I measured a 5V DC voltage from a lab bench power supply. The probe is a 10x which I properly compensated and the scope was calibrated prior to the measurement. I just forced the trigger to get a reading, but when I had it in AUTO and with a larger timebase the AVG was the same. The ONLY difference between the two measurements is the vertical adjustment. Why would I get such a different reading? Also I double checked the voltage with a multimeter and I got 5 volts exactly. My readings were the same with a properly compensated 1x probe.

First reading: enter image description here

Second reading: enter image description here

  • 3
    \$\begingroup\$ Does your scope change ADC bits at different vertical scales? \$\endgroup\$
    – DKNguyen
    Aug 20, 2020 at 22:50
  • 1
    \$\begingroup\$ I've heard of Rigol having problems with the AC coupling circuit; I wonder if this is related. Does it go back and forth with the vertical adjust, or does it change over an extended time? \$\endgroup\$
    – mbedded
    Aug 20, 2020 at 23:15
  • 1
    \$\begingroup\$ What offset does the channel have with the input shorted? Have you performed a self-calibration? \$\endgroup\$ Aug 21, 2020 at 3:14
  • \$\begingroup\$ @DKNguyen, I actually have no idea... \$\endgroup\$
    – tlubes
    Aug 21, 2020 at 5:22
  • 2
    \$\begingroup\$ 20mV on what scale? MSO5204 specified gain accuracy is +-3%, offset is +-0.1div +-2mV +-1.5% of offset. Scopes are usually not very accurate voltmeters! \$\endgroup\$ Aug 21, 2020 at 6:17

1 Answer 1


I actually texted Rigol and this was their answer:

"The measurement accuracy is driven largely by the range setting. In this case, you would get more accurate by zooming in on the vertical scale even if you moved the offset until the ground was off of the display.

The limitation is the number of bits per division on the display is more or less fixed, so you zoom in vertically to reduce the quantization error.

You can also use the High Resolution Mode in the Acquire menu. This is especially useful for signals like this because it allows the instrument to eliminate quantization error by averaging between the bits. So, you should definitely try this even before you spend time optimizing the range. It may erase much of the difference."


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.