Measuring Large resistances

On a sunny day (Thu, 12 Nov 2009 21:12:29 +0100) it happened Kreyen wrote in :

Sure, 100 volt in 100 MOhm = 1 uA. How accurate can you measure current?

Are you trying to measure a resistor or something else? What range variations are you getting? Did you consider surface leakage and other interfering factors? What level of accuracy do you need?

I'm trying to measure 3 resistances to be used for calibration purposes.. I'm getting variations of several percent while I'm looking into a t least a 0.1 percent accuracy level.

Kreyen

Well we got this expensive meter at the Lab at work so I can't see how accurate its current abilities are ... as I'm now at home sweet home . But I'll check tomorrow and let you know the values. It is a HP and a very professional looking device though.

Kreyen

On resistors in those high range, even finger prints on the resistor can affect accuracy. Solution: Clean them with isopropyl alcohol and let them dry, then try measuring them again.

Some meters have a nanoSiemens range which is 1/Resistance, they usually give accurate readings. Other that that a megohmeter that uses high voltage to measure resistance is usually used.

Shaun

I was about to suggest something similar to Jan's but decided to ask for more details. For that level of accuracy, a bridge would be your best bet. With the jury-rigged method, it goes without saying that the supply would have to be stable and accurate to better than 0.1%, with a similar requirement for the measuring instrument.

Perhaps you could try your lab ohmmeter again, making sure that everything is squeaky clean and perfectly dry. In still air. Varying thermal gradients might also have affected the readings.

100M isn't that high, well, depending on what kind of accuracy you're looking for.

How much instability are you seeing? How much above 100M?

LMC6042 has 2fA (typical) input bias current and costs a couple dollars one-off.

A 100M 1% resistor with 50ppm/K tempco runs around five dollars. A 1G 1% resistor with similar tempco is maybe double that.

Most inexpensive DMMs on the lower ranges have very high input impedance so the LMC buffer might not be necessary.

Spehro Pefhany a écrit :

High value resistors are (depends on the model) somewhat unstable. I currently have a batch of Dale 1G/1%. They measure fine with 100V bias which is the datasheet measuring conditions. When measuring them at low voltage, they're all over the place, from +3ish% to +7ish%. Yep, not even grouped... On the contrary I've some Caddok rated

I've got a fluke 189 multimeter that reads 100M ohms just fine. (Ohmite brand.) Since the meter has 10 Meg input impedance they must do some tricks in software. Which suggests that you might measure a good 10M ohm resistor and then try putting the 100M in parallel. See if you get better numbers... of course you will have reduced accuracy, but it might work. At least you might learn if it's the meter or the resistors that are flakey.

George H.

A quick+dirty way to do it is:

Measure a power supply or 9-volt battery with your DVM. Then measure it again with the unknown resistor in series.

Most DVMs have a 10M input resistance on their volts ranges. So Rx and the 10M DVM form a voltage divider, and you can do the math on that. You should be able to measure at least a Gohm or so to a couple of per cent accuracy.

If your DVM has an infinite-impedance option (many bench meters do on their 10-volt ranges) you can make a divider from Rx and a 1G resistor, measure that with the infinite DVM, and do that math. Digikey has 1G resistors.

A Keithley 610C measures up to 1e14 ohms. I got one on ebay for $150.

I also built this:

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

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

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

John

One cute method is to build a current-controlled oscillator, with a CMOS op amp connected as a Schmitt trigger, a voltage reference, a small MOSFET for the reset, and a good integrating capacitor. When you stick a current into the capacitor, the output frequency gives you the current. (No, this isn't the most accurate possible VFC, but it works pretty well for ~1% accuracy or a bit better.)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net

My 17 years old Beckman DM27 xl multimeter has a 2G range on it. It can't do 0.1% accuracy at 2G, of course, but it does a good job for me (I use 1G resistors in HV biases, encounter them in HPGe detectors preamps etc.). Got it at about $100 back then.

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

------------------------------------------------------

Original message:

I just measured some Megaohm resistors with my Fluke 867b. accuracy 0.5%

I measure two - 66 Megaohm resistors at 20% tolerance (not very accurate)

measured 1 - 65.839 Megaohms measured 2 - 69.742 Megaohms

Also I measured one 200 Megaohm resistor that came out of some medical equipment, unknown tolerance

measured 196.46 Megaohms

then readings were stable +/- .1 count in nanoSeimens and repeatable.

You maybe working in an electrically noisy environment, try moving you meter to a different location where there is no extra cables around, maybe a noise filter would help too.

Shaun

"A quick+dirty way to do it is:

Measure a power supply or 9-volt battery with your DVM. Then measure it again with the unknown resistor in series. "

Excellent, I knew there must be some way to use the 10Meg input impedance of the DMM. I'll try it tomorrow on some 1G R's I've got.

George H.

Most hand-held DVMs are 10 Meg input and have a 200mV full scale setting; equivalent to 20nA full scale - allowing one to use reasonably low voltages for deriving large resistance values.

*That* is a color of a different horse..the DVM used would have to have that accuracy, leaves out handhelds!

Fred Bartoli wrote in news:4afc7e1b$0$21967$ snipped-for-privacy@news.free.fr:

Nonsense;they have the same impedance on their lowest range as the others,unless they disconnect the input divider.They don't.

(I did that on a TEK DM501 to make cal measurements for a 577/177 curve tracer.)

also,I checked a Harbor Freight DMM and it was only 1 Megohm input R. (and the first one was way out of calibration,read a 1.5v alkaline cell at

1.9v)

seems like they could just use carbon film deposition on glass substrates and laser-trim them,all machine handled.

--
Jim Yanik
jyanik
at
localnet
dot com

You could build a simple bridge circuit that depends only on the *ratio* between two resistors to determine the unknown. Put a known precision resistor in series with the unknown value as the lower resistance of one leg of the bridge. Then, the other leg using a second precision resistor and decade box and helipot / trim pot of suitable value. Put a null voltmeter between the two leg midpoints and adjust pot to get a null. Then, calculate the ratio with a few sums to determine the unknown. Bridge excitation voltage doesn't need to be precision or stable either.

You should be able to get very high accuracy with this route, depending on the bridge supply and dvm resolution. Null methods resolve to zero load on the unknown leg divider at null...

Regards,

Chris

Maybe *your* handhelds.