1.5 year-old topic, but the unit I got this week seems to have the same problem. Some
combinations of sensitivity (volts/div) and frequency (time/div) seem to be worse than others: just now, 1V/div and 100ns/div was the worst combination - see picture. Looking that at picture you'd think it's a ~200mV noise, but turning the sensitivity to 50mV/div reduces the noise to ~20mV. Almost as if the noise was a fraction of a div, whatever the div unit was. No specific frequency, it can happen with any time/div setting. Sometimes, turning the unit off and on again almost completely clears the noise, then eventually it comes back. Only happens to channel 2 and regardless of probes being attached, which makes me hesitant to believe it is external interference.
What annoyed me the most about this unit though, is the loud humming noise when the screen brightness is reduced. I find 100% brightness a tad too bright (and probably unkind to the battery), but I cannot turn the brightness down as I can't stand the noise.

And I wish it had the feature of summing and subtracting the two channels -- it doesn't.
Those are my complaints, yet I do
not regret purchasing it! As an occasional hobbyist and given the low price point, its compact size and built-in battery (that allowed me to directly measure mains just for fun, aided by a couple of resistors to simulate a 100X probe), I am not aware of a better value alternative.