High Voltage DC Blocking

[nicknails]

I have a need to measure AC ripple on the output of high voltage DC converters (anything in the range of 1kV-20kV). My thought is to filter the DC with multiple caps, then use an op-amp to amplify the ripple to make it easier to measure on a scope. All I have is a TDS1002B that goes down to 20mV/div. On the lower voltage stuff the ripple is in the 20-40mV range and nearly impossible to see on the scope. My current setup is a 0.01uF in series with a 10M and measure across the resistor. This works but there's a transient spike on the cap. My understanding is that if I put multiple caps in series, the transient spike will be smaller and smaller on successive caps, but I have no idea how to calculate the capacitance or how many caps are needed. The ripple frequency is anywhere from 60k-300kHz.

I'm assuming I'll need separate setups because of the voltage range, which isn't a problem. I have some nice AD822 op-amps that I'll be using, which can run up to +/-15V, so I'm thinking I can get the ripple in the 4-8V range.

Please no rants about high voltage and killing myself   . I've been doing HV DC/DC converters for many years (up to 12kV), just never had a need to measure such a small ripple.

[ebclr]

Why not use a high voltage probe ?

[ovnr]

Transient during capacitor charging: Just clamp it with a TVS or Zener or something to that effect - you've already current-limited it with the series resistor (beware maximum working voltage, etc).

[nicknails]

Why not use a high voltage probe ?

This would require the scope to be set to 1000X, making it 2V/div minimum.

Transient during capacitor charging: Just clamp it with a TVS or Zener or something to that effect - you've already current-limited it with the series resistor (beware maximum working voltage, etc).

I have no experience with TVS diodes; I'll do some research and check that option out. Thanks.