Measuring small capacitance with KS34465A & TH2830

[mawyatt]

When still employed a few years back we had all the proper instruments for measuring components and such, so getting timely decent measurement results wasn't difficult since someone else did the measurements

Now retired and having to do everything myself, but still "working" on a project, I need to make these measurements.

This ongoing project requires some components that are tightly matched and with tight ratios. Getting accurate matched components before COVID was difficult, now next to impossible without very deep pockets, so the decision was made to select matched components from a large batch of lesser accurate components, which provided the perfect "excuse" to sacrifice some project funds for a proper LRC meter

After considerable research we selected the Tonghui TH2830 LRC meter, TH26011CS Kelvin leads & TH26009B precision tweezers which has all the features for accurately measuring & binning components, but still working the learning curve on this new instrument. Out of curiosity we decided to check the capacitance measurement capability of the KS34465A & SDM3065X using some leaded film type capacitors covering the range of 1nF to 1uF, the DMMs compared well to the LRC meter. As expected electrolytic, and higher capacitance ceramic capacitors (X5R and X7R types) didn't agree well due to the different measurement techniques (for another discussion).

Then we turned to the lower value capacitors with the COG dielectrics which are SMD and require tweezers. We didn't have any proper tweezers for the KS and SDM DMMs except for the cheap two lead tweezers shown here.

https://www.aliexpress.com/item/1005002619578037.html?spm=a2g0s.9042311.0.0.523e4c4dUHGAQc

These tweezers don't shield the leads to the probe tips like the 4 wire TH26009B and thus subject to external influences and have almost 50pF distributed capacitance, so how is one supposed to get decent measurements at lower capacitances

Turns out the KS34465A has a few "features" that can help, first is the handy "null" feature to null out resistance in the test leads, this also works on capacitance!! Second is the MX+B "scaling" function to add extra resolution, also works on capacitance!!

We were able to get reasonable readings in the 10s of pF utilizing these cheap tweezers employing the features above and some careful technique. If the tweezers are gripped like you intend to use them and the tips carefully positioned opened about the same opening as the intended SMD chip, then the null invoked, the residual probe (and hand) capacitance can be effectively removed from the measurement.

This technique can work in a pinch when you don't have a proper LRC meter & fixture handy, but need to get a reasonable estimate of a small value chip capacitor, or match chips caps.

We've ordered another set of cheap tweezers, that may prove a little better. These have BNC type connectors which will need to be replaced with banana connectors for the DMM use.

https://www.aliexpress.com/item/1005001722184966.html?spm=a2g0s.9042311.0.0.10554c4dVoFphY

Anyway, hope this helps when in a measurement pinch.

Best,

[nctnico]

If you need to do a lot of measurements on SMD components then do yourself a favour and get a proper SMD test fixture for the LCR meter. These aren't insanely expensive (the one from GW Instek with part number LCR-09 is around US $200) and save a lot of hassle.

[mawyatt]

Agree a proper SMD fixture is in order after we get more familiar with this instrument.

Best,

[oz2cpu]

Tonghui TH2830 should be able to tell capacity ? AND series resistance at the SAME time ?

and in inductor mode ?
L and R at the same time ?

[2N3055]

Tonghui TH2830 should be able to tell capacity ? AND series resistance at the SAME time ?

and in inductor mode ?
L and R at the same time ?

Yes it can and much more... it is very capable instrument.

[oz2cpu]

thanks, i go look on some more detailed measurement demo videos
if anyone know some , please add links.

anyone tried to place it on top of siglent equipment ?
looks like same size ? will they fit nice and stable on top of each other ?

[mawyatt]

The TH2830 can measure R, L and C in both parallel and series equivalents with D and Q. Also can measure Z, Y, R+-jX and G+-jB.

Did some comparison R measurements with the KS34465A and very good agreement. Comes with full calibration sheet showing calibration results.

Supplied with a very nice Kelvin TH26011CS probes and TH26048A Test Fixture for leaded components. Also consider getting the TH26009B Kelvin Tweezers, these are really quite good for measuring SMD devices.

CAUTION!!!!! Do not purchase from any AliExpress site, none are authorized resellers of any Tonghui Products. OEM Tonghui will not honor the factory warranty if you purchase from an unauthorized site.

We purchased from BEICH Official Store on AliExpress and an issue with a 100 ohm range came up (likely a firmware issue, since it's revision 1.0). We contacted the OEM only to find out about the unauthorized AliExpress seller BEICH, and that the seller must provide any warranty support.

So now we are stuck waiting to see if Tonghui will provide some help, maybe with a firmware update. Tonghui understands our situation and may decide to help out, so purchase from only authorized sources.

We did open up the case to take a look and see if anything was loose. The internal construction is superb, and they are using a nifty toroidal power transformer for the mains power...first class indeed!!

Best,

[oz2cpu]

thanks as always premium feedback from you :-)

please tell me : what authorized dealer you recommend I goto ?
here is Europe there is a problem with VAT and shipment cost, so I prefer a contry in the EU if possible.

and please try to put it on top of your siglent equipment ? signal generators, multimeters, loads, power supply or similar with rubber edges.
and the other way too, fit ? stable ? or looks really bad ?
can I take off the handles ? so more rows can be placed closer ?

I am a little bit crasy when it come to my lab design,
and posibilities to place equipment where it makes most sense in easy access

[2N3055]

thanks as always premium feedback from you :-)

please tell me : what authorized dealer you recommend I goto ?
here is Europe there is a problem with VAT and shipment cost, so I prefer a contry in the EU if possible.

and please try to put it on top of your siglent equipment ? signal generators, multimeters, loads, power supply or similar with rubber edges.
and the other way too, fit ? stable ? or looks really bad ?
can I take off the handles ? so more rows can be placed closer ?

I am a little bit crasy when it come to my lab design,
and posibilities to place equipment where it makes most sense in easy access

Batterfly from Italy sells them.

https://www.batterfly.com/shop/en/testandmeasurement/test-componeti/tonghui-th2830-lcr-meter

[mawyatt]

I'll send a PM to Tonghui contact, if I get OK from them.

2N3055 beat me to it

Here's how the TH2830 sits on a HP34401A. The handles seem to just pop out.

Best,

[nctnico]

thanks as always premium feedback from you :-)

please tell me : what authorized dealer you recommend I goto ?
here is Europe there is a problem with VAT and shipment cost, so I prefer a contry in the EU if possible.

and please try to put it on top of your siglent equipment ? signal generators, multimeters, loads, power supply or similar with rubber edges.
and the other way too, fit ? stable ? or looks really bad ?
can I take off the handles ? so more rows can be placed closer ?

I am a little bit crasy when it come to my lab design,
and posibilities to place equipment where it makes most sense in easy access

Batterfly from Italy sells them.

https://www.batterfly.com/shop/en/testandmeasurement/test-componeti/tonghui-th2830-lcr-meter

That is a good suggestion there. I have been eyeballing the Tonghui meters as well for a long time but never found a good source for them until now 

[oz2cpu]

does not count mike with agilent equipment :-)

--

See the attached test fixtures, what would you get ?

both ?

I kind of miss an SMD version ?? does it exist ?
as far as i understand, there is nothing inside the test fixture ?
only a mecanical stable solition, each side of the spring loaded contacts, connect to the centre, of each of the 4 bnc
right ?
just like the kelvin test leads.

[2N3055]

does not count mike with agilent equipment :-)

--

See the attached test fixtures, what would you get ?

both ?

I kind of miss an SMD version ?? does it exist ?
as far as i understand, there is nothing inside the test fixture ?
only a mecanical stable solition, each side of the spring loaded contacts, connect to the centre, of each of the 4 bnc
right ?
just like the kelvin test leads.

I think this one from GW-Instek would fit...

https://www.batterfly.com/shop/en/gwinstek-lcr-15

[TimFox]

I used this fixture with impedance analyzers that used a 50-ohm reflection bridge (as shown in the photograph):
https://www.coilcraft.com/en-us/other/smd-a/
Both the Coilcraft and the GW-Instek unit show the mechanical requirements to connect to a SMT part, but the GW-Instek is appropriate to 4-wire Kelvin connections.
The SMD-B fixture is designed for 4-wire BNC connections.  https://www.coilcraft.com/getmedia/33893caa-0cca-48fe-b203-410cc6f83c41/smd-b.pdf
(Their website is a bit hard to navigate for these fixtures.)

[nctnico]

See the attached test fixtures, what would you get ?

both ?

I think this one from GW-Instek would fit...

https://www.batterfly.com/shop/en/gwinstek-lcr-15

I have this one and it is very nice to use for (small) SMT parts. It is being sold at many stores under various brands. The one I have is branded RS-Pro (bought from RS because I needed it quickly).

Kelvin clips are almost mandatory to have. The slotted adapter seems not very useful to me; connecting the Kelvin leads is easier.

[oz2cpu]

>I think this one from GW-Instek would fit...
>https://www.batterfly.com/shop/en/gwinstek-lcr-15

nctnico ?
so this means it is CONFIRMED to fit ?

the 4 bnc mounts distance and all, is an approved industry standard ?
or just pure luck if fit or not ?

[mawyatt]

Update, Tonghui provided a updated firmware and now the issue with the 100 ohms is corrected 

So the OEM Tonghui stepped up and helped us out, so hat's off to them

Regarding SMD Kelvin measurements, haven't found anything as good as the TH26009B tweezers. These have a guard shield up to the probe tips, and this allows very good low capacitance measurements by proper calibration of the shorted tweezer and open tweezer with the supplied special cal fixture. You really have to try these out to see how smooth and easy they operate!!

The unit comes with the Kelvin Clips TH26011CS and the TH26048A Test Fixture for leaded components, also a calibration short is provided for the TH26048A. We found this TH26048A Test Fixture quite useful when testing leaded components for sorting, much quicker than using the Kelvin Clips.

The advantage of the fixtures with the fixed BNC connections is repeatability, and I prefer them over the individual BNC cables. They are just mechanical interfaces for the interconnects, cable interface and instrument BNC interface, but well built and sturdy fixtures.

Now that our new DMM6500 just arrived, we have some precision comparisons to make between the TH2830, KS34465A and new DMM6500. Also have a few different types of Kelvin Clips & Tweezers to compare when time permits.

Anyway, this Tonghui TH2830 seem like a very good value and accurate LCR Meter. If anyone decides to give one a try, please remove the cover, there's no warranty seal to beak      and take a look at the unique mains toroid transformer

Best,

[nctnico]

>I think this one from GW-Instek would fit...
>https://www.batterfly.com/shop/en/gwinstek-lcr-15

nctnico ?
so this means it is CONFIRMED to fit ?

the 4 bnc mounts distance and all, is an approved industry standard ?
or just pure luck if fit or not ?

I used it on a GW Instek LCR meter but I do believe the distance between the BNC connectors of LCR meters is an (unofficial?) industry standard.

[oz2cpu]

thanks a lot mike, i am SO close to push the purchase button,
but cant find any video material showing real usage,

any takers ?

I like to see a few caps with value and ESR displayed at the same time,
how to change test frequency, for the ESR to make sense on small versus large value.

and the same performed on a few inductors,
please also try to add a series or parallel resistor,
and see if it pick the right results,
I assume it is able to measure both series, and parallel resistance on caps ?
one will be very low, and the other will be very high of course

[mawyatt]

No video but a very quick set of measurements, used different Kelvin clips for each Meter.

Some Vishay Resistors

TH2830                     DMM6500 (rec'd new 10/14/21)                 KS34465A                       

120.029                    119.993                                                   119.993
3.00235K                  3.00299K                                                 3.00298K
13.0061K                  13.0072K                                                 13.0073K
75.0473K                  75.0514K                                                 75.0523K
2.00229                    2.0024                                                     2.0003 (not Vishay)
0.12056                    0.12056                                                   0.1197 (not Vishay)

Film Caps   ESR @ 1KHz
1.00876uF (0.0179)  1.0087uF                                                  1.0089uF
0.47122uF (1.346)    0.4791uF                                                  0.4792uF
102.287nF                102.65nF                                                  102.66nF
2.25264nF                2.282nF                                                    2.257nF

Electrolytic HF Low ESR type {Single C measured at 1,10K, LR measured at 100K since past SRF)
423.195uF (0.06748) @ 1K     OverRange (DMM650)                  425.38uF
626.246uF (0.06225) @10K
152.046nH (0.07656) @100K  (measure cap with L & R series mode) 

C with ESR and ESL, so resonate frequency of C ~20KHz agrees with LCR meter. Interesting that we could get an equivalent RLC series large electrolytic cap model with just two measurements, series C & R and series L & R. The ESL and series resonate frequency seem reasonable since the cap long leads were probed at the ends.

C model:
        C          ESR           ESL
*-----] (-----^^^^----~~~~~----*

Series Resonate Frequency ~ {1/((LC)^1/2)}/(2*Pi)




Inductor       ESR
226.175uH (0.27928) @ 1K
225.473uH (0.37788) @ 10K
220.041uH (4.36245) @ 100K

L with ESR

Anyway, hope this helps with some very quick, crude measurements with the TH2830.

Best,


BTW the case & rubberized moldings on this TH2830 are also superb, equal or actually better IMO than any of the Siglent, HP, Agilent, Keysight and Keithley gear we have. This device just looks like it belongs in a high end lab.

[mawyatt]

Here's a cheap SMD device including video, that might fit the Tonghui, don't have any idea if it will work, nor how good.

https://www.aliexpress.com/item/1005002561014252.html?spm=a2g0o.detail.1000014.7.7fdd4e8eKzjN7D&gps-id=pcDetailBottomMoreOtherSeller&scm=1007.33416.213724.0&scm_id=1007.33416.213724.0&scm-url=1007.33416.213724.0&pvid=8fbaed83-db51-4b3f-8479-fac00af499cb&_t=gps-id:pcDetailBottomMoreOtherSeller,scm-url:1007.33416.213724.0,pvid:8fbaed83-db51-4b3f-8479-fac00af499cb,tpp_buckets:668%232846%238113%231998&&pdp_ext_f=%7B%22sceneId%22:%2223416%22,%22sku_id%22:%2212000021152791964%22%7D

Here's the OEM SMD fixture TH26008A.

https://www.tonghui.com.cn/en/product_144.html

Best,

[Kean]

I have a TH26008A which I use on my CKT5000 LCR Meter.  No problems with compatibility.
Let me know if you want any measurements (physical or electrical).

[mawyatt]

I have a TH26008A which I use on my CKT5000 LCR Meter.  No problems with compatibility.
Let me know if you want any measurements (physical or electrical).

How does this work for measuring small value capacitors (1~10pF) and inductors (10~100nH, including ferrite bead types)? With small physical size 0603 and 0805 types, do you get good repeatability?

Thanks for the help.

Best,

[Kean]

I have a TH26008A which I use on my CKT5000 LCR Meter.  No problems with compatibility.
Let me know if you want any measurements (physical or electrical).

How does this work for measuring small value capacitors (1~10pF) and inductors (10~100nH, including ferrite bead types)? With small physical size 0603 and 0805 types, do you get good repeatability?

The CKT5000 gives pF and uH measurements with 5 decimal places.  I performed open and short calibrations on the TH26008A SMD fixture after a 1 hour warmup, then grabbed a few parts: 1pF (C0603C109C5GACTM), 6pF (generic), 39nH (L0603C39NJRMST), and 100nH (L0603CR10JRMST).  All 0603 size.

With the capacitors I could get stable and repeatable measurements in pF to 3 decimal places (+/-1 LSB) - so 1.003-1.004PF and 5.876-5.877pF (measured at 1kHz).  I ran several measurements and sweeps from 1kHz to 1MHz and the values ranged from 0.996-1.006pF and 5.865-5.877pF.  I also ran a sweep up to 5MHz, and I could see the capacitance drop off above 1MHz.  The SMD fixture is supposedly rated to 120MHz....

With the chip inductors, readings were a bit unstable at low frequencies but quite stable and repeatable at 100kHz and above.  At 1MHz I measured 0.045uH 327mohm and 0.121uH 514mohm.

I did some quick comparison measurements with my Tonghui TH2822C and tweezers, and readings were consistently higher and not as repeatable.  Typical readings were: 1.034pF, 5.954pF, 0.053uH, 0.131uH.

I swapped parts in an out of the SMD fixture several times, switching measurement modes and frequency.  When using the fixture and returning to original settings the readings were often spot on, or over the course of a couple of hours still within 0.010pF of the original measurements.

Smaller than 0603 parts is possible with the fixture (I've done 0402), but it becomes very fiddly.  I actually lost the 1pF cap near the end of these tests, but that was while using the TH2822C tweezers.

I didn't originally bother trying the tweezers that came with the CKT5000 because I knew it would probably lead to parts flying...  but I just ran a quick measurement of the 6pF cap at 100kHz with the tweezers and reading was 5.56-5.60pF.  So definitely less stable and presumably less accurate.  I did re-do the open/short cal after changing to the tweezers.  The CKT supplied tweezers are nicely built, and feel quite solid compared to the ones that came with the TH2822C.

Hope this helps.  This is way overkill for most of the stuff I do.

ETA: I forgot to mention, after cal the fixture read 20mohm when shorted.
Edit 2: Oops, I've got a TH2822C not a TH2822A

[Kean]

Attached is photos showing the workings and guts of the TH26008A SMD test fixture.  It uses PTFE parts for the holder and probe isolation.
You can see each pair of red wires go into some sliding PTFE cylinders, and they then connect into the gold plated probes.
I'm not sure I want to open it up any more, but I am quite curious how the wires are terminated.
I've had some problems with getting good connection on older parts that are probably a bit oxidised, but otherwise I've found it works really well.