Best CHEAP LCR meter for small capacitors and inductors?

[Martin72]

Other question - is it normal to do calibration on DE-5000 every time you turn it on?

Depends on how accurate you want to have the measures.
Ambient temperature/humidity are changing plus warmup of the meter itself.

[thm_w]

If you didn't already have the DE-5000 then I would say consider these: https://shannontweezers.top/

[tautech]

If you didn't already have the DE-5000 then I would say consider these: https://shannontweezers.top/

Why ?
To be properly equipped you need both.

[thm_w]

If you didn't already have the DE-5000 then I would say consider these: https://shannontweezers.top/

Why ?
To be properly equipped you need both.

Just because its not entirely clear what OP is doing.
If its in circuit measurement then the tweezers are great to have. If its measuring a part, very precisely, prior to install, then maybe the mentioned SMD jig with a DE-5000 would be more repeatable.

[ranchero]

Just because its not entirely clear what OP is doing.
If its in circuit measurement then the tweezers are great to have. If its measuring a part, very precisely, prior to install, then maybe the mentioned SMD jig with a DE-5000 would be more repeatable.

I don't think it's even possible to measure such small values correctly in-circuit. So - I always remove them. The tweezers you've recommended only go up to 10kHz, so they will not measure small inductors.

[pqass]

I know you're probably not looking for a DIY LC meter now, but in case you've hit a wall, consider the LC-meter project below (or at least the oscillator part).

I'm currently implementing the LC+LM311 oscillator [on perf-board] just to validate the approach before committing further. See attachment.  I'm using a bread-boarded [reciprocal] frequency counter (which I happen to have handy) but any bench counter will do.  Also, I've removed the need for relays in my version; replaced with a DPDT and BJT.

See Part 1 of the video at
  10:28 for the theory of operation.  (https://youtu.be/KhJiE4gL5T4?t=628)
  16:26 for the six formulas involved  (just one frame at https://youtu.be/KhJiE4gL5T4?t=986)
  17:54 for the algorithm on how to determine C-DUT or L-DUT  (https://youtu.be/KhJiE4gL5T4?t=1074)
        To summarize:
        (a) first measure the freq. without any C-DUT attached,
        (b) then turn on calibration cap BJT (still no C-DUT),
        (c) then, attach a C-DUT (w/o cal cap) and measure freq.; use formula 5 to determine value.
            OR attach L-DUT (w/o cal cap) and measure freq.; use formula 6 to determine value.
        Calibration steps (a) and (b) only needed every once in a while.

I've successfully measured down to 16pf (all I had) just using manual recording of the frequencies with/without DUT and calibration cap into a spreadsheet.  Measuring inductors yields values close to whats printed, although, in the 10s of uH to 100mH; I don't have any nH inductors.  The chosen L+C values has problems with >1uF C-DUTs so I'll be exploring other combinations to determine ranges. 

FYI: I'm just a hobbyist (no EE) and as such, am not (yet) knowledgeable on the intricacies of those fancy HP LCR meters on eBay. But this appears to fit my C and L measurement needs and may be worthwhile for you. Also, in case you're wondering, the reciprocal frequency counter on the b-b [74HC4020+74HC74+ATTINY84] is based on www.instructables.com/High-Resolution-Frequency-Counter/ (good to 7 digits) but I've done a major rewrite of the software.


Original LC meter project links:
Hackaday IO: https://hackaday.io/project/178081-build-an-lc-meter
Github: https://github.com/coreWeaver/LC-Meter
Look for parts 1 thru 5: https://www.youtube.com/channel/UCSwxZFW1XBvgM92nI7XMwSA/videos 

Measuring Femto Farads:

[ranchero]

I would love to participate in this project, but unfortunately I'm not in the position when I can spend time on it. 

[mawyatt]

Generally small inductance is more difficult to measure than small capacitance. One fundamental reason is inductance is highly dependent on length, whereas capacitance is inversely so, and at a moderate frequency a small capacitance has a relative high impedance whereas a small inductor is low impedance.

As you move up in frequency the capacitive reactance drops and the inductance reactance increases, however for small L like 100nH is only ~ 0.063 ohms at 100KHz and a small capacitance of 100pF is ~ 15.9K ohms. So getting a good "zero" reference for the L is difficult, but relatively easy for C.

As mentioned by 2N3055 a VNA is maybe a better choice is some cases, and for small inductances this is the case!! The usual VNA is a low 50 ohm instrument that allows moving the "reference plane" up to the DUT fixture, thus removing the setup and fixture "length", LCR meters don't do this, although some of the lab types do have a "cable length" option which compensates for the test cable phase shift.

Measuring small capacitance ~100pF can be achieved with even a DMM capacitive measurement in a pinch, one must carefully null out the residual and keep the cables fixed in place (tapped to the workbench for example). Heck, Dave even showed how to measure Inductance with a DMM in his April 1 Video 

Best,

[thm_w]

Just because its not entirely clear what OP is doing.
If its in circuit measurement then the tweezers are great to have. If its measuring a part, very precisely, prior to install, then maybe the mentioned SMD jig with a DE-5000 would be more repeatable.

I don't think it's even possible to measure such small values correctly in-circuit. So - I always remove them. The tweezers you've recommended only go up to 10kHz, so they will not measure small inductors.

It just depends on the circuit in question, there is no blanket answer. If you know a specific inductor has nothing in parallel with it, then it will measure OK.

Those tweezers can measure a 0R 0603 at ~5nH in circuit. BUT thats very much in the noise, and I would not consider it repeatable at all. Which is where a jig is better (consistent distance, pressure, etc.).

[Vojtech]

As Mawyatt mentioned, measurement of nH at 100 kHz needs mOhm resolution, so 4 wires Kelvin is a must. A quick test of ET432 with shorted inputs at 100 kHz - resolution is down to pH, the result is stable to some hundreds of pH. Capacitance measurement at 100 kHz with open input – offset is some tenths of pF (can be nulled out), the resolution is 1 fF, the result is stable to 1 fF. Of course, the moving of cables changes measured capacitance with some tenths of pF. I have no etalon of low caps or inductors, so I am not able to verify absolute accuracy. Still, its meaning is questionable – with such low values parasitics in real PCBs play a significant role so absolute accuracy is questionable here. I think ET432 is better than DE-5000 with still an acceptable price.

[iMo]

I know you're probably not looking for a DIY LC meter now, but in case you've hit a wall, consider the LC-meter project below (or at least the oscillator part).
I'm currently implementing the LC+LM311 oscillator [on perf-board] just to validate the approach before committing further.

I built that famous LC meter myself some 20y back, wrote my own fw at that time. It works great.. The lowest C and L values are measured at something like 500-700kHz frequency.
PS: I would definitely use a relay with the 1n reference capacitor (not the npn transistor as you do)..
PPS: And do not forget the "Null" button where you measure the L or C of the probe wires before the actual measurement

[ranchero]

Just ordered ET432 with Kelvin clip while Black Friday sale is going on (about $130US with all discounts). Will compare with ET-5000 when I will receive it.

[hgjdwx]

LQ-9101 is better than DE-5000

[ledtester]

About the LM311-based LC meter... I found another design which uses a 74HCU04D instead of the LM311:

https://www.pcbway.com/project/shareproject/LC_meter_of_using_the_franclin_oscillator_9b1ebd81.html

The description there mentions a "Dr. Hegewald" and I was able to find this discrete LC-tank exciter circuit credited to him:

https://home.kpn.nl/a.van.waarde/lc.htm

[mawyatt]

Here's a post on using a cheap SMD Fixture connected to the DE-5000, well more kludged that "connected" 

Please note the entire thread and the earlier reference SMD 4.7pF cap used with the two SMD Fixtures and the Lab Grade LCR Meters for measurements.

https://www.eevblog.com/forum/testgear/cheap-smd-lcr-fixture-the-good-bad-and-ugly!/


Best and Happy Thanksgiving!!

[mawyatt]

LQ-9101 is better than DE-5000

Looks interesting but apparently only supports 10KHz, DE-5000 supports 100KHz!! This is an often overlooked specification, and the higher frequency range instruments need to employ more sophisticated circuits and algorithms, and why above ~100KHz the LCR meter price rises exponentially!!

Believe the DE-5000 utilizes a unique custom chip set that employs a high resolution SD ADC as well, also believe this is the same chip set utilized in the popular ET432 since both have identical Cal procedures (30 counts) & frequency ranges, and very similar specs. Soon we'll have a direct comparison of these two, as ranchero has both available

Best and Happy Thanksgiving

[thm_w]

As Mawyatt mentioned, measurement of nH at 100 kHz needs mOhm resolution, so 4 wires Kelvin is a must. A quick test of ET432 with shorted inputs at 100 kHz - resolution is down to pH, the result is stable to some hundreds of pH. Capacitance measurement at 100 kHz with open input – offset is some tenths of pF (can be nulled out), the resolution is 1 fF, the result is stable to 1 fF. Of course, the moving of cables changes measured capacitance with some tenths of pF. I have no etalon of low caps or inductors, so I am not able to verify absolute accuracy. Still, its meaning is questionable – with such low values parasitics in real PCBs play a significant role so absolute accuracy is questionable here. I think ET432 is better than DE-5000 with still an acceptable price.

Yeah that looks quite good.
Wonder if ET430 can be hacked to ET432. ET433 seems like it might have different hardware inside.

[balnazzar]

Hello,

is an indirect measurement via the resonant frequency with L or C connected in parallel an alternative?

Best regards
egonotto

Beware of indirect measurement if you are not super-experienced. Parasitic stuff in the components will make you calculations' result quite unreliable.

[mawyatt]

This may be the chip set utilized in the DE-5000 and ET432. The chip set appears to incorporate bridge approach and contains most of the necessary components, including high res ADC (likely high bit count Sigma-Delta type), to realize a quality LCR result.

Believe this chip set is the main reason these handheld LCR meters perform and yet are reasonably priced, classic example of custom chips improving performance and reducing costs.....integration rules

http://www.cyrustek.com.tw/wp-content/uploads/ES51920.pdf

Best,

[Vojtech]

ET432 has a higher resolution and better accuracy than mentioned chip, so different hardware has to be used.
To thm_w : ET432 comes with Kelvin clips, ET430 does not.

[TimFox]

For a while, IET (successor to General Radio) offered their DE-6000, which appears to be a rebadged DER meter with tighter specs.
It is no longer available from IET, and was in the same package as the DE-5000, but a nicer color.
https://www.ietlabs.com/pdf/Datasheets/DE-6000.pdf
IET's data sheet for their (also discontinued) DE-5000 (in yellow):
https://www.ietlabs.com/pdf/Datasheets/DE_5000.pdf
Slightly better specs on the more expensive 6000:  at 1 kHz and the 200 nF FS range, the 6000 spec'ed 0.2% and 2 digits, while the 5000 was 0.3% and two digits.
When it was available, I recommended the IET DE-6000 to a colleague because it was available with a traceable calibration, which his job required.

[MathWizard]

LCR tweezers sound nice. But yeah if the good ones are starting around $300, I'm in no rush.

I have a cheap Mastech LCR metter but it has Kelvin probes. But it's not the best at measuring large caps like +1000uF. I can't remember it's rating, but some times it seems to do it just fine, but other times, it just never gets a reading, even at lower frequencies too.

I don't really need another LCR meter, but I would like a better 1, but I'd love the self-contained chopstick ones too. The case of the mastech meter is really big, but it's mostly hollow. But well it does have a nice huge screen tho.

[tautech]

LCR tweezers sound nice. But yeah if the good ones are starting around $300, I'm in no rush.

Maybe you missed the development of these at half the cost of the big boys:
https://www.eevblog.com/forum/testgear/design-a-new-precision-lcr-tweezers/

[mawyatt]

As mentioned earlier, this setup for SMD with the DE-5000, a modified TL-21, and a cheap SMD fixture works extremely well, and is highly repeatable. We get results that agree quite well with our IM3536 & TH2830 lab grade Bench LCR Meters. The DE-5000 does have True 4 wire Kelvin type capability, with separate Guard Ground which are utilized in this setup, and a likely reason for the performance .

Highly recommend setup for those interested in quality SMD measurements, quick, accurate (compared to lab bench quality LCR meters), repeatable, that we prefer over tweezers.

Edit: BTW the reading here was from a quality Venkel 0603 4.7pF COG cap that measured 4.71pF on 12/23/22. Also, the modified TL-21 and SMD fixture were connected, plugged into the DE-5000, and the DE-5000 wasn't even recalibrated for this quick measurement


Of course YMMV.

Best,

[vinlove]

I bought this Chinese LCR meter from eBay about 10 years ago.  It is still working OK.  It was cheap, but now I cannot remember exactly how much I paid for it.

https://www.amazon.co.uk/dp/B096MJZ6ZL/ref=sspa_dk_detail_7?psc=1&pd_rd_i=B096MJZ6ZL&pd_rd_w=8mC9s&content-id=amzn1.sym.15c0cc83-c6c5-4d44-aa3d-0de17a9f3682&pf_rd_p=15c0cc83-c6c5-4d44-aa3d-0de17a9f3682&pf_rd_r=SY4C0ES1R53NC7H64422&pd_rd_wg=qeS0x&pd_rd_r=8fe0b2d6-72d5-46f4-a213-8e0fb633ee2a&sp_csd=d2lkZ2V0TmFtZT1zcF9kZXRhaWw