor at each test point. Turns out I had used 10nF by mistake. I put in 100nF caps and the transient went down! But is still there, and is showing even on known-good power supplies.

 

[KJDS]

Are you using the scope in high impedance or 50 ohm mode?

[Phaedrus]

1 Meg.

[Jay_Diddy_B]

Hi,

Have you tried with the power supply under test plugged directly into an outlet?

I don't know if your Extech AC power supply is either making a contributing to the noise or increasing the impedance in the ground lead.

I like to use this set up for measuring ripple and noise. Since we a dealing with RF, this circuit is matched for 50 Ohms.




Here is a picture of my test cable and the DC block. I use a 1uF capacitor because a standard (RF) DC block has too high a corner frequency.



I do a check by connecting the signal and ground to the ground on the power supply output. I should get no signal, if I do there is common mode noise between the scope and power supply grounds.


Jay_Diddy_B

[Phaedrus]

Hi Jay,

Switching between the Extech and various wall outlets was among the first things I tried. Unfortunately, this doesn't seem to be as simple an issue as mains interference.

[Jay_Diddy_B]

Hi,

I wasn't think that it was mains interference. I was think that the Extech AC source might be a switching, class D, amplifier type device, and contributing to the noise.

Jay_Diddy_B

[Phaedrus]

Fixed it!




Changed a number of things, but the biggest difference was decreasing the length of the electrolytic capacitor leads. Shortened each of them by 5-10mm (for a total length of ~10mm) and the ripple instantly dropped by 30mV and the spike almost disappeared. My readings now approximately agree with our Europe lab.



Thanks all! Helped a lot! 

[juanfermed]

Great you fixed it, but does that mean the  capacitor leads were picking up some kind of noise? or the inductance of the leads was causing that spike to appear? What was the cause of the problem ultimately? I am a total noob, cant guess. Thanks!

[Phaedrus]

I think it was inductance in the leads. Why else would it coincide so perfectly with the peaks in the waveform? I'll need to make a spice model or something...