Oscilloscope reading voltage when probe is sorted.

[RoadRunner]

I have Rohde und Schwarz RTB2004 with vertical specification attached as per image.

When I turn scope on it is is reading about ~300mV on instrument's DVM when turned on. But it settles down to around 20mV on 2V/div range (Total 20V full scale on screen), after one a minute or two of operation.
All channel have similar behavior. All read few hundred millivolts initially and then slowly settles down under few millivolts.

This value of a constant offset seems to be variable depending on range vertical setting, it is in uV when scope is set to 1V/DIV or less and reading around 500mV when scope is set to 50V/div (500V full scale.

My question is, Is this normal, If yes Which specification of scope defines it. Is this what called "Zero error"?

In Attached image you can see screen shot of scope, all channels are at 0V offset but they are not perfectly over one another. Is this normal?

" alt="" class="bbc_img" />

[Caliaxy]

Not sure about the warm-up time, but 20mV for 2V/div seems under 1LSB (the ADC full range should be wider than the max 20V range that fits on screen, so 1LSB should be < 20,000mV/2^10 bits). I don’t think you can get much better than that.

[RoadRunner]

Thank you all of quick answers, I assume its normal. I was little worried when offset appeared on specific ranges. I guess may be way too large screen of instrument is also a little culprit showing even small changes.

Regards

[tautech]

when probe is sorted.

Do you mean shorted ?
If so, how ?

Let's say you've connected the probe reference (gnd) lead to the probe tip, if so you've just made a fine little RF loop that will collect close by interference and feed it into the scope. From boot some scopes have a cold auto-adjust feature which could null any minor interference out and have measurements return to an apparent normal.
Remember firstly probes are designed to collect and convey signals to the scope even if there is no apparent connection where for sanity checks you'd be better advised to use input Ground coupling to check for any measurement offsets in a scope.

[Damianos]

When testing for input offset, short directly the inputs or remove everything from them (probes, cables). Allow the device to warm up and recheck. Also, after warm up, run the "auto calibration" (or whatever it is called), especially if there is a big temperature change (of the environment) from the previous calibration.