[TCML] ocilloscope: Yep, I blew mine, too

[EVP]

    Remember that such an HV resistor is useful for extending the range 
of a DC voltmeter but isn't necessarily good for high frequencies unless 
it is capacitance compensated with a parallel capacitance.  For flat 
frequency response the RC time constant of the HV resistor/capacitor 
combination needs to be equal to that of the scope input end.  If you 
make a 100:1 probe with 100 meg series resistance and the scope input 
resistance is say 1 meg with 20 uufd capacitance then the effective 
capacitance across the 100 meg resistor needs to be 20 uufd x 1 meg/100 
meg = .02 uufd.  Not the same as zero at all.  One way to check the HF 
response is to feed a square wave into the divider and look at the rise 
time of the output waveform.  Most scopes have such a square wave output 
for that specific purpose.  The voltage rating of the compensating 
capacitor needs to be equal to the maximum input voltage of course and 
one way to accomplish that is to use a number of larger capacitors in 
series across the multiplier resistors.  If they are too large 
[overcompensated with overshoot on rise] it's easy to add additional 
capacitance across the scope input end to get flat response - much 
easier than trying to adjust the multiplier capacitors.

Ed

Bill Noble wrote:

> 1. you will find that a vacuum tube scope is much more tolerant of 
> overvoltages with low current
> 2. you can make an HV resistor by putting many LV resistors in series 
> - make your 100 meg (10 is way too low) resistor out of 100 1 meg 
> resistors in series, put it in a plastic sleeve and fill sleeve with 
> HV grease