Measuring a 60V test point but needing 1Gohm input impedance meter?

[MattPadden]

Hi,

My first post here, so a friendly wave to you all.

I’m doing some repair work on a high-end condenser microphone. One of the test points I need to measure is the DC bias supply to the capsule, stated on the schematic as normally being 60V +/-1V DC. However, the capsule is extremely high impedance and the requirement for the measuring device is for a minimum impedance of 1Gohm to avoid loading the test point.

DMMs like the HP 3478a and many others have suitably high input impedances, but only on the lower voltage ranges up to 2 or 3V or so. My question is whether there is a meter that offers 1Gohm or more at higher ranges, or whether I need to look at an alternative method of measuring the test point, such as a null meter. Clearly I don’t want to try anything outlandish given the very delicate and very expensive nature of the microphone.

Matt

[MattPadden]

For reference, the mic is a Schoeps CMC5 with MK5 capsule. Schematic courtesy of sdiy.org is here: http://www.sdiy.org/oid/mics/Schoeps.gif

The test point in question is U6, located at the cathode end of D1 where it connects to R4 and C6.

[doktor pyta]

You can use external power supply which set on 60V.
Connect - of the PSU to the mic's 0V and use high input impedance DMM as a differential voltmeter.

[Marco]

Just measure with your DMM and multiply by 11/10.

[MattPadden]

Differential voltmeter seems the way to go here. Many thanks.

[bdunham7]

A high-voltage probe might work, the 40KV like the Fluke 80K-40 or BK Precisio PR28A  typically have about a 1G input resistance and are 1000:1 with a 10M DMM.  You should calibrate it against a known 60V source because they aren't all that accurate.

[DaJMasta]

Yeah a high voltage divider probe is a good bet, but the bias supply for the capsule can probably provide a lot more current than it actually needs - if you measure on the upstream side of that 1M resistor you may just read 60V as normal, whereas measuring after it will have its current restricted - I was messing around with a Rode NT-1 and observed this because the final stage of the biasing circuit and the first stage of the input are about the same.

[David Hess]

Electrometers, in contrast to multimeters, usually support high input resitsance to +/-200 volts.  Maybe you can find a used Keithley 617 for a reasonable price.

Your measurement could also be made with a JFET or CMOS operational amplifier configured as a voltage follower with high voltage bootstrap of the supply pins which is how electrometers do it.  Alternatively there are some high voltage JFET and CMOS operational amplifiers which can do it without the additional power supply bootstrap circuits.  But this method requires a high voltage power supply.

In a pinch what I have done in the past is to put a 1 gigohm or 10 gigohm resistor in series with my 1% 10 megohm input multimeter to make an instant x100 or x1000 probe.  That turns the 200 millivolt range into a 20 or 200 volt range.

[shakalnokturn]

If you're only doing this once and the DMM you're using is good enough I'd follow David Hess's cheap solution.
The problem is that many DMM inputs vary from 10 to 11Mohms depending on the range used.
For maximum accuracy you'd want a divider that lands you close to full-scale of a range with a little margin depending on expected maximum voltage.
With a known voltage source and precision series resistor (it can be the one estimated for the actual need >990Mohm) calculate the DMM's input resistance on target range.
Once you know the ratio for that range you can make the measurement.

[MattPadden]

David Hess, thanks for that great idea, and shakalnokturn thanks too for the finetuning tip.