R not R anymore..

Maybe they need drying out? A week's soak at 50'C or something?

Grant.

Both resistors behave the same way on both meters?

John

to

an

e,

ry.

I recently had problems with 10Gohm resistors. I first tried cleaning them with Flux Remover with no effect at all. I then soaked them for a half an hour in tap water and let them air dry for two hours and that worked.

to

an

e,

ry.

Hmm, six months ago it was winter, low humidity? Maybe the 'sense' resistors inside the GR meters need cleaning?* Do both meters read the same amount? Blow inside the meters and see if that changes things? All sorts of sneak paths for a wee bit of current to flow.

George H.

Did you happen to expose the resistors to high voltage (DC especially) in the meantime?

Best regards, Spehro Pefhany

The blue resistor

ftp://jjlarkin.lmi.net/Keithley_1gig.JPG

is a 1G thickfilm, mounted into a Pomona dual banana plug. I have most of the decade steps like this, 1K to 1T. I tried pawing the 1G and 10G resistors, putting fingerprints all over the Pomona and the resistor body, with no noticeable effects.

John

Hmmm...worth trying. Thanks.

Yes.

In a lot of places, tap water is worse than water off the roof. Will try the drying option first, then if need will use DI soak / rinse / soak / rinse / rinse / dry.

All scales are done the same way: Rx is the top of a resistive divider; the bottom resistor is changed for the scale (1Meg..100,000Meg in steps of ten). Never seen this before - but highest resistor used then was a measly

The GR uses 500V and most of that is across Rx during measurement. These resistors are rated for a few KV or so.

The resistor looks just like the ones i have, except the 100G is longer. And, yes,since the 1G resistors are so cheap, i have a number and some got baerhandled with no observable R changes. Where did you get that 1t resistor and how much did it cost?

Use distilled, not de-ionized. water.

ed to

er an

lace,

r dry.

in)

r"

ly

My method is to use the gain of a non inverting amplifier. I use the LMC6001. I put a 10gig resistor in the feedback. The resistor under test is the gain resistor. I inject a 10 mV, 250 milliHertz square wave into the non inverting input. I measure the amplitude with no gain resistor and the amplitude with the resistor to be tested as the gain resistor. You need some known gigaohm resistors to calibrate it. I measured those using resistor bridges, working my way up from a 10 meg I can measure directly with the meter. The main purpose of the LMC6001 jig is to measure photodiode shunt resistance. I don't know how accurate it is but my results are good enough to know when a 10gig is measuring 6gig.

You've gotta be careful around baers.

It's a sample from IMS. An 0805 surface mount! I don't recall the price right now, but they're not expensive. I'll see if I have it somewhere.

John

I threw this together for testing diodes, mostly.

ftp://jjlarkin.lmi.net/99A260A1.JPG

ftp://jjlarkin.lmi.net/99A260A3.JPG

ftp://jjlarkin.lmi.net/99S260A.JPG

The measurement floor seems to be 10 or 20 fA. The low-end measurements take minutes to settle out.

You can plug in two resistors and verify their ratio, so you can build a calibration train all the way from, say, 1M to 1T.

John

Baking the 100Gohm resistor restored its value. But the same treatment for the 10Gohm resistor "helped" only by a factor of two, cleaning procedure helped some more and a "soak: at 500V on the GR made it crawl up to 80 percent of rated value. So i used PGA slosh / rinse / soak / rinse to no avail - then used DI followed by a bake. Now the damn thing reads 20 percent HIGH! Any clues to reason?

Don't they prevent Forest Faers?

mike

y.

DI

I don't know what your resistors look like. Mine are ceramic chips that have a zig-zag conductive path. If you managed to etch the path slightly thinner the resistance would increase.