Advice for measuring high volts dc

[dnix71]

I have an instrument that puts a static charge on a gas tube and would like to know how to accurately measure the voltage without buying a electrostatic volt meter or oscilloscope. The instrument puts out pico amps at what is supposed to be 900v dc, but I think the HV regulator has failed (it's also a gas tube).

I have a 35 year old Micronta 22-208 with 32 meg ohm constant on the 1kv dc channel, but even that loads down the circuit and shows only about 600vdc. My digital meters are not high quality but they all range off scale (well above 1kv) before dropping back to about 35 v when trying to measure hv to ground.

I have a bunch of 100v zeners on order to fix things if the HV side is too high, but I would like to be accurate about it. I opened the 22-208 and the 32meg ohm resistor is just behind the probe plug dedicated to up to 1kv dc. Adding more resistance in series made the voltage read drop. The 22-208 is a dual FET design and that may be why adding more resistance to one side doesn't help. The dual FET is actually visible on a small board next to the battery as a 6 pin can (looks like maybe a 30809).

I also tried 150 meg ohms shorting out the supply and measuring the voltage across a 4.7kohm resistor that is inline the probe power feed to protect the probe. That doesn't get me any better readings. I don't have any gig ohm resistors.

Anyone here got a gig ohm axial they would sell me? I made the 150 meg ohm by strings 10's together and verifying the string with a 200 meg ohm meter.
Big 1kv electrolytic caps are rare and not entirely safe. I thought about letting the instrument charge a big cap and read that.

[Tritium]

http://www.google.com/products/catalog?q=fluke+HV+probe&oe=utf-8&rls=org.mozilla:en-US:official&client=firefox-a&um=1&ie=UTF-8&tbm=shop&cid=11108751438087845567&sa=X&ei=DuMLTvPvIuTw0gGI2JGNDg&ved=0CG0Q8gIwBA

Not real high but not cheap either. I used these while working in the Semiconductor industry.

Thurmond

[joestue]

how accurate do you want this?

For about as much work as its going to be to get a 1% error at 900 volts and pico amps with resistive dividers you could throw together a 900 volt supply and with two cheap dmm's you can get to 1%
If you can get the differential voltage down to 1 volt then that's about a gigaohm with a common 10megohm dmm.

[madlabs]

If it is a meter you love, then a real high voltage probe is the way to go. I used to play with a lot of HV and have several probes, one of them even a gas filled one that I can use to 20kV. If it is a meter you don't love and you dont need real accuracy, then try the voltage divider and see what you get. Use 1/4 or 1/2 watt resistors so that the arc doesn't jump the resistors. Did that when I was young measuring a 12kV transformer. Was a neat show but ended badly. Next one I made was under oil. Be careful.

For static stuff, getting a good measurement is hard, I'm sure most HV probes will load it down some.

Jonathan

[letERblow]

I work with instruments that have 900 V low current supply, geiger counter tubes. We made up some probes for the field techs to use with their DVMs that did not load the supply.

PM me your address at     don2424 at bridge mail dot com     and I will send you a 1Gig ohm resistor and schematic if you're interested. also have some 2G ohm.

Don

[dnix71]

letERblow I'm sending you a PM. To everyone else, the addon probe looks like a good idea, but how does it work? I have Harbor Freight Cen-Tech 95683 meter with a dedicated 2G/20M probe ohm port but I can't find the probe that is supposed to go with that meter.

Also, would having 32M ohm throw off the Fluke probe listed above, or is the 10M ohm number listed just the minimum recommended? I don't mind spending $100 for a nice add on probe.

I also have a Commercial Electric/Mastec/Sinometer MS8260A digital meter that has a 1kv scale but it also loads down the circuit too much.

The Micronta is a museum piece now, but there are few meters anywhere with even 32M ohm built-in. Heathkit made this one http://www.heathkit-museum.com/test/hvmmm-1.shtml that had 20k ohms per volt on the dc scale. I could never afford Heathkit, though. About 28 years ago when I moved down here there was still a Heathkit store on Broward Blvd in Plantation. The Layfayette Electronics in Plantation on SR7 is still in business. Motorola has a plant in Plantation, so there are people here with old school skillz.

[joestue]

the fluke 80K-6 is just a 75 Mohm and a 75K ohm resistor which form a 1000:1 voltage divider.

http://www.powermeterstore.com/crm_uploads/fluke_80k-6_high_voltage_probe_instruction_sheet.pdf

[dnix71]

So the Fluke won't work. The 75M ohm is not nearly enough. I need 10 times that. I would also have to replace the 32M ohm internal with a 10M, which wouldn't be that hard.

[boB]

So the Fluke won't work. The 75M ohm is not nearly enough. I need 10 times that. I would also have to replace the 32M ohm internal with a 10M, which wouldn't be that hard.

Maybe you can calibrate it at a lower voltage and just use the input impedance of the Fluke meter as part of the divider...  As long as it doesn't change
when you go up in voltage.

boB

[dnix71]

I tried 150 m ohms and even that wasn't enough. Has anyone here used an ACL 300B? They claim 50v minimum accuracy non-contact, which is 5%, at 1/2" from the test piece.

I can get one for $100 used with a 7 day return in case it doesn't work. New they are $400+

[dnix71]

I got it. I had a Duh! moment at work today. Radio Shack sells 10M ohm resistors. 5 packs and 10 in a pack of assorted sizes. I made a string of 150M and a string of 100M and put my $40 Commercial Electric/Mastech/Sinometer 8260A from Home Depot in the middle and set the scale to 20ua. When I turned the instrument on the scale read 3.75 ua. 3.75 x 250 = 938v which is right where it should be.

Using only the 150M string the voltage sags to 875, but that's still within range and no probe should ever load it down that much anyway.

[letERblow]

DINX71

In my last post for the 1G resistor i miss typed the email address.
should be don2424 at bridgemail dot com

great to see you may have solved your problem. The resistor offer still stands.

don