...
The voltage marked on the scope is not the RMS value . Either it is Vpp like on the Siglent models for example or Vp on Rohde Schwarz models i.e. . Here in Europe we have 644 Vpp ( 320Vp) . So you overload your scope even it is 400 V rated .
It is actually worse than you describe.
For just the case of signal measurement, someone might think that a x10 probe will increase the peak input voltage rating. However this only applies with DC coupling; AC coupling allows the capacitor in series with the input to charge to the average DC voltage which would happen if you wanted to see the AC ripple at a high DC potential. They make special x10 and high voltage probes with built in internal shunts to handle this situation but the ones lacking this sure do not advertise it; at most their instructions say something about not using the probe with AC coupling and I've never seen a standard x10 probe with that warning anyway. Now you know why the maximum input voltage for a standard x10 probe is the lower of the probe and oscilloscope input voltage ratings and why a standard x100 "high voltage" probe may only be good to the oscilloscope input rating of 400 volts peak or whatever.
And speaking of input coupling, why did all of those old oscilloscopes include ground coupling anyway and why implement it at the input instead of later in the signal chain where it would be easier? Maybe Rigol did it the right way? Ground coupling precharges the input AC coupling capacitor ... but not on a Rigol. And without precharging, a 400 volt peak input can momentarily become 800 volts peak which is a great test to see if the oscilloscope's input protection circuitry really works ... or not. Have there been any mysterious Rigol input failures?
For floating measurements things are even worse. What is the maximum float voltage? The designed breakdown voltage between the AC line and ground can be surprisingly small. Obviously the oscilloscope has to handle the situation where hot and neutral are reversed but the rating in addition to this can be surprisingly small.
Old oscilloscopes included this specification because the manufacturer's knew oscilloscopes would be used with a floating ground and may even have recommended it for certain measurements. Today there is no reason to include it because nobody will go against the recommendations not to float the oscilloscope's ground, right? RIGHT?
Now where are those included probe schematics so I can see if they use an internal shunt. And where are those oscilloscope schematics so I can see how ground coupling works and what limitations there are on floating measurements? The results are in; the lawyers won and the engineers lost.