Sensitive LCR Meter Measurements

[mawyatt]

After using the Special Fixture with the IM3536 this morning we decided to try the Tonghui TH2830 LCR meter while riding out Tropical Storm Debbie. Here we've copied from post 21 above and added the TH2830 Results with the Special SMD Fixture.

We first did the Special SMD Fixture, then the Standard SMD Fixture. Afterwards we decided to reinstall the Special Fixture, do a quick Zero Cal, and rerun the test DUT Resistors, note the repeatability in the Special SMD Fixture results.

Resistor Value
in Milliohms       Special SMD Fixture  (Repeat see text above)          [number] With TH2830      Standard SMD Fixture     

1                       0.80                           (0.80)      {0.82}     <0.81>      [0.79]                         ~1.23
5                       4.84                           (4.82)      {4.82}     <4.86>      [4.81]                         ~5.45
10                     9.90                           (9.87)      {9.86}      <9.92>     [9.82]                          ~11.98
15                     14.70                         (14.71)     {14.72}    <14.73>   [14.69]                        ~13.29
20                     19.83                         (19.84)     {19.85}    <19.85>   [19.83]                        ~17.84
40                     39.87                         (39.88)     {39.88}    <39.88>   [39.88]                          44.87
50                     49.87                         (49.91)     {49.91}    <49.93>   [49.96]                          48.29
100                   100.21                       (100.23)    {100.21}  <100.27> [100.25]                        99.28
   
"~" indicates some variability in readings
Edit: 4~5 hours later redid the measurements shown as {#} for additional results.
Edit: From <#> post 24 below.

These results hint at the good repeatability with this Special SMD Fixture when using the same LCR Meter for low Z DUTs, but also show good agreement between the different brand quality LCR Meters we have.

Hopefully others can benefit from this interesting concept of keeping the Kelvin Force and Sense connections separate and allowing the SMD DUT conductive (solder) end caps make the Kelvin Connections, seems to work a treat in our case!!

BTW we've been "pinged" about making this Special SMD Fixture available as a kit. A couple possibilities are worth considering, one would use the same repurposed LCR Kelvin Clip Box with Lever Arm BNCs like shown, the other a custom enclosure design which would be an aluminum "U" design like the repurposed LCR Kevin Clip Box and Lever Arm BNCs (important for a quality design). 3D Printed components would be created as needed.

Anyway, let us know if interested.


Argh Counsel has advised against this, due to liability


Best,

[Martin72]

I have now written to them and asked for a data sheet, together with an explanation as to why.
The things look so similar, but it may well be that the cables from Sourcetronic do not come from TongHui or are “different” and are manufactured by TongHui for Sourcetronic.
We will know soon.

As I didn't get an answer from them either privately or from work, I called them "today".
They are in the process of getting this information, because they don't have it themselves....

[KungFuJosh]

Translation: TongHui definitely made it, and didn't supply good information. 😉

[Martin72]

That's how I see it too, because when I brought the name into play, they reacted rather narrow-mindedly.

[KungFuJosh]

That's because they have a lot of experience dealing with them. 😉

[EC8010]

I stopped using leads like those shown when I discovered that the capacitance of their coaxial cables gets converted into 100nH of series inductance. Until I discovered that, I thought all electrolytic capacitors had an ESL of about 100nH. 

I've found the best way to measure components is to hang a die-cast aluminium box directly off the analyser's four BNCs and have the Kelvin clips on unscreened leads. I also got fancy and bolted a vertical steel bar down the middle of the back of the box. Why? If you cut two pieces of tin plate and fold a right angle on their ends, then a few neodymium magnets to hold the tin plate in place allows you to have a variable length screen between the two Kelvin clips. Provided the screen is adjusted to the right length, I find my stray capacitance is only about 5fF - allowing me to measure little capacitors accurately. My large box is 8" x 5" x 3" so I bolted it to the ends of a pair of 1/2" steel rods that go under the analyser to take the weight off the BNCs.

[mawyatt]

Not sure "what leads" you are referring to?

If it's those in post #21 & #25 that come out the side of the small fixture box, they are not coaxial and in fixed position within a sturdy small shielded enclosure that directly attaches to the LCR meter with the 4 BNCs. The H and L sides are shielded within the box as shown. Since these leads don't move when installing the fixture, nor when making measurements, their effects are somewhat removed by the proper Open/Short Calibration.

The "trick" with this fixture is allowing the SMD DUT end terminals to make the actual Force and Sense connections for Hcur, Hpot and Lcur, Lpot respectively. This effectively removes the contact issues with SMD devices and has shown excellent repeatability with very small Z DUTs.

This fixture works well up to 8MHz limit of our IM3536 we've found, even with low Z (or higher Z components). We developed this specialized fixture for low Z SMD components, and it easily outperforms our best Tonghui Kelvin Clips for SMD devices, especially with very low Z components.

What will be interesting (if the custom PCBs ever arrive) is will driving the Guard section underneath the active PCB contact area (the IM3536 has Active Guard capability, which significantly improves Kelvin Clip use as previously mentioned) improves results, this area is vacant of any direct PCB conductor with the present custom PCB.

Anyway, in our case this unique SMD fixture has produced excellent results and displaced the normal SMD LCR Fixture and Kelvin Clips for low Z SMD component measurements. Still use the regular SMD fixtures for higher Z DUTs and rarely employ Kelvin Clips for anything other than leaded components (which is what they are intended for).

BTW curious as to your LCR Meter setups, please post some images

Best,

[EC8010]

Not sure "what leads" you are referring to?

I was thinking of the leads shown in the photograph in the first post.

This is my large box. The Kelvin clips are Mueller. The Hameg measures R + jX at 68 frequencies from 20Hz to 200kHz (excluding mains frequency) and reports directly to a spreadsheet. Spreadsheet subtracts a S/C calibration run, or on high impedance stuff an O/C calibration run of 5fF in parallel with the instrument's noise floor. Spreadsheet then least squares fits a model to the curve, calculates manufacturer's uncertainties and adds error bars. So far, I've about six different models for capacitors and inductors. It's noticeable that tightening the two screws at the bottom to tightly close the lid reduces measured uncertainties due to hum. The two tin plate screens slide against one another to give variable length.

[mawyatt]

That's a nice setup!!

You've effectively eliminated most of the disturbances by placing the DUT inside the shielded enclosure. Only issue might be the variation in movement of the DUT and Kelvin Clips & Leads between the Open Cal and Changing DUTs since within the box you have parasitic capacitance in all directions to the shield. The movable shield is clever, can you maintain the DUT and Kelvin Clips leads positions well enough for pF measurements with good repeatability?

Also understand your statement about a bar for support, that aluminum box looks pretty heavy hanging on those 4 BNCs!!

Nice work

Best,

[EC8010]

Thank you. Yes, there's stray capacitance to the shield. Unavoidable. For the smaller box, I lined it with 10mm thick hard foam. No idea what the plastic is, but (being foam) it will have a nice low permittivity and prevent the DUT getting too close to the shield. I didn't have any more hard foam to use for the large box.

Surprisingly, repeatability isn't a problem provided I'm careful not to let the movable shield get too close (<2mm) to the DUT and provided I arrange the DUT to hang (if its smallish) such that a little movement isn't a problem. The biggest problem is that the Mueller Kelvin clips don't always make a good contact - they have flat faces rather than corrugations to make their contact and their springs aren't all that powerful. The limiting factor for 1pF capacitances is actually the instrument itself. The impedance of 1pF at any frequency below 200kHz is quite high and that dominates the uncertainty at 1% (if I remember correctly). In general, the box allows the instrument to perform to its specifications. Which is nice.

The box itself isn't heavy, but I often test big heavy lumps like transformers in there, perhaps weighing a kilo or two, which would definitely be unkind to the BNCs. Those M5 screws with penny washers go into the rods.

[mawyatt]

We created some crude LCR Meter plotting software in Python, here:

https://www.eevblog.com/forum/testgear/lcr-meter-plot-software/

Obviously we're not that proficient in coding, and need to spend some time wrt the UI, adding GUI would be nice, but GUI is something we've never done before. Was "hoping" some clever software types would pickup on this but never happened, so it's in our "to do list", but way down the list!!

Best,

[EC8010]

You've got the tricky bit done; getting the data out of the instrument. That was done for me. But I wrote the spreadsheets that calculate impedance from the model and then allow the Solver to do the least squares fit to find model values.

I looked at your impedance plots - they're all familiar, although they go to a higher frequency than mine. ESR is pretty much constant with frequency. I think what you are seeing when it apparently increases at low frequencies is imaginary capacitance, which, being imaginary, is at right angles to real capacitance, putting it on the real plane and looking like resistance, but inversely proportional to frequency. It's in series with capacitors and quite a good indicator of the dielectric. The ideal capacitor has imaginary capacitance that is much larger than the real capacitance, thereby making its series contribution small.

[mawyatt]

V5 of the Special SMD fixture PCB arrive and assembled, see preliminary results here starting at #22:

https://www.eevblog.com/forum/projects/different-type-lcr-smd-fixture/msg5600601/#msg5600601

Best

[Martin72]

I have now written to them and asked for a data sheet, together with an explanation as to why.
The things look so similar, but it may well be that the cables from Sourcetronic do not come from TongHui or are “different” and are manufactured by TongHui for Sourcetronic.
We will know soon.

ST26011A 300kHz
ST26011B  5Mhz 300kHz
ST26011C 300kHz

The website will be changed accordingly.
But there will still be no data sheets.

[KungFuJosh]

ST26011A 300kHz
ST26011B  5Mhz 300kHz
ST26011C 300kHz

Did they note any differences between them?

[Martin72]

After they hadn't contacted me for 3 weeks, they called me today.
It was an interesting conversation...
They all had the same bandwidth of 300kHz.
I was like ah OK, and what about data sheets?
They didn't really go into that, instead they told me that they had looked at Keysight, among others, where it was also only 300kHz....
I think they didn't get any information from the actual manufacturer.
It will be some OEM stuff where they order their name on it.
Since the R&S adapters/cables look exactly the same, I would be inclined to ask them for data sheets.

[KungFuJosh]

Since the R&S adapters/cables look exactly the same, I would be inclined to ask them for data sheets.

They do look identical. https://www.rohde-schwarz.com/us/knowledge-center/videos/r-s-lcx-test-fixtures-video-detailpage_251220-1386305.html

However, I found nothing else on their website except for that video.

ETA: Here's the only "datasheet" I could find. More like sales brochure though: https://assets.testequity.com/te1/Documents/pdf/rs/lcx_datasheet_en_3609-8309-32_v0100.pdf

[mawyatt]

All those R&S fixtures look like Tonghui. Wouldn't be surprised if R&S LCR Meters are from Tonghui, probably similar to some TH types but with different front panels and maybe different firmware with same PCBs??

Best,

[mawyatt]

I have now written to them and asked for a data sheet, together with an explanation as to why.
The things look so similar, but it may well be that the cables from Sourcetronic do not come from TongHui or are “different” and are manufactured by TongHui for Sourcetronic.
We will know soon.

ST26011A 300kHz
ST26011B  5Mhz 300kHz
ST26011C 300kHz

The website will be changed accordingly.
But there will still be no data sheets.

Not surprised

Personally don't think these types of Kelvin Clips are good enough for repeatable results at 300KHz, certainly not the cheap clips you often see with lesser instruments!!

Best,

[Martin72]

I think so too.
The 5Mhz was simply utopian for this type of line.
With fixtures you can easily outperform this bandwidth.
Oh yes, the fixture from Sourcetronic is on its way.

[mawyatt]

BTW we did some measurements the other day with the cheap common 4 Wire Kelvin Clips and our DMM3500 for some low valued precision 2512 SMD chip resistors. Was a real PITA, even had a couple chips "flip out" with these clips

Did find them after crawling around on the floor the next morning tho

Then we used the Special Custom Split-Kelvin SMD Fixture with the DMM6500 and this was much easier, and we appreciated why we developed this Specialized SMD Fixture 

Also tried the Standard type SMD Fixture and couldn't get anything reasonable as the Zero Cal is too much dependent on the Zero DUT. Just not repeatable down to the levels necessary for good low milli-ohm measurements IMHO. This isn't surprising as these fixtures are more for higher Z components and the termination of the Force and Sense at the Plunger Base doesn't compensate for the variations in contact resistance at the plunger tip to DUT interface.

Best,

[Martin72]

All those R&S fixtures look like Tonghui. Wouldn't be surprised if R&S LCR Meters are from Tonghui, probably similar to some TH types but with different front panels and maybe different firmware with same PCBs??

Best,

I don't think so.
The LCX meters are manufactured in Malaysia, just like the inexpensive spectrum analyzers from R&S.
The test leads/fixtures will certainly be from Tonghui.
I wanted to buy an LCX100.
But only the LCX200 can be upgraded to 10Mhz.
However, all LCXs can be upgraded with a graphics plot.

[hpw]

BTW we did some measurements the other day with the cheap common 4 Wire Kelvin Clips and our DMM3500 for some low valued precision 2512 SMD chip resistors. Was a real PITA, even had a couple chips "flip out" with these clips

Did find them after crawling around on the floor the next morning tho

Then we used the Special Custom Split-Kelvin SMD Fixture with the DMM6500 and this was much easier, and we appreciated why we developed this Specialized SMD Fixture 

Also tried the Standard type SMD Fixture and couldn't get anything reasonable as the Zero Cal is too much dependent on the Zero DUT. Just not repeatable down to the levels necessary for good low milli-ohm measurements IMHO. This isn't surprising as these fixtures are more for higher Z components and the termination of the Force and Sense at the Plunger Base doesn't compensate for the variations in contact resistance at the plunger tip to DUT interface.

Best,

A week ago, I liked to measure a larger Tantalum Cap (as 2512 SMD sized) using my SMD fixture.

Conclusion:

- 2512 sized: impossible to have any contact

- 402 sized: impossible to have not slipped them under the contact needle

- also missing a nice leakage measurements of 30 old Tantalum (100u / 10V) ...
  Did then with an Siglent 4000 Power supply on 5V & 1K resistor & SDM3065 current graphs almost down to ground noise.
 

[mawyatt]

A week ago, I liked to measure a larger Tantalum Cap (as 2512 SMD sized) using my SMD fixture.

Conclusion:

- 2512 sized: impossible to have any contact

- 402 sized: impossible to have not slipped them under the contact needle

- also missing a nice leakage measurements of 30 old Tantalum (100u / 10V) ...
  Did then with an Siglent 4000 Power supply on 5V & 1K resistor & SDM3065 current graphs almost down to ground noise.

If you were using the typical SMD LCR Fixture, ones we have can be "opened up" to accommodate the 2512 by sliding the stationary plunger back giving more space. You can use a Zero Ohm 2512 for the Short Cal, or better yet a small section of #12 Cu solid wire, or small rectangular Cu plate.

Even with this effort it's still difficult to get a good Zero Cal reference for reasons previously mentioned, the Custom Split Kelvin SMD fixture is a better choice IMO.

0402 SMDs are indeed difficult. We found that some of the SMD fixtures don't have the plunger thin shaft to tip seated directly on the Insulated "V" shaped support. Placing "shims" under the support can help, we even 3D printed a custom "V" support for one of our cheap SMD fixtures.

If one is constantly working will small 0603, 0402 SMDs, might be worthwhile to create a custom 3D printed "V" support for this, even honing down the plunger tips to a smaller diameter (please don't do this to a quality fixture, use one of the cheap types!).

BTW how do you like the 4000?

Best,

[hpw]

If you were using the typical SMD LCR Fixture, ones we have can be "opened up" to accommodate the 2512 by sliding the stationary plunger back giving more space. You can use a Zero Ohm 2512 for the Short Cal, or better yet a small section of #12 Cu solid wire, or small rectangular Cu plate.

Even with this effort it's still difficult to get a good Zero Cal reference for reasons previously mentioned, the Custom Split Kelvin SMD fixture is a better choice IMO.

0402 SMDs are indeed difficult. We found that some of the SMD fixtures don't have the plunger thin shaft to tip seated directly on the Insulated "V" shaped support. Placing "shims" under the support can help, we even 3D printed a custom "V" support for one of our cheap SMD fixtures.

If one is constantly working will small 0603, 0402 SMDs, might be worthwhile to create a custom 3D printed "V" support for this, even honing down the plunger tips to a smaller diameter (please don't do this to a quality fixture, use one of the cheap types!).

BTW how do you like the 4000?

Best,

Yes, as at best to have various "V" shaped plug's for that matter.

>> SPD4121X

Well, still missing the service manual so how to custom calibrate as an opened task even no copy calibrated to user settings copy menu

The GUI somewhat different as RIGOL, so to think different. In addition, RIGOL one channel with low voltage.
I needed for oven 4 individual channels, as to go to four times 12V/1.5A.
IMHO also the connections and cables should have 4 individual color pairs, otherwise
red/black to label.