Why are they called "resistors"

Archie uses only the finest $ 100 nippers, so of course he wouldn't want test instruments by Keithley.

Only the finest will do for him! LOL

JL > I told you a number of times that the current application is high JL > impedance but not high voltage.

Archie won't let that deter a perfectly good rant.

An interesting question is whether fingerprints which have no measurable effect on resistance at low voltage somehow have more effect at higher voltages. Short of outright arcover, I mean.

Ohms is ohms, seems to me.

John

Sure. How does that relate to fingerprints on resistors?

John

At high voltage sharp points start a "corona" effect, and when other conductive surfaces are near, they start participating. Not so, at low voltage. So yes there can be differences at some voltages.

You think transformer varnish has outgassing problems?

You ain't real bright, johnny.

Absolutely. Sniff a new transformer. You can smell the varnish. In UHV, that means it will outgass massively, approximately forever. Transformers are especially bad because they are big.

There are some special epoxies that can be used, in very small dabs, in UHV systems

Do a little googling on UHV practice. It's interesting.

John

Post a graph.

John

Transformers are also especially bad because there's a huge surface area of varnish, lots of gas traps, and you can't bake them out.

Other than that, they're great. (Of course in real UHV you can't use circuit boards either--besides the porosity, they die during real UHV bakeouts.)

Cheers

Phil Hobbs

Google for corona.

Here's an amplifier that someone (no names, please) manufactured for use in UHV.

ftp://jjlarkin.lmi.net/AMP_UHV.jpg

(The feet at the bottom are mine.)

It's a drilled ceramic slab, all hand-wired on the backside, with bare copper wire. The fet opamps have only radiation cooling, and I roughly calculated that they run about 130C. So there's so much bias current they had to use 1M resistors to ground the inputs. So the Johnson noise is horrific. Overall, about 80 nV/rthz.

Never allow chemists to design electronics.

John

Opinions, no numbers.

John

Is that soft solder and plastic packages? Need a lot of pumping to get it down into 1e-10 torr!

Cheers

Phil Hobbs

I heard numbers like -12!

John

Atmosphere, I believe. Torr, no way.

Cheers

Phil Hobbs

The K is basically a constant-current source into the DUT, and the terminal voltage is measured by a super-hi-Z voltmeter. I think on that range my 1G resistor was seeing 1 volt.

Fingerprints don't conduct at low voltages?

Why would the resistance depend on the terminal voltage, short of corona or arcing? If it does, how much more voltage would be significant, and why?

If the fingerprint resistance isn't nonlinear, the shunting effect is the same at 1 volt and at 1 kilovolt. Ohms is ohms.

The Keithley measurement, and the statement that it didn't visibly change as a function of the fingerprint state of the resistor. That puts the fingerprints at the Tohm range, or higher. Fingerprints may start to matter with Tohm resistors; I did measure a 1T 0805 resistor, mounted on a Pomona plug, and manhandled the plastic of the plug to see if that would change the measurement. It didn't seem to, but the Tohm measurement is delicate and not super stable anyhow. And it's

See? No evidence, just hand waving and opinions.

No, I love quantitative data on component bahavior, especially stuff that's not currently documented.

I posted the Keithley measurement. It didn't change with serious fingerprinting.

I think fingerprints are over-rated as a leakage source. Maybe I'll do some experiments on PC boards. You could too, you know, make some measurements instead of hand waving and pushing Ohms Law around. It's not hard to measure fA with simple home-made gear. I can think of a couple of fun ways to do it.

Come on, make some real measurements.

John

hives

ly wiped

our

ll do it.

e'll see.

tion

Say do they put some 'better' coating on high value resistors these days. I'm using some 1G's and was all worried about finger prints and such... But I notice no finger print problems... at least at the few percent level. (This is also a low voltage application Max of ~10 Volts)

George H.