LCR tweezers 100 kHz.

[iet]

I will give another example. Ceramic capacitor 0.47 pF. The distance between the capacitor terminals exactly matches the distance between the tweezers tips.

[mawyatt]

I will give another example. Ceramic capacitor 0.47 pF. The distance between the capacitor terminals exactly matches the distance between the tweezers tips.

At ~ 0.48pF @ 10KHz the DUT impedance is ~33MΩ and test current is only ~30nA 

At this high Z level we would put more faith in a mechanically & electrically stable SDM SMD Test Fixture with an LCR Meter than tweezers where hand placement, tweezer orientation, DUT orientation & mounting all can have significant measurement influences.

Best,

[iet]

LCR Reader offers its own method for measuring extremely small inductance values.
https://www.lcr-reader.com/documents/AutoTestCon_2022.pdf

[mawyatt]

LCR Reader offers its own method for measuring extremely small inductance values.
https://www.lcr-reader.com/documents/AutoTestCon_2022.pdf

Well that's an interesting paper, as it seems they don't even know how to properly calculate the impedance of an inductance, which leaves everything else in question

That alone kept us from reading any further than the introduction, as obviously this paper hasn't had any peer review!!

Thanks for the link, this is a good example of why folks shouldn't believe and value everything they get on the internet. Even papers that seem to "Look Official", and emphasizes that one should know basic electrical & measurement fundamentals and limitations of such before believing everything they read.

As we said many times KTI (Know Thy Instrument), especially true for extreme measurements  

Best 

[shabaz]

To be fair, even seminal papers have had errors occasionally (some of which are not detected for decades). I noticed the error in the abstract, a decimal place slip-up, granted it shouldn't be there if it was checked. But it would be good to know if there's any deeper mistake in the paper or not (especially if maybe if the authors could see this and explained themselves why they are confident in what they wrote, and accepted the mistake).

[Sensorcat]

To be fair, even seminal papers have had errors occasionally ...

An error is OK, but simply everything is defective in this paper, such that actually no method is disclosed. The author does not describe his measurement setup, instrument settings, part numbers of specimen used, calculations performed. The results shown in the plots just spring into existence. And then this fuzzy language, that does not explain, does not provide a design or research goal, draws no conclusion. Formal errors on top: ridiculous references, two sections with the same title. A joke camouflaged as a scientific paper. Pauli would probably say: "It is not even wrong!"

[shabaz]

There are many papers where numbers are summarized, with the raw data not there. Anyway, that's a different thing to what I mentioned. I don't know if the paper has other mistakes or not, and it may well be the quality is lacking in other ways.

However, even for a novice, it seems to indicate quite clearly what (at least part of) the secret sauce is that may (or may not, I know nothing of their product) provide better results at small inductances for their LCR tweezers. I don't know what the specification is for that product (I'm not in the market for spending that much on LCR tweezers), but it certainly makes clear that the specification relies on some assumptions about the geometry of the inductor for low inductances. For that at least, the paper was interesting. By the way, please note I'm not being argumentative nor even playing devil's advocate here. I'm not interested in that. I just happened to look at that paper out of interest and noticed that error. I'm not in the market for these LCR tweezers, nor do I own any. I do use other tools to measure inductances occasionally.

[Kean]

I only skimmed the "paper", but it seemed to basically offer a rule of thumb (fudge factor) for getting "more accurate" measurements... for when you are using the wrong tool for the job.

I didn't spot the calculation error someone mentioned, but the graphs all using different color keys was very poor form.

I haven't gone searching the LCR Reader/Siborg website to see if they offer more detail on accuracy specs.  Other manufacturers have those handy charts that show how measurement accuracy is affected by multiple variables - most notably the combination of value vs test frequency.

[egonotto]

LCR Reader offers its own method for measuring extremely small inductance values.
https://www.lcr-reader.com/documents/AutoTestCon_2022.pdf

Well that's an interesting paper, as it seems they don't even know how to properly calculate the impedance of an inductance, which leaves everything else in question
....
Best

Hello,

I can only see one formula in the rather confusing paper. And this seems to want to correct a measured value by an offset and a factor.
Could you please tell me where an impedance of an inductance is calculated incorrectly.

Best regards
egonotto

[shabaz]

Hi,
It's in the abstract (they mention 6 milli-ohm and 60 milli-ohm, but their decimal point is wrong by one place.
As you say, they are simply correcting by an offset+factor, based on the measured value. It doesn't seem novel, just seems to be some form of small justification of why they do it. Relies on the fact that they are hoping that a small inductance is likely to be in a small (0603 or smaller) package.

[mawyatt]

I only skimmed the "paper", but it seemed to basically offer a rule of thumb (fudge factor) for getting "more accurate" measurements... for when you are using the wrong tool for the job.

I didn't spot the calculation error someone mentioned, but the graphs all using different color keys was very poor form.

I haven't gone searching the LCR Reader/Siborg website to see if they offer more detail on accuracy specs.  Other manufacturers have those handy charts that show how measurement accuracy is affected by multiple variables - most notably the combination of value vs test frequency.

In the Introduction, a 1nH Impedance at 100KHz is ~0.6 milli Ω, not 6 milli Ω, and at 1MHz ~6 milli Ω, not 60 milli Ω as stated.

One could "forgive" this as a decimal slip (twice) except it's right up front in the Introduction and the whole subject is about measuring low Z inductance where this is the Parameter of Discussion!!

Obviously no peer review and note how the paper is configured as an IEEE type paper, kind of implying it's an IEEE paper which it's not!!

Anyway, agree this seems more about using the wrong tool for the job, a little advertising for the their LCR tweezers, and using "fudge factors" without any details of how these were derived. Very little valuable technical content IMO.

Best

[egonotto]

Hello,

But there is one good point in the paper, namely:
"[2] Chapter: Inductance Measurement, by Michal Szyper, Measurement, Instrumentation, and Sensors Handbook, Edited By John G. Webster, Halit Eren, eBook Published January 31, 2017, CRC Press"


Best regards
egonotto

[shabaz]

..this seems more about using the wrong tool for the job

"the wrong tool for the job"
That's a good point, but it raises the question, what is the job? It depends. Taking me as an example: If I'm making the effort to measure an inductor precisely, then I'll measure it ideally at the frequency of operation, and sometimes within a larger network, and I'm prepared to spend the time setting it up, there will be a time investment of say 10 minutes or more, depending on what tools I'm using and if I decide to plot the results or not. It's rare I would want precision, because many circuits can be trimmed in one way or another to reach precision without requiring the inductance to be initially precise. Besides, there may be plenty of temperature dependence, or component tolerance, so one might not fully rely on the inductance precision anyway.

However, if I ever had a need to do a spot-check to confirm the ready-made tiny SMD inductor I've picked up is 10nH or 22nH nominal, then I believe users like me would not need to care if the measurement error is in tens of percent. For that reason, from my perspective I could only get interested in a pair of LCR tweezers if:

(1) they supported measurement with a granularity of at least 10 nH for the low inductances, and
(2) if they were cheap(er), because I would not be using them for precision measurements

Maybe (1) can only be achieved well in a small form-factor if there's really a 4-wire measurement going on fairly close to the probe tips (if it cannot be possible right at the probe tips) - I don't know if any LCR tweezer manufacturer is doing that or not.

From the few photographs in this thread, it does seem to appear that nH and tens of nH can be measured in a tweezer form-factor to at least some degree of success. There's still not a lot to go on, but it looks plausible from the photos. But I'd like to see more evidence, and a lower cost. Until then, personally I'm only interested in handheld or desktop instruments; if I'm spending several hundred $ then I don't see the benefit of tweezers when for a similar cost, I can get a handheld instrument that will work down to nH.

Of course, just because LCR tweezers do not meet my needs, doesn't apply to others, they will have a different job to apply it to.

For instance, if I was doing production work, then that's a different kettle of fish. Maybe the product being produced uses no inductances that low. In that case, I could find the tweezers very useful perhaps. And I'd be prepared to spend a decent amount on the tweezers more for the reliability rather than precision.
I don't know, because I don't do production work.

[Kean]

"the wrong tool for the job"
That's a good point, but it raises the question, what is the job?

When I used that phrase earlier, I was referring to what iet was trying to show.  And to a lesser extent, what that paper was promoting (i.e. use of their products beyond specification)
I don't know iet's use case or intentions, but it seemed to be showing off or shilling products.  Maybe just fanboy-ism.

In general I do use SMD tweezers regularly (not quite daily), but only for a quick component value check.
e.g I have small container of passives that have come off my PnP machines that I keep handy for prototype or rework.  I know it is full of my most common values, so saves me having to pull out a bag from stock.

I wanted some SMD tweezers for so long, but the price was too high for my usage.  I now mostly use the Shannon tweezers, and I'm very happy with them.
I also have some cheap ones which are rubbish.  I don't recall the brand/model but they are bright yellow and quite bulky, but handy to keep at my other workbenches.

If I want to do any kind of precision measurement, I will use my bench LCR (a CKT5000) with an appropriate fixture.  I often use it to graph frequency sweeps, which just isn't practical with tweezers.
Sometimes I will need to go beyond 5MHz and will use my tinyVNA or LiteVNA64, but then the setup/fixturing is even more important.

[shabaz]

I couldn't tell if it was astroturfing either, which is annoying if it is : (

What you mention about a quick component check is pretty much all I would want too, with that caveat of needing checks below 100 nH. So for the same reason, I find these tweezers too high priced for my usage too.

It would be nice if a manufacturer addressed the market for those who do not actually need much accuracy from a tweezer device, but do want the granularity down to nH or tens of nH. Maybe that market just isn't all that large : (

$395 (which is just one midrange price for the LCR-Reader, there are even higher!) is crazy.. it puts it squarely outside of what I would value such a product (because I would not want to pay for accuracy I did not need for quick spot checks). Personally, even $200 is too high for the very basic use I would put it to, so unless it comes down in price a _lot_, it's an uninteresting product for me.

Totally agree with the setup time with the VNA, which is massive compared to LCR tweezers. I don't mind that, since if I'm using the VNA then I do want the deeper insight, and that's time well spent usually. I always end up learning something each time!

[Sensorcat]

What you mention about a quick component check is pretty much all I would want too, with that caveat of needing checks below 100 nH. So for the same reason, I find these tweezers too high priced for my usage too.

It would be nice if a manufacturer addressed the market for those who do not actually need much accuracy from a tweezer device, but do want the granularity down to nH or tens of nH. Maybe that market just isn't all that large : (

Yes, I guess this specific market must be small, because people who need to measure inductors less than 100nH will usually have other ways to do that, most probably also at higher frequencies. But I also think that no one wants to offer that in the lower price range because small inductance ranges are difficult to implement with the tweezer form factor:

• Higher test frequencies are necessary, but that will drive the bill of materials up.
• Higher test currents will help, but that will deplete the small battery fast.
• 4-wire connection will improve accuracy and stability, but is tricky with tweezers (who wants to have an instrument that always complains that the connection to the DUT is incomplete?). I'm not aware of any offer, either tweezer instrument or tweezer test fixture, that brings 4-wire connection to the DUT, such that contact resistance is eliminated, too. And exactly that would deliver RF properties of the device, both real and imaginary part.

So it's 'wrong tool for the job'  for small inductors by design principle, I would say. For anybody who does not need that and is price sensitive, this could be an option:

https://www.mastech-group.com/global/en/ms8911.html

That is not a recommendation, I never had one and do not know how well it performs. But it looks quite attractive on paper.

[shabaz]

That looks quite good for users who need a quick check and (like me) don't care about high accuracy in this form-factor.

I guess if there was some matrix of these parameters like price, granularity for low-value components, and size, and speed-of-use, in the absence of LCR tweezers with that sensible price and granularity at the low end, then I'd probably gravitate toward something like Extech LCR200 attached to 2-wire tweezers for the times I didn't want any accuracy, and wired to a DIY 4-wire measurement setup for the times I did. It's cheaper than the LCR Reader product, and more expensive than the Mastech, and not as speedy as picking up LCR tweezers, but probably would suit me more.

I've had my eye on that Extech for quite a while (and might bite sooner if I come across a bit of a discount, even though the price seems very fair, I can't complain), but have just resorted to other tools for the time being.

[mawyatt]

Another issue with tweezers, either the fully complete types or the types for use with a bench LCR meter is the effects of the users hands on the measurements. If the tweezer handles host unshielded wires to the blade tips, the users hands can have an influence on the results, this is more pronounced with higher Z measurements such as small capacitors.

Some quality tweezers for use with bench LCR meters like the Tonghui TH26009B and various clones, have shielded cables up to the tweezer blade tips for all 4 wires in a Kelvin arrangement. This works well shielding the users hand effects and also extends the tweezer frequency range to 15MHz for the TH26009B.

Best

[iet]

What you mention about a quick component check is pretty much all I would want too, with that caveat of needing checks below 100 nH. So for the same reason, I find these tweezers too high priced for my usage too.

This LCR tweezers may match your request.
TH2822M-230$.
https://www.ebay.com/itm/225599830680

[shabaz]

Thanks for the link, I'll check it out soon.
I have a Tonghui bench LCR meter, I like it.

[mawyatt]

Thanks for the link, I'll check it out soon.
I have a Tonghui bench LCR meter, I like it.

Caution, Tonghui is very strict about dealing with products from unauthorized dealers/resellers, could end up with no warranty and have to deal with warranty claims from the reseller.

Also, check the ebay resellers rating and reviews, very questionable source IMO.

Best

[iet]

For anybody who does not need that and is price sensitive, this could be an option:

https://www.mastech-group.com/global/en/ms8911.html

That is not a recommendation, I never had one and do not know how well it performs. But it looks quite attractive on paper.

Those tweezers that have appeared recently can also measure inductance up to 100 nH, based on the technical characteristics. I can definitely say about Fnirsi ST1.
If we consider only inductance measurement, then it is cheaper than MS8911.And its measurement speed is higher.
Measuring smaller inductance values ​​is a slightly different circuit, components, implementation and price. Everything has its own cost.

[iet]

Information from the LCR Pro1 on accurately measuring small inductance.
https://www.lcrresearch.com/post/use-lcr-pro1-smart-tweezers-style-lcr-meter-to-measure-small-inductors-accurately
Now I'll tell you where the LCR Pro1 tweezers can help in real practical work.
When I made the tweezers, the 100 kHz ESR meter, I had to measure the inductance of the primary and secondary windings of the transformer very accurately. Here, nanohenries are not so much needed as high resolution. There must be a strictly multiple ratio of inductances. Like 550 nH and 550 uH. Then we can easily get the rest of the parameters.
This also applies to checking the inductance of the windings of factory transformers.
The second use of the tweezers is measuring the inductance of SMD calibration resistors. Because the smaller the inductance of the resistor, the smaller the error in calibrating the scale.

[mawyatt]

Information from the LCR Pro1 on accurately measuring small inductance.
https://www.lcrresearch.com/post/use-lcr-pro1-smart-tweezers-style-lcr-meter-to-measure-small-inductors-accurately
Now I'll tell you where the LCR Pro1 tweezers can help in real practical work.
When I made the tweezers, the 100 kHz ESR meter, I had to measure the inductance of the primary and secondary windings of the transformer very accurately. Here, nanohenries are not so much needed as high resolution. There must be a strictly multiple ratio of inductances. Like 550 nH and 550 uH. Then we can easily get the rest of the parameters.
This also applies to checking the inductance of the windings of factory transformers.
The second use of the tweezers is measuring the inductance of SMD calibration resistors. Because the smaller the inductance of the resistor, the smaller the error in calibrating the scale.

That link is interesting, we like the resolution of the Pro1 Tweezers down to pH, 10X better than the lab Tonghui LCR Meter with TH26009B Tweezers, we must remember that a pH @ 100KHz is just 628 nΩ (yes nano ohms) 

The TH26009B Tweezers for LCR Meters are some of the better tweezers available IMO, and the cost reflect this. They have shielded wiring thru the handles up to the blade tips and do a good job with various size SMD components.

We have both the Tonghui TH2830 LCR Meter and the TH26009B Tweezers and decided to run a quick test of SMD inductors @ 100KHz like the article. We did the Open Short Cal with the Tweezers a few times, even cleaned the blades with IPA, but can't get a repeatable Short reading within at least +-0.5nH as often as we tried. This uncertainty also reflected sometimes even higher in various SMD Inductor measurements in the 10nH to 500nH range shown with various SMD sizes starting at 0201 like shown below. If we change the our hand pressure, move hands around, or change tweezer orientation we can get readings of just about anything we want around the inductor value

This type behavior is exactly why we invested in various SMD Fixtures for SMD devices, some OEM, some modified and some custom, where we needed accurate and repeatable results

Best

[iet]

It would not be entirely correct if I comment on the capabilities of these tweezers for the developers.
I can only answer for myself. I never post things on the forum that have not been tested in practice. Any low values ​​of the measured quantity must have a certain stability of repetitions. For me, this is the criterion of truth. If there is no stability, then it is necessary to analyze the entire measurement chain.
I appreciate your efforts in implementing different devices for different types of measurements. In general, I like people who do something and not just write long texts with deep thoughts on the topic of measurements. And it is even worse if they come up with different fantasies, attributing them to the topic starter.
In such an unstable range, it is difficult to make comparative measurements, being in different conditions.
I do not have a cool desktop device to check the result. I will now make a short comment and show the measurement results. If you can repeat the same on your cool device, I will be very grateful to you.
A good tweezers in the nanohenry range should react to a piece of copper wire. Such measurements were made with a piece of copper wire 0.65 mm, information is in the article from LCR Reader. I also made such measurements with a piece of copper wire 0.65 mm: 5 mm, 2.5 mm, 1.25 mm.