Rigol DS/MSO2xx2A-S -- How to Calibrate Generator Offset?

[TurboTom]

I've got a "pimped" Rigol MSO2072A-S and lately found the AWG to have some annoying DC offset (several tens of millivolts). Of course, I could compenste it by adding a DC offset via the controls but that's not the way it should be. Running the scope's internal calibration wouldn't help and I didn't find any information how to calibrate the AWG section of the scope. Does anybody know if there's a way to calibrate the AWG with the scope automatically -- maybe by linking the AWG output(s) to the scope's inputs?

Even though the usefulness of the internal AWG is limited, I'ld appreciate any input on that issue. I'ld just like to have the instrument running as it was meant to be...

Thanks and all the best,
Thomas

[TurboTom]

Peter-
thanks for this information. Strange enough, any inquiry that's referenced to a certain channel won't work on my "pimped" MSO2072A-S. I found exactly the same commands from your list in the memory dump of my scope (what else should I expect) but apparently they are supposed to be used differently compared to your DS1xx4Z-S. Or is there a neccessity to enable calibration mode first on my model?

I actually checked the inaccuracies at the output of the generator module and they are substantial and actually behave quite strange. First of all, there's a substantial gain and offset error when increasing output amplitude above 900mV (DC output). At that level, the hardware is reconfigured and probably another gain stage is inserted into the signal path (audible switching of a relay).
The worst inaccuracies I get when I generate an offset square wave with eiter the upper or the lower level at 0V (i.e. 2Vpp square with either +1V or -1V offset). If the offset is positive, the 0V level is a tiny bit above 0, maybe 10 to 20mV. I wouldn't worry about that. But when the offset is negative, the 0V level is more than 200mV below zero! And funnily both channels behave very similar.

To me it almost seems that the AFG has lost its calibration more or less completely. But what might have caused that? Maybe while performing a firmware update? Could there be a (hidden) self calibration function for the generator as well where the outputs are routed to the scope section's inputs? But how should this be invoked?

I guess I might just ring up Rigol Europe - I had been in touch with them before and they always were very friendly and helpful - maybe they can give me a hint. Yet I'm not to sure if I want to send my highly modified scope there...

Cheers,
Tom