Accurate Capacitance Meter Reccomendations

[lennybogzy]

I need a very accurate capacitance meter with capabilities down to + .01 puff. Any reccomendations?

 

[berkeman]

I need a very accurate capacitance meter with capabilities down to + .01 puff. Any reccomendations?

What is the application? That's mighty fine accuracy, and it will likely require relatively frequent recalibration/compensation.

Have you looked at the Fluke and Agilent websites?

 

[lennybogzy]

We just need something good. The fluke multimeteres arent cutting it. On second thought, i think I'll settle for 0.1pf. The difference in price for another 10th is astounding.

http://www.teknetelectronics.com/Search.asp?p_ID=24662&pDo=DETAIL

Basically I want to be able to tell the difference between a 1pf and a 2pf and I'm afraid that with the kind of accuracy that I'm seeing in meters at that level, it will be difficult.

 

[berkeman]

We just need something good. The fluke multimeteres arent cutting it. On second thought, i think I'll settle for 0.1pf. The difference in price for another 10th is astounding.

http://www.teknetelectronics.com/Search.asp?p_ID=24662&pDo=DETAIL

Basically I want to be able to tell the difference between a 1pf and a 2pf and I'm afraid that with the kind of accuracy that I'm seeing in meters at that level, it will be difficult.

Even a basic meter will be able to tell you the difference between 1pF and 2pF, and many offer 0.1pF precision.

The key for you is to have a stable setup, with short connection cables, and characterize the parasitic capacitance of the setup. You're not going to hand-hold cables and get 0.1pF accuracy.

What are you measuring? How short can you make the cables? Do you have some reference capacitances that you can use to prove out your setup?

 

[lennybogzy]

I'm reading the accuracy for the U1701 meter and it says, for example, with a range of 1000pF and a resolution of 0.1pF the accuracy is (1% + 10 counts of least significant digit). What does that mean?

For example with a reading of 2.0pF, or 2.2pF what are my margins?

 

[mdjensen22]

Either one of these should work...
http://www.bkprecision.com/products/model/890B/dual-display-capacitance-meter-to-50-mf.html
or
http://www.bkprecision.com/products/model/830B/dual-display-capacitance-meter-to-200-mf.html

As Berkman said, the cabling and any stray capacitance needs to be solid in order to get the accuracy you mentioned.

There are often guard traces that can be used to help with this if needed. Many capacitance meters you just plug the component directly into the meter (no cabling). This should be the most ideal approach.

 

[lennybogzy]

a lot of the caps i need to measure are RF caps and consequently not axial lead. The only thing that concerns me is the tolerance, the margin of error. I want this as low as possible to get the most accurate reading.

I have a fluke 289 DMM and this thing can barely measure 0.1nF

 

[waht]

A 1.0 puff capacitor as measured at 1 KHz by a standard LCR meter could be entirely different at 100 MHz for example.

So keep in mind that any standard LCR meter will not be able to capture the true behavior of capacitors in the RF range because of their test frequency usually ranges from 120 Hz to 1 MHz.

If accuracy and precision is required, then consider an Impedance Analyzer:

http://www.home.agilent.com/agilent/product.jspx?nid=-33831.536880679.00&cc=US&lc=eng

The device will measure the true capacitance as it is at any RF frequency.

 

[lennybogzy]

Two questions:

1) Are you saying capacitance changes with frequency?

2) how can that meter measure true capacitance at "any RF" if it only synthesizes 1 MHz to 3 GHz?

 

[f95toli]

Two questions:

1) Are you saying capacitance changes with frequency?

2) how can that meter measure true capacitance at "any RF" if it only synthesizes 1 MHz to 3 GHz?

Capacitance (as well as inductance) is a lumped element property; it is a valid approximation if your component is much smaller (lets say 1/10th) than the wavelength you are are working at but this is often NOT the case at RF frequencies.
Now, one "solution" which often works quite well to this is to model the component as a small circuit and take things like stray inductance etc into account.
Another solution is to simply say that the effective capacitance is frequency dependent.

The answer to 2 is that it can't; but a measurement between 1 MHz and 3 GHz will obviously give you the capacitance in that range; and should allow you to extrapolate to somewhat higher values. That said, for anything over a few GHz you should probably use a VNA instead and simply use S matrices instead of lumped element values anyway.
Manufacturers of RF components will often supply you with specs that can be used to simulate the component in e.g. Microwave Office or similar.