Accuracy of LCR meter?

[Faringdon]

Hi,
Do you agree this LCR meter  cannot realistically measure 37uH or so?...because its accuracy  is stated as 2% of 2mH = 40uH.

LCR meter
https://www.farnell.com/datasheets/1955257.pdf

[rs20]

Yeah, I suspect it'd be fairly useless for that, but I think your reasoning is wrong.

37uH would get rendered as "0.037" mH, except it would it would be off by up to (2% of 37 uH (negligible; ~ 1uH) PLUS 8 counts (where the least significant digit here is uH, so ~ 8uH)), so anywhere from "0.028" to "0.046" mH would be reported in the case of a nominal 37uH inductor. In other words, about +/- 25% error on the inductance value. So pretty hopeless.

[Martin72]

Hi,

because its accuracy  is stated as 2% of 2mH

According to the sheet you´ve linked, there is no remark that the 2mH range got 2%.
It´s only stated, the BEST accuracy will be 2%, not on which range.
Apart from this, 2mH as lowest range makes it useless.
It should have 200µH range at least:

https://www.amazon.de/Kapazit%C3%A4ts-Induktivit%C3%A4ts-Pr%C3%BCfvorrichtung-Meter-Selbstentladung-Strecken-Anzeige/dp/B07BR2N7GB/ref=asc_df_B07BR2N7GB/?tag=googshopde-21&linkCode=df0&hvadid=266451026830&hvpos=&hvnetw=g&hvrand=7066970650968613726&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9043666&hvtargid=pla-444110103102&th=1

My recommendation :
Save a little bit more money and buy a DER DE5000.
Starting at appx 100 bucks, depending on the accessoires, there´s no better choice for the money, imho.

[bd139]

Buy Peak LCR45 from CPC. That will measure reliably down to 0.3uH which is outside it’s specified capability. I’ve tested one against a fixture with a signal generator and scope and a 6.5uH inductor was measure at 6.6uH.

I wouldn’t buy the DER. There’s no warranty service or calibration options unlike the Peak.

[Martin72]

Warranty: Who cares about it, when buying ultra-lowcost stuff like DER or peak.
Calibration: It´s the same, open and shorted probes.
We got LCR40 and LCR45 at work for quick checking, can do a comparison between them and the DER, if it´s interesting for anyone.

[bd139]

Please do. Interested in results.

As for warranty I do. Stuff on the cheap end likes to drop dead.

[TimFox]

The DER unit is cost-effective.  I have checked my unit's calibration against secondary lab standards, and was amazed at the accuracy.
You do have to run through the calibration routine with the test connections you intend to use for the measurement, and you do have to be careful not to connect a charged capacitor to the terminals.
My only complaint is that if you measure a "capacitor" whose actual reactance at the (high) test frequency is inductive (above SR), it does not return a negative value, so you don't know that it is inductive.
You need to use the phase angle measurements to check on the sign of the imaginary component.
In doubtful cases, I would prefer a real/imaginary component readfout, but the inexpensive DER does not support that.

[mawyatt]

2nd the DE-5000, really good value. Here's a few comparisons we did some time ago with a Tonghui TH2830 lab type LCR meter.

Reading/ESR and TH2830 with supplied Kelvin Clips and DE-5000 with supplied short Alligator clips except as noted.

Type                                 TH2830                              DE-5000
10uH @100KHz               9.39761/0.1754                  9.366/0.13
470uH @100KHz             449.188/3.8841                 449.8/5.1
470uF @100Hz               439.802/0.13115                440.6/0.1
100uF @ 1KHz                91.6142/0.43874                91.72/0.43
1uF @ 1KHz                   1.00896/0.0181                 1.0097//0.02
0.1uF @ 100KHz            99.5913nF/0.289                 99.52nF/0.28
2.2nF @ 100KHz            2.18710/11.435                  2.193/10.72

#12 Wire   @ 100KHz    51.123nH/0.00206               43nH/0.00         51nH/0.0 with special custom Kelvin Clips

0.1 ohm   @  DCR             0.101054                           0.11
1 ohm       @ DCR             1.00540                             1.01
10 ohm     @ DCR             9.97630                             9.97
100 ohm   @ DCR            99.7219                             99.61
1K ohm     @ DCR            998.001                             998.0

Judge for yourself how good the DE-5000 behaves!!

Best,

[bicycleguy]

I'm really  liking my Shannon LCR Tweezers.

There discussed here:
https://www.eevblog.com/forum/testgear/design-a-new-precision-lcr-tweezers/msg3473886/#msg3473886

Kind of new, (device and for me), but 3 firmware updates since I bought  .  I've been to busy to study the specs, but in use it matches my 121GW.  Haven't bothered to calibrate it yet.  Repeatedly measures a 10uH inductor as 9.966uH.

[tautech]

I'm really  liking my Shannon LCR Tweezers.

There discussed here:
https://www.eevblog.com/forum/testgear/design-a-new-precision-lcr-tweezers/msg3473886/#msg3473886

Kind of new, (device and for me), but 3 firmware updates since I bought  .  I've been to busy to study the specs, but in use it matches my 121GW.  Haven't bothered to calibrate it yet.  Repeatedly measures a 10uH inductor as 9.966uH.

100% with you on smart tweezers and been using them for 15 years on all TH and SMD work.
Good quality ones and it seems Shannon's are, negate the need for other LCR equipment unless you're chasing high precision in which case you'd be better off with a mains powered bench LCR meter.

[abdulbadii]

How SMD C can be known onboard when there often possibility of lowering ohm parallels
need to separate a pin and  so ordinary probe should be used finely anyway

[bd139]

Indeed. To get an accurate impedance measurement of a part it needs to be in a fixture or you're getting an impedance measurement of the network. That might be good enough but it's not good enough for most diagnostics hence why I don't own SMD tweezers. Lifting a leg or the whole SMD part is advisable and you can usually make do without the tweezers then.

If you must have tweezers, Peak sell 'em: https://www.peakelec.co.uk/acatalog/smd03m-lcr-and-esr-test-tweezers.html

[fyl2022]

you may try a benchtop lcr meter, MATRIX LCR METER MCR-5200 looks nice, it has good accuracy and quality, they sell on amazon, here is the buying link:
https://www.amazon.com/MATRIX-apacitance-Insulation-Resistance-%EF%BC%8840Hz-200kHz%EF%BC%89/dp/B07ZCJ1319/ref=sr_1_1?crid=GFLDWRR2RHG2&keywords=MCR-5200&qid=1656574710&sprefix=mcr-5200%2Caps%2C665&sr=8-1

[BravoV]

Accuracy for LCR meters that use Cyrustek ES51919 and ES51920 based chip sets are pretty good. Example : DE-5000, MS-5308 or UNI-UT612 and etc.

For calibration there was an info shared by other fellow here in this forum ->

For those who would like to perform unit calibration, here is the information and it is also works on Mastech MS5308, UNI-UT612 or whatever using Cyrustek ES51919 and ES51920 ship set.

For DE-5000, adjust DC 500mV between TP1 and TP2.

For others, adjust DC 500mV between pin 26 and pin 28 on ES52920.

Using my cheapo MS-5308 compared with probably 5 digits priced  Wayne-Kerr 6440B LCR meter  ...

[rs20]

Hi,

because its accuracy  is stated as 2% of 2mH

According to the sheet you´ve linked, there is no remark that the 2mH range got 2%.
It´s only stated, the BEST accuracy will be 2%, not on which range.

I... I realize that this is what a literal interpretation of the column title might imply, but surely this is wrong and the column is just poorly labelled. There's no way that ±(2%+8) figure applies to only one range; and that the manufacturer is making absolutely no claim whatsoever about the performance in the other ranges?

[AVGresponding]

Hi,

because its accuracy  is stated as 2% of 2mH

According to the sheet you´ve linked, there is no remark that the 2mH range got 2%.
It´s only stated, the BEST accuracy will be 2%, not on which range.

I... I realize that this is what a literal interpretation of the column title might imply, but surely this is wrong and the column is just poorly labelled. There's no way that ±(2%+8) figure applies to only one range; and that the manufacturer is making absolutely no claim whatsoever about the performance in the other ranges?

Wouldn't surprise me if the accuracy dropped off to 5% or so, and maybe near 10% at the top ranges.



As BravoV's post shows, meters from the well known brands like DER, Uni-T, Peak etc are all good enough for a hobbyist. There are always gotchas ofc; high voltage performance can only be realistically tested using high voltages for example, such as for leakage current/breakdown on caps.

[BravoV]

There are always gotchas ofc; high voltage performance can only be realistically tested using high voltages for example, such as for leakage current/breakdown on caps.

Yes, the only limitation using this kind of handheld LCR meter is DC high voltage biased measurement, eg: measure a capacitor at their rated high voltage as its characteristics may be different compared to the low voltage measurement signal generated from the handheld LCR meter as its battery powered. Imo, this is not super uber critical for hobbyist.

But for enthusiast, for capacitor leakage measurement, its "relatively" easy with only a DMM with 10M Ohm input and an adjustable DC power supply.

Example -> Quick n dirty method to measure cap leakage

... and you don't need an expensive DMM like this example, just need the one with 10M Ohm input. It worked very well. 

[ebastler]

you may try a benchtop lcr meter, MATRIX LCR METER MCR-5200 looks nice, it has good accuracy and quality, they sell on amazon, here is the buying link:

Any chance you are working for the seller? Right, thought so. Thank you for the impartial recommendation. 

[mawyatt]

Monitoring capacitance vs voltage is an interesting experiment to perform with LCR meters. The high K dielectrics used in common ceramic capacitors cause a dramatic reduction in capacitance with applied voltage, aluminum electrolytics also show a capacitance voltage dependance, although tantalums are much better. Generally the "film" types are OK. Monitoring semiconductor diodes is also interesting to see how the junction capacitance varies with applied voltage, these include Schottky, regular signal, rectifier, Zener and transistor junctions.

The cost of a lab type LCR meter that supports a wide range of applied DC voltages (or currents) is high, and many of the moderate cost lab type meters without this feature can be "augmented" with a circuit to isolate the DC from the AC measurement and allow accurate measurements in the presence of DC bias. Here's a link to a DIY device for such.

https://www.eevblog.com/forum/projects/bias-network-for-lcr-meter/msg4091638/#msg4091638

Best,

[Martin72]

Hi,

because its accuracy  is stated as 2% of 2mH

According to the sheet you´ve linked, there is no remark that the 2mH range got 2%.
It´s only stated, the BEST accuracy will be 2%, not on which range.

I... I realize that this is what a literal interpretation of the column title might imply, but surely this is wrong and the column is just poorly labelled. There's no way that ±(2%+8) figure applies to only one range; and that the manufacturer is making absolutely no claim whatsoever about the performance in the other ranges?

What makes you sure to think it isn´t so ?
As I´ve read the "specs" I´ve stumbled over the word "BEST".
Best means to me that there will be also "normal" and "worst" accuracy - but not shown in this case.
2nd pic below shows a snip of the specs from peaktech 3730 lcr-meter, nearly similar.
That´s more plausible, there are 2% and 5% accuracy mentioned and when you state only the 2%, it will be the BEST accuracy.
(Same on R and C)

[rs20]

OK, I stand corrected, thanks for taking the time to provide evidence. I'm appalled that they think it's OK to present their performance specs like that, but I guess I'm just naive.

[bd139]

Worth pointing out that specifications are always conservative here or open to interpretation because fixture and lead capacitance and inductance is a function of the measurement methodology employed by the user. They may make up a significant portion of the measurement. This is unlike resistance which is only an issue on very low resistance measurements.

Ergo it’s better not to worry about a couple of percent and a few counts and work out what the tolerance and device spread of your design is and make sure you window values within that.

[AVGresponding]

Also worth pointing out the laughably wide manufacturing tolerances on some components, notably electrolytic caps which can be as wide as +20/-80%

Edit: Might have been -20/+80% that I saw once. Either way, measuring the capacitance there would have limited use, ESR would be the better measurement.

[bd139]

Usually -20% +80%. Unless it's ChongX and then you might as well ignore the label and look at the size of the can and make a finger in the air estimate.

[mqsaharan]

Accuracy for LCR meters that use Cyrustek ES51919 and ES51920 based chip sets are pretty good. Example : DE-5000, MS-5308 or UNI-UT612 and etc.

For calibration there was an info shared by other fellow here in this forum ->

For those who would like to perform unit calibration, here is the information and it is also works on Mastech MS5308, UNI-UT612 or whatever using Cyrustek ES51919 and ES51920 ship set.

For DE-5000, adjust DC 500mV between TP1 and TP2.

For others, adjust DC 500mV between pin 26 and pin 28 on ES52920.

Be careful while calibrating these LCR meters. You need a meter with Hi-Z input impedance. Even a 10M input impedance meter loads down the reference voltages.