Of course word should have gotten around by now that we must not overload the inputs of modern DSOs – or any scope with split path input buffer, which are almost all scope designs today (Picoscope 4262 is one of the feew exceptions).
Just as a reference, here are tests with the Siglent SDS2504X HD, similar to what TurboTom did:
First the Overload test. 10 kHz triangle, 6 Vpp amplitude, viewed at 50 mV/div with an additional offset of 1.5V:
SDS2504X HD_Ramp_OVL
Now the offset challenge. A 350 mVpp triangle is fed into channel 2 as reference. Channel 4 gets a copy of that signal with 6 V DC offset. No worries, of course.
SDS2504X HD_Ramp_Offset
Fnally the question, why the heck does everyone always want to overload the oscilloscope input? Why not just use zoom if we want to inspect a signal in more detail?
The last screenshot shows a ten times zoom into the signal (5 mV/div in the zoom window) which has essentially the same effect as the overload method, just with the difference that we can feel better because we stayed away from unspecified territory.
SDS2504X HD_Ramp_Zoom
People will now bitch: “oh, we cannot compare the cheap rigol to an immensely expensive Siglent!”
First, there’s no point in having a scope advertised as 12 bits when its performance is just not there, no matter how cheap it is.
Second, The Siglent SDS1000X HD will perform just as well as the 2000 (except for the bandwidth and MSO option) and it will most likely be very affordable.