Products > Test Equipment
Best CHEAP LCR meter for small capacitors and inductors?
<< < (13/13)
mawyatt:
Interesting, even Hackaday refers to your XJW01 "Calibration" with the Kelvin leads attached, see snip below.

Calibrate the meter
It's time for calibration. It's not a super-obvious procedure, but it is fairly easy.

First, measure the included resistors as accurately as possible. There should be a 40R, 1K, 3K, 9K, 10K and 100K resistors. Go nuts with 50 power line cycles, averaging, 6,5 digits, 4-wire measurements and all that. We will use the measurements you've done to calibrate the meter to.

Once you've measured the resistors to an accurately enough degree, it's time to start the calibration procedure:

Turn the unit on
Perform a short test lead calibration. Clamp the Kelvin clips together and press the "Zero" button
Perform an open test lead calibration. Disconnect the kelvin clips and press the "Zero" button
Hold down MENU for approximately 3 seconds, this will bring up the options menu
Press RNG to enter calibration mode, the display should now show "Z0"
For the calibration, we will adjust the instrument to exactly match the previously measured resistor values (as well as some other parameters). Use the CLR (up) and SP/ESR buttons (down) to adjust the values. Use FREQ to go to the next parameter and QTY to go back.
Set the values of the different parameters as follows:
Step   Target Value   Notes
Z0   Z=0   Short the Kelvin clips
Z1   Z=0   Short the Kelvin clips
Z2   Z=0   Short the Kelvin clips
R1   40R resistor
Connect the resistor and try to match what you previously measured with your multimeter
R1X   θ=0   Leave the resistor connected
R2   1K resistor
Connect the resistor and try to match what you previously measured with your multimeter
R2X   θ=0   Leave the resistor connected
R3   10K resistor   Connect the resistor and try to match what you previously measured with your multimeter
R3X   0 pF   Leave the resistor connected
R4   100K resistor
Connect the resistor and try to match what you previously measured with your multimeter
R4b   100K resistor
Connect the resistor and try to match what you previously measured with your multimeter
R4X   0 pF   Leave the resistor connected
G1   3K resistor
Connect the resistor and try to match what you previously measured with your multimeter
G1X   0 pf
Leave the resistor connected
G2   9K resistor
Connect the resistor and try to match what you previously measured with your multimeter
G2b   9K resistor
Connect the resistor and try to match what you previously measured with your multimeter
G2X   0 pf
Leave the resistor connected
PhX   θ=0   Connect the 1K resistor
After calibration, make sure to save your changes with the Hand/Auto button!

https://hackaday.io/project/175641/instructions

Not trying to divert nor get into a useless argument as so often happens, nor play word games.

Just curious how one does as you say "a Calibration to the Instrument BNC terminals". Since these instruments are 4 terminal measurements, how does one Calibrate such.

How do you connect a reference resistor or capacitor (or reference short) directly to the 4 BNC terminals without some sort of cables/leads, or does the instrument have a "built-in" calibration means which includes input switching relays and Reference components inside the instrument?

If one has a known set of Calibration cables/leads which are well known and characterized, I can see how these could be utilized with the external Reference components and the lead effects "backed out" to the BNC connectors, sort of like with a VNA.

If this is true then I can see how one could do an Open/Short Compensation with whatever cables/leads or fixture that is attached to the 4 BNC instrument connectors and Compensate for such by means of mathematically removing these external effects.

BTW we do agree with the basic HP terminology regarding LCR Calibration and such, but when the 2 lab grade instruments we posses utilize completely different terminology, as does the Handheld LCR meter, and all different as shown, then the term "LCR Calibration" becomes a much more broad coverage term.

Our best guess to all this is the XJW01 establishes it's internal Calibration coefficients with the supplied Kelvin Clips by means of measuring various external reference components as shown in the Hackaday article, and does not mathematically remove these Kelvin clip effects to move back to the BNC terminals, please advise if this assumption is incorrect. Whereas a proper lab grade instrument utilizes a special "fixture" of know effects to establish a measurement reference plane at the BNC terminals at the instrument Calibration and Verification (which requires such with lab grade certified reference fixture and components), and then the Calibration/Correction/Compensation/Normalization or whatever we call it, the Short and Open are performed with the intended usage cables/leads or fixture. The instrument will mathematically remove the effects of this Open/Short procedure, and in some cases will also take into account the effects of the measurement cable length.

Anyway, not trying to get into a useless argument here, just trying to understand how this instrument achieves said results. If we had one of these XJW01 instruments then we would know how all this works, and with the relatively low cost might be worth acquiring just to learn something :)

Best,




2N3055:
I'm with Mike here..

I look at impedance meter as I would to low frequency VNA.

In VNA parlance calibration is, well, calibration and compensation of all (possible) parasitics out. Math behind it can be quite complicated as there are many possible math models the can be used. It establishes the scale and accuracy of measurements. And since it is very dependent on parasitics of interconnect and fixtures you use to measure DUT, you calibrate with fixture in place...

With RCL meter open/short calibration actually calibrates zero and open (infinite impedance) values at end of fixture. It is calibration, maybe not full one but at least those two pins are established. Fact that LCR meter has internal adjustments of circuitry is internal to how hardware functions... If that doesn't work well, adjustments need to be made to device... So I would call that repair or adjustments in this context.

This is how word calibration is usually used in this context. 
Metrology guys would tell you calibration is just a check what values your device measures when measuring reference values.... And no adjustment is included with that word...
precaud:

--- Quote from: mawyatt on February 12, 2023, 04:21:34 pm ---Just curious how one does as you say "a Calibration to the Instrument BNC terminals". Since these instruments are 4 terminal measurements, how does one Calibrate such.

How do you connect a reference resistor or capacitor (or reference short) directly to the 4 BNC terminals without some sort of cables/leads, or does the instrument have a "built-in" calibration means which includes input switching relays and Reference components inside the instrument?
--- End quote ---

Using 4-terminal standards with BNC's, like HP 16074A and others, that attach directly to the outputs, and are characterized at the BNC's. Your lab-grade LCR meters were undoubtedly calibrated that way.  I'm pretty sure there have been threads about making DIY versions of the 16074A.

If you don't have access to those (few do), then use the most-compact fixture you can find to calibrate the instrument.

THEN, your Open/Short routines will more accurately compensate for other fixtures and leads.

If you calibrate an instrument at the end of test leads, and then use a fixture, or shorter leads, the Open/Short vector math can not accurately correct for a negative impedance difference.

Impedance via a VNA is not the same as 4-terminal balanced bridge.

Again, I highly recommend the HP/Agilent Impedance Measurement Handbook. It doesn't matter what brand LCR meter you have, they all use the same basic technique developed by Henry Hall for General Radio decades ago.

The procedure given for the XJW01 is for end-users using items supplied with the instrument. The instructions are fine except for directing users to use the supplied Kelvin clips. If you do that, then you will have increasing uncertainty/inaccuracy at the impedance extremes. Which is exactly the OP's concern.

Think about what is going on in calibration versus compensation (no matter what words you choose for them) and it will become clear to you. Calibration stores internal offsets for known impedance values (typically zero and near-full-scale per range) mostly to correct for manufacturing tolerances. In the "old days" trimmer pots were adjusted for it. These are simple offets added or subtracted from the values that the A/D converters are delivering, and set the vector starting points. Open/Short compensation then "extends" these points to the end of your test leads, using vector (real & imaginary) math.

In an ironic note, I borked my XJW01 yesterday by inadvertantly connecting a charged 'lytic to it. The LCR meter I use the most, an HP 4276A, is totally fine with charged caps and I sometimes forget that other meters aren't. Sigh.
scopeman:
You probably just blew the protection diodes. Usually the meter comes with spares. IIRC they are just 1N4148 MELF Diodes. Have fun taking it apart to replace them!

Sam
W3OHM
precaud:
Yes, it was an easy fix, easy to access, too.
Navigation
Message Index
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod