Measuring extremely high impedance, extremely high voltage sources

I'm working on a geiger-counter, and trying to get an accurage measurement of the high-voltage Tube Bias.

The issue I'm having is that the impedance of the high-voltage source is really, really high, I'd guess somewhere in the range of 100-500 MΩ.

Right now, I have 4 10MΩ resistors in series with my scope probe, which is set at 10X (I've measured the scope input impedance to be 10MΩ). This should produce a effective 50X probe, with a 50MΩ input impedance. Then, since the Input impedance of the prove is still enough to significantly draw-down the high-voltage rail, I'm using a sampling scope to capture the voltage as I touch the scope probe to the high-voltage rail.

Unfortunately, I'm still getting voltages that are far higher then they should be (it should be ~500V. I'm measuring ~1.2 KV). However, as this is a spark-fun product, and the circuit that I have so far studied has been extensively studded with truly WTF inducing design decisions, I am having trouble determining if the incorrect voltage measurement is due to the circuit not behaving as designed, or measurement error on my part.

Are there any other simple ways I can measure a high voltage source with an extremely high source impedance, to cross-check my results?

• Argh. Of course, now that I wrote the question, it occurs to me that I can measure the voltage with various numbers of series resistors (10, 20, 30, 40, 50MΩ) and calculate the source impedance of the HV generator. This should let me extrapolate to the unloaded voltage, I think. I'm not sure if a transformer used in a flyback topology has a linear response to loading. – Connor Wolf Mar 15 '12 at 7:59
• Re my comment about linear loading response - It's horribly, horribly nonlinear. – Connor Wolf Feb 19 '13 at 11:41

Use a variable high voltage power supply to generate a voltage that is about as high as the one you want to measure.

Place a voltmeter between both voltages and regulate the HV power supply until you can't see any difference.

Now both voltages are the same and the voltage source you want to measure isn't loaded.

The HV power supply either should have a calibrated scale control or its impedance should be low enough to measured its voltage directly.

        use reading of voltmeter
to regulate HV power supply
+----------------------+
|                      |
v                                        high impedance
variable HV ----------------(V)--------------- voltage  source
power supply                                   you want to measure

• That's brilliant. And so blindly obvious now that you said it. I don't know if I have a 500V-1KV power supply, though. – Connor Wolf Mar 15 '12 at 10:19
• Mains multiplier gets to 500V easily and 1 kV not to hard. This can be of lower impedance and a meter can help you read it. Note that above 500 VDC you want to ensure that your meter is rated appropriately and ensure that if it dies you don't. – Russell McMahon Mar 15 '12 at 10:37
• Hah. I found the high-voltage power supply I thought I had seen at work. 0-3KV, 0-40 mA. It uses four 8068 vacuum tubes for the output stage! – Connor Wolf Mar 15 '12 at 23:47
• @Russell McMahon - I'd prefer to avoid using doubled mains. Mains connected electronics are another layer of danger I would prefer to avoid, and I don't have an isolation transformer handy. – Connor Wolf Mar 15 '12 at 23:48