High value resistors: what use are they?

[tggzzz]

If you had resistors in the range 10e9 to 1000e9, what would you use them for? Assume they are "properly mounted" with BNC connectors.

Why? There are some at a local technical college, and they might get rid of them.

[CalMachine]

I generally use high resistance values in conjunction with stable voltage sources to get low current.  Not 100% sure if that is a proper use for them, but it seems to be fairly repeatable and reproducible.  IE; I've used our 732B with our Russian 1GOhm standard to get a fairly accurate and stable 10 nA source.

[VintageNut]

I agree. I have a Fluke 731B that I will use with 10G and 100G resistors to check my KE487 picoammeter at its lowest measurement range. When I have both the 731B and the resistors calibrated, I will use the 731B and the resistors to calibrate the 487.

[daqq]

Popular uses include resistive dividers for feedback for high voltage applications (as in convert a 0V-10kV to a 0V-1V range, whilst not sucking any noteworthy current), biasing, interesting measurements such as http://cds.linear.com/docs/en/lt-journal/LTJournal-V25N2-00-df-LTC6268-GlenBrisebois.pdf .

[Cerebus]

If you had resistors in the range 10e9 to 1000e9, what would you use them for?

Use them in a transresistance amplifier to measure the input leakage current on the high-Z buffer amplifier that I'm currently designing (once it's built, obviously).

[Neilm]

I work with a lot of high voltage so I often use resistors of that order of magnitude or more. I have found myself saying "we only have to measure 100G"

[TimFox]

Be careful with high-voltage applications.
There are high-value high-voltage resistors, that are hefty and usually long.  Caddock makes some nice ones that are several inches long.
There are also high-value resistors such as used in Keithley electrometers that are not rated for high voltage, but are useful in making transresistance amplifiers for sensible voltage levels.

[T3sl4co1l]

Useful?  Sure!  Just don't get your filthy fingerprints all over them!! 

 

Tim

[guenthert]

To set low currents, e.g. the Keyley 617 electrometer's calibration procedure calls for a 100GOhm resistor (and a Teraohmmeter to determine the actual value of that resistor).

[VintageNut]

To set low currents, e.g. the Keyley 617 electrometer's calibration procedure calls for a 100GOhm resistor (and a Teraohmmeter to determine the actual value of that resistor).

I believe that the modern day instrument of choice for measuring a 100G resistor will be a MI bridge. I will find out later this month and report back. I have a DIY 10G mounted in a small enclosure and I have a 100G KE5155 on the way to me, hopefully in time for my visit to the calibration lab. I also have a Fluke 5450A which has resistances up to 100M. I will find out what is used to calibrate the 5450A.

[tggzzz]

Thanks for some interesting replies and links; nighttime reading

 It may be relevant that the BNC connectors are ~6" apart and they are marked as "transfer resistors". The effects of surface leakage in the BNC connectors and cable effects could be "interesting"

Now I have to decide whether to get the resistors. Ah, decisions, decisions.

[Neilm]

 It may be relevant that the BNC connectors are ~6" apart and they are marked as "transfer resistors". The effects of surface leakage in the BNC connectors and cable effects could be "interesting"

Transfer resistors makes it sound as if these are used as transfer standards. That is something that is stable enough that it can be measured with a unit that is known to be calibrated (for instance by a full calibration lab, national standards agency etc), then it can be measured by an instrument that the user wants to calibrate. This will allow the user to "transfer" the calibration of the knowwn instrument to the unknown instrument. (I know that is simplistic). The uncertainties of measurement will increase.

Given two BNC connectors, I would have thought that the inner parts were connected to the resistor and the outer part connected to the guard. I would hope that only one of the connectors was connected to the guard or else there will be a releatively low resistance across the terminals meaning the tester has to supply a lot more current.

[tggzzz]

Well, I'm picking these up next week, all being well. Hopefully that is 10Gohm and the other is 1Tohm.

Now to find a use for them; if not maybe I'll sell them.

[e61_phil]

The connectors look like MHV instead of BNC.

[Conrad Hoffman]

Yes, MHV type. The question I'm surprised nobody has asked is, how do you DIY a decent gigohm and higher resistor? What materials might lend themselves to being the resistance element? Can you simulate a high value resistor to moderate frequencies with a floating battery powered opamp circuit, using easily obtainable values, to an accuracy higher than and actual device?

[tggzzz]

The MHV connector observation leads to a number of thoughts:

• bugger, more expense
• phew, I least I didn't ensure a BNC made a connection
• ah, so that's how large voltages can be applied
• I like this forum: it has interesting topics, interesting responses, and (even though I try) I end up learning more than I could possibly teach others

Thanks!

[grouchobyte]

When you need atto-amps from mega-volts they come in mighty handy
Obviously that comes at a price and some major tradeoffs, including a slight risk of death by electrocution...