Power supply offset - too high expectations?

[arg]

This is my first post here, and I'll like to express my gratitude to all the contributors at this forum. I've been lurking around here for some time now and this is a great place to learn.

I just got a new bench power supply and I'm not sure if it is the power supply or my expectations that needs to be calibrated. The power supply is a GW Instek gpp 4323. The power supply has separate soft power off for each channel, but with power off on the main channel I still measure 7 mV. I have measured with two different dmms, and when reversing the multimeter probes the result is -7mV. So my questions are: Is this normal? In general, does soft power off only mean "set to 0V"? Is 7mV acceptable for "power off", or am I expecting too much?

[soldar]

With what load do you measure that voltage?

[arg]

So the 7 mV was measured without any load. Just measuring over a single resistance I get these numbers from my ancient Fluke 79ii (voltage, resolution 1 mV) and Uni-T 61 (current):

All measurements with channel turned off.

8 kOhm, 7 mV, 1 muA (micro Amp)
1000 Ohm, 5 mV, 5 muA
100 Ohm, 2 mV, 12 muA
12 ohm, 1 mV, 13 muA
1 ohm, 0 mV, 10 muA

When the channel is set to ON, but output set to 0 V, for 100 Ohm, I measure 1 mV, 3 muA which is different from above when channel output is set OFF.

I probably shouldn't worry, but I just replaced a cheap Korad 3005 clone that had 70V spikes when hardpowering off even if soft power was already off. I have not checked this new supply for that kind of behaviour yet.
   

[exe]

I think when it's "off" they just set voltage to zero, which is often not really zero. 7mV looks absolutely safe, also because the residuo current is very small. May be it's just a leakage current of a pass element.

Some power supplies, to get really down to zero or even a bit below, have a current sink inside that pulls down output to internal negative rail. In this case output can be even negative too.

[JustMeHere]

I'm seeing the same thing on my TP3005P.  It could be capacivitve coupling.  It's 7 mV too

[arg]

Thanks for your replies, I'm less worried now.

[ArthurDent]

My two GW Instek power supplies are pretty close, but not exact to the digit as you can see in the photo. One thing to keep in mind is that your supply only has local sensing at the output terminals so the voltage reaching your load will be less by an amount depending on lead length, wire size, and current drawn. This error will probably be far greater than the 7 millivolts you're worried about.

That's why the PSM-2010 shown in my photo has the ability to do remote sensing to compensate for voltage drops on leads, in connectors, etc., by sensing the voltage right at the load in critical circuits using two additional leads that don't carry any real current. I really don't think it's something you should lose sleep over. 

Here's a good explanation of what I've described.

https://community.keysight.com/community/keysight-blogs/general-electronics-measurement/blog/2018/01/25/remote-sensing-is-important-for-your-power-supply