Cheap SMD LCR Fixture, the Good, Bad and Ugly!

[mawyatt]

Thought this might be a useful recollection after reading:

https://www.eevblog.com/forum/testgear/best-cheap-lcr-meter-for-small-capacitors-and-inductors/25/

We have a couple cheap non-OEM SMD Fixtures for use with lab grade LCR meters, OK know it's not wise to strap on a cheap fixture to a quality lab type instrument, here's the story. We did have a quality OEM set of Tweezers (Tonghui TH26009B) that helped with some SMD component selection and verification for a major project. When the precision SMD components we required became unavailable we were forced (actually major redesigns 4 times to accommodate available components!!) to use what was available and do selection. These were thousands of components and after using the OEM Tweezers for awhile (tedious as heck!!) decided to get a proper SMD Fixture, however the OEM SMD fixtures were unavailable with deliveries "expected" in many months, so we took a change on a couple available cheap SMD Fixtures that "looked" OK.

The build quality isn't great in both fixtures, but sufficient. After some time playing around before committing to actual use (precision component selection) they seemed OK, but a little concerned about the case lid grounding. One lid was completely painted, the other partially painted but both relied on mounting screws for grounding, fixed this with a knife & drill bit scraping some paint off. Also noticed the internal BNC grounds were not connected on one that had the star ground washers, so we corrected this by soldering in grounding wires. Then repositioned the internal wires towards the middle of the enclosure to minimize parasitic shunt capacitance, this is between the internal shield between the L and H sides.

After we suffered thru the measurements long ago we began to notice more uncertainty in the readings, especially in one fixture, then it began to fail DC Short calibration. We found the threaded insert that is orthogonal to the plunger contact, where the Force and Sense wires come together between the BNCs and plunger was slightly loose. So was the other plunger insert. The plungers were removed and these inserts well tightened and everything reassembled with good results again!! Then we began to experience some increased uncertainty with the other SMD fixture and found the same result, loose threaded insert. Both Fixtures perform well now and serve our needs.

A couple tidbits on getting repeatable results with these SMD Fixtures (probably holds true for the OEM fixtures as well). When you do a Short and Open cal the fixture moveable plunger moves, this also moves the internal wires slightly. For best results we often use a similar size SMD Zero Ohm for the Short calibration, then use a spacer on the plunger arm to open the contacts to about the same space as the expected DUT. Because the wires move, the internal fields in the fixture change slightly influencing the results. Using a similar size Zero Ohm SMD for Short Cal and same contact spacing for Open Cal you are helping keep internals similar for the DUT measurements.

Of the two SMD Fixtures shown the one in front with the 4 BNC lever arms is the better build and mechanically more stable, however both achieve very repeatable and comparable results on either of our TH2830 or IM3536 Lab LCR meters. Recall both were under $100, not a OEM replacement, but if one doesn't mind tinkering around then not too bad a alternative.

BTW the Ugly one has a grey 3D printed SMD support, we needed to replace the original because it was useless for small devices which is what we normally use.

Anyway, hope this helps some folks considering a cheap SMD Fixture for LCR meters.

Best,

[mawyatt]

Hers's what the insides look like. First is the better of the two cheap SMD Fixtures IMO. Note the baffle shield between L and H sides and how the Force and Sense wires go from the BNCs to the plunger insert, they are soldered here and this becomes the effective end of the Force/Sense capability. What this means is that the rest of the plunger from this point on becomes part of the DUT, and can only be compensated for by means of Open & Short calibration.

BTW it's a warm fuzzy when you can get a 0603 4.7pF COG to agree within 10fF between instruments & SMD fixtures (4.6912pF vs 4.6927pF), even when swapped (4.6905pf vs 4.6898pF). Repeatability between same instrument and fixture is even better


Best,

[mawyatt]

Upon viewing the second image above it became obvious that the internal wires could use a little trimming to shorten up and reduce the uncertainty regarding the internal E fields and reduce the series Z of the fixture. Honestly, don't know why we didn't do this long ago, but we didn't

Here's the internal view with shortened leads.

While playing round with this fixture decided to see how well it would work with a KS34465A DMM reading small capacitance, so gathered up some BNC adapters and cable and kludged (hooked) things up as shown in second image. Used the Math MX+B function and Long Averaging Function, and after allowing things to settle did a careful Null (the setup is floating so hand position can influence Null) to remove the residual capacitance (~68pF) with the fixture Open with a spacer in the Lever arm to position the plunger contact tips about the same spacing as the intended DUT, which in this case was the 0603 SMD 4.7pF reference cap mentioned earlier above. Believe it or not here's the result, 4.804pF with last digit bouncing round some but without any trickery

Then we decided to kludge up the adapters for the DE-5000 as shown in the 3rd image, note we used the Guard to the Fixture case ground. With the KS34465A the fixture "floats" and is more susceptible to surrounding effects, with the DE-5000 Guard utilized this effect is reduced considerably. Also note the DE-5000 is using direct Banana to BNC, this defeats the 4 wire Internal Capability but we don't have the unique Split-Banana plugs, so did the best with what was on hand. This resulted in a stable reading of 4.69pF!!

Edit: Added 1, 10, 100pF COG SMD Caps measured with TH2830 & DE-5000 with SMD fixture as shown. 1.01927/1.00pF, 9.95836/9.95pF and 101.080/101.17pF. This was done without an additional Cal on the DE-5000 after a couple hours off after initial Open/Short Cal ~3 hrs earlier, the TH2830 was left On the entire time of ~3 hrs with only initial Open/Short Cal.

Anyway, just some fun stuff with these fixtures and other instruments. As always, YMMV!!

Best, and Happy Thanksgiving

[mawyatt]

The kludge setup with the DE5000 above got us thinking. We acquired the DE-5000 long ago on the reliable recommendations provided (ones that had them and made comparative measurements), and planned on using this for some component selections (mostly small Cs), also got the TL-22 tweezers and then modified a TL-21 for proper Kelvin Clips. We found the Kelvin clips were not much better than the TL-21, also have a bunch of 6 1/2 digit DMMs with 4-Wire Kelvin capable and cables so no need for precise resistive measurements from the DE-5000 which isn't anywhere near the DMMs anyway. So the Modded TL-21 with attached Kelvin clips just never got used as intended.

After realizing we could use the SMD Fixture with the DE-5000 as shown above, and get good results (mainly for SMD capacitors), the thought of making an adapter with the TL-21 to 4 BNCs for use with the SMD Fixtures. Also, the core DE-5000 is a true 4 wire system, so we didn't want to defeat this as we did above with the BNC kludge adapters. So we cut the Kelvin clips off the modded TL-21 and reconfigured it with 4 BNC female connectors (could have used male but then would require the BNC barrels) we had on hand. The cables were kept short, maybe a little too short as you can see, and the workmanship isn't good (need to redo this sometime), but the resul is very good!!

We did a cal and pulled a Venkel 0805 COG 0.3pF reference cap we measured with the TH2830 at 0.31914pF, the DE-5000 shows 0.31pF!!

Best

[joeqsmith]

I like the idea of a spring loaded fixture like these for rapid change out.  I built something similar but for characterizing parts in several MHz.    They work alright.  You can see the one in the first post.     

https://www.eevblog.com/forum/rf-microwave/shunt-capacitance-of-1206-smd-resistors-jeroen-belleman-december-2010/

[mawyatt]

Nice DIY fixture!!!

Agree, these work very well and much preferred than having to solder the DUT into a test fixture, then insert fixture into measurement setup, then remove the test fixture, then unsolder the DUT to utilize.

The OEM types of these SMD Fixtures, like the Tonghui TH26008A, claim a frequency range of DC to 120MHz, although we only have a LCR meter (IM3536) capable of 8MHz.

We've utilized these SDM Fixtures to cycle thru many devices (more than I care to remember!!) and they made quick work of a boring procedure and produced good repeatable results!!

For under a $100 and a little rework not a bad addition to one's fixture tool kit, apparently even if they have a handheld LCR meter

BTW the design of these SMD fixtures is not obvious from the images. They use a nicely machined (brass?) and plated "plunger" that slides inside a hollowed tube insulated sleeve that extends longer than the plunger. The tube slides inside the larger shinny (silver color) plated metal support (brass?) which is secured to the Lid. The plunger insert has a deep threaded hole orthogonal to travel, where a long brass thin bolt or rod is attached thru the insulator tube which has a hole. This rod fits within a ~ 13mm slot cut in the Al top cover, on the underside two wires, Force and Sense are soldered to the rod end and go to the BNC connectors, and allows the plunger to slide along the slots.

One plunger side is held with a thumb screw pressing into the insulating sleeve, the other is arranged with a spring loaded arrangement and a lever arm and the insulated sleeve extends well past the end to hold the Spring and Lever Arm. So the only active metal is the two plungers and they are arranged so the spring loaded lever plunger is on the Low or L side, and the other the H side of the LCR Meter. Very clever arrangement that is mechanically & electrically stable by maintain good control of the fields inside and outside the fixture. Likely this may have been copied from an old HP LCR fixture. 

Here's what the inside of the modified TL-21 looks like.


Best,

[joeqsmith]

Here's my shunt fixture.  The Teflon looks hacked up because it was cut with a knife. 
https://www.eevblog.com/forum/projects/20db-rf-attenuator-seeking-feedback-to-improve/msg2974530/#msg2974530

A member here was trying to make a fixture with pogo pins.  I'm not sure how that worked out for them.   

I have used some test PCBs to characterize parts by soldering them.   Agree, it's a pain and having to insert the test board after changing components is hard on connectors.   My balls are not big enough to solder with the test board still attached to the test equipment.   

Istvan Novak published three articles in EDN on low cost fixtures.  The very first was using solder wick and coax to make shunt thru measurements. 
https://www.edn.com/solder-wick-trick-characterizes-bypass-caps/

[nctnico]

I have used some test PCBs to characterize parts by soldering them.   Agree, it's a pain and having to insert the test board after changing components is hard on connectors.   My balls are big enough to solder with the test board still attached to the test equipment.   

I have created some boards with SMB connectors for such purposes. Quicker to connect / disconnect. But there are also plug-in SMA adapters. Recently I received a few of those.

[mawyatt]

One of the nice things about these fixtures is they are 4 wire Kelvin type fixtures, so the cabling, connectors and such are effectively removed. The Force and Sense wires converge at the threaded rod tip just before it's screwed into the tapered plunger, so this leaves just the tapered end of the plunger from the thread rod not sensed on each plunger. When doing the LCR Open and Short Calibration this finishes up the fixture normalization task fairly well. Some of the lab grade LCR meters utilized SD ADCs (up to 24 bits I believe) and can resolve into the Giga ohm impedance magnitudes, so can accurately resolve small capacitances at low frequencies, this is what we needed for work as everything was high impedance.

Anyway, this little DE-5000 LCR meter has surely earned it's keep along side the two lab grade meters, one of the better instrument values around IMO.

Best,

[mawyatt]

I have created some boards with SMB connectors for such purposes. Quicker to connect / disconnect. But there are also plug-in SMA adapters. Recently I received a few of those.

Have those as well, think just about everyone does, but way too much of pain to use, especially considering how repeatable these SDM fixtures behave when properly setup and utilized.

Best,

[joeqsmith]

Using the VNA to make shunt thru measurements is similar to your low frequency Kelvin type connections.  The problem is the two ports of the VNA share a common reference which causes a loop and leads to errors at lower frequencies.  At higher frequencies, the cables are lossy which takes care of the problem.  One way to solve the ground loop problem is by placing a common mode transformer in-line with the measurement port.  The problem is designing such a transformer. 

You mentioned starting out using the tweezers.  I have a set of Kelvin tweezers for my RLC meters.   They are fine for larger surface mount parts.  Not sure using them would be faster than the fixture you're using.   Sounds like in your case, they were not.  In one case, I guess you have to place the parts on a non-conductive work surface and clip on.  The other, you have to insert them into the fixture.    Certainly both options are faster than soldering.

[mawyatt]

Many of the SMD were 0603 even 0402, had difficulty handing these small SDM with the OEM Tonghui Kelvin LCR tweezers, very tedious and error prone, and the cheap no-name types are really difficult. With the SMD fixtures you just pull the lever, remove the previous device (use small SS quality tweezer like Techni-Tool), grab and drop in the next SMD and get the measurement. The fixture Insulated "V" groove places the SMD in the correct position for the plunger to engage. On one of the SMD Fixtures we 3D printed a custom "V" groove insulator because the one provided wasn't working well with the smaller SMD components, this is the Grey V Insulator in the images above.

For low Z components like very small inductors (nh), the VNA is a better instrument, but our precision needs were Cs and Rs, not Ls. We even had to study the voltage effects & thermal (self heating) on such, may recall we developed a fixture to allow supplying an external bias voltage with the LCR fixtures, but that's another story!!

Regarding the transformer, yep that's a difficult problem, ran into a similar issue with the Open Loop Bode Plots using a transformer made from a CM Core.

Best,

[joeqsmith]

I wouldn't want to use the tweezers I have for parts smaller than 0805.   I have used them with 0603 but it's a real pain.   Even with the fixers I made, I would need to rethink how to hold smaller parts like this as what I have now would certainly allow them to twist.   The V-grove is a good idea.  In my case, I have different shims made out of coax with various thickness.  For 0603, you have to hold the part you want to test in place while letting the spring put compression on it.   

PDN measurements like I mentioned will be mostly made of Cs, well, and Ls but the Ls may be the Cs acting as Ls.. The same for Rs..  Most of my lower frequency designs don't use very small Cs or Ls.    Similar to your external biasing of the RLC fixture, we do this with PDN measurements as well.   After all, you have to power up your device at some point..    With the VNA you can see how the impedance changes with voltage and how stable it is with frequency.

[precaud]

Also noticed the internal BNC grounds were not connected on one that had the star ground washers, so we corrected this by soldering in grounding wires.

Note that some LCR meters do not want all of the BNC barrels to be grounded and interconnected at the fixture, so check before you do this. HP's do, but Wayne Kerr comes to mind as one that does not.

[mawyatt]

Also noticed the internal BNC grounds were not connected on one that had the star ground washers, so we corrected this by soldering in grounding wires.

Note that some LCR meters do not want all of the BNC barrels to be grounded and interconnected at the fixture, so check before you do this. HP's do, but Wayne Kerr comes to mind as one that does not.

Thanks, we did!!

Both of the Lab Grade LCR meters we have (Tonghui TH2830 & Hioki IM3536) all input BNC grounds are connected, and on the IM3536 they are also connected to the "Guard" Input, the TH2830 has no "Guard" Input.

Interestingly, the TH2830 has the BNC grounds also connected to the Mains Ground return, while the IM3536 does not, it's galvanically isolated.

Best,

[ExaLab]

I received my test fixture a few days ago. It appears to be exactly the same product presented by mawyatt in previous posts (to be exact, the one described as the better of the two...)

As can be seen from the image, the right support is isolated from the case by means of two insulating washers.
This solution, in my opinion, has no reason to exist, given that it exposes the measurement system to inevitable interference from the surrounding environment. I will remove the two washers making sure the related support is electrically connected to the case.

What do you think?

[mawyatt]

Unless the plunger assembly is different than what we have those bolt insulators should not be used. The plunger assembly has the center "shaft" insulated from the plunger casing with the white plastic coaxial pieces in each plunger assembly. The conductive plunger shaft sits inside the coaxial plastic surroundings, so the plungers are isolated from the plunger casing and aluminum box.

You want everything at ground potential inside and outside the box as best you can, this also means the plunger case and  box "lid" where the attachment screws create a ground connection to the box. The box is at ground due to the 4 BNC connectors. Make sure the 4 leads (2 pairs) that go thru the box to the plunger are tight, these were loose on one fixture version we had and caused measurement repeatability issues. The leads are soldered to a single long threaded brass insert that screws into each plunger shaft, this was loose and needed tighting.

Also make sure everything is mechanically stable including BNC connectors. Having a stable electrical and mechanical environment as "seen" by the DUT is important for accurate and repeatable measurements.

Once you have reworked these inexpensive fixtures, they preform well we've found

Best,

[ExaLab]

Thanks a lot for your prompt reply.
As you suggest I'll remove the two insulating washers and I will check that everything is tightened properly...

I'll use this tool with my Fluke PM6304 (and may be with the ET432 too). To do this, I'm going to make a pair of interface adapters.

I measured the distance between the BNC connectors of this tool and it seems to be exactly 22.0mm (66.0mm exactly between the central pins of the two outermost connectors).
The doubt I have is the following: is this the standard in use (i.e. a pitch of 22.0mm) or rather the most used standard is the 7/8" (i.e. 22.225mm)?
When making interface adapters, to increase their compatibility with possible further tools, I would like to be sure to use the most commonly used pitch!

On your branded LCR analyzer could you please measure with a caliper the distance between the two outermost connectors and tell me if it is closer to 66.0mm or 66.7mm? I don't have analyzers that use this type of BNC interface (My LCR meter PM6304 uses a proprietary interface based on a Lemo connector...).

[ExaLab]

I finished reworking the test fixture and the result is very satisfying.
Here are some of the upgrades made (some of them absolutely necessary!!):

1) The right nylon insulating cylinder was reworked (it had an anomalous hole in correspondence with the tightening knob which caused the short circuit between the measurement contact and the case...)
2) To avoid excessive mechanical play, I replaced the M3 screw of the lever with a more suitable M4 one
3) Replaced the rigid PVC internal cables with extra-flexible silicone sheathed cables
4) Replaced the four support rubbers with more suitable and professional products
5) Repainted the entire unit with two-component polyacrylic varnish (RAL 7035). I did this last operation at the same time as painting the containers used for the two interface cables.

For the pair of interface cables I used Hammond die-cast aluminum enclosures (1590G2) and Amphenol BNC connectors.

Now I have really stable and repeatable measurements with both LCR analyzers at my disposal.
Thanks to those who presented this low cost test fixture (with many defects but... all fixable).

[mawyatt]

Very nice work!! Like the cables and the box approach with the handheld.

Must apologize for not responded to your question, completely missed it

We also measure 66mm, and 22mm between BNC centers.

An area we've had some difficulty in making quick but repeatable measurements is with some SMD devices where the connection pads are on the bottom, common to some surface mount Polymer Caps, Tantalum and Inductors we use. These required some fiddling to get just right in the SMD fixture "V" groove, the "V" we 3D printed works better for these devices but still isn't as good as we would like.

Been thinking about this for awhile now, and liked the way the plungers Lever Arm worked with our custom Tek577 fixture shown below.

https://www.eevblog.com/forum/projects/smd-test-fixture-for-the-tektronix-576-curve-tracer/msg4553639/#msg4553639

So we came up with a similar type fixture to use with these SMD devices and bench type LCR meters. PCB should be heading to fab house this week along with some other projects. This allows the plunger Lever Arm to be on either side and the PCB can be flipped over for another set of surface contacts for larger devices. For first try we'll just go with standard plating, and if it works well may consider gold plating for the contact area.

Best,

[ExaLab]

Cables inside the sheath and box approach are absolutely necessary for repeatability of ac measurements (especially when measuring low impedances where the "weight" of the cables inductance is more evident...).

Thanks for your kind reply. For the two boxes I used the pitch measured on the test tool at my disposal, i.e. 22.0 mm. (I've been lucky!!)

[mawyatt]

For the best results the cables should be low capacitance shielded types, where each signal (L-Cur, L-Pot, H-Pot, H-Cur) are within individual coaxial shield cables. Placing these 4 cables inside a sheath is a good idea and reduces measurement uncertainty due to cable movements. Using a box as the breakout for the BNC to fixture connectors provides better mechanical stability and thus better measurement repeatability.

As a side note, in the SMD test fixtures, when one does the "Open" calibration the calibration should be done with the fixture plunger spacing the same as the expected DUT. The reasoning is that the plunger fringe capacitance should be the same as when the DUT is evaluated, and the internal two wires on the plunger movement side (L-Cur and L-Pot) slightly move and the fringe capacitance to the fixture interior case (ground) slightly changes. Same goes for the calibration "Short", use a Zero Ohm Short of the same spacing as the expected DUT, here we usually use Zero Ohm SMD devices.

All good concepts for better repeatable measurements

Best,

[ExaLab]

The use of a box for bnc connectors has a big advantage: it prevents movements between the "Force" and the related "Sense" cables, thus avoiding altering the short circuit trim parameters.

However, I do not agree with the procedure you use for the "Short Circuit Compensation".
For this step you must not use any thickness between the measurement points, otherwise you lose an important parameter of the DUT which substantially coincides with the equivalent series inductance associated with its physical length. (as example, think to the series resonance of a capacitor... ).
More, during SC Trimming the effects of the fringe capacitance to the fixture interior case is absolutely negligible (due to the voltages near zero...).
Even leading companies like R&S and Keysight require short circuit trimming with the tips touching each other.

Personally, in some specific cases, in order to avoid including in the measurement the contact resistance between the tips and the DUT,
I carry out the short circuit trimming by interposing between the tips a thin layer of tin (of negligible thickness compared to the length of the DUT).

Different ideas that I hope will help both...

[Kean]

As a side note, in the SMD test fixtures, when one does the "Open" calibration the calibration should be done with the fixture plunger spacing the same as the expected DUT. The reasoning is that the plunger fringe capacitance should be the same as when the DUT is evaluated, and the internal two wires on the plunger movement side (L-Cur and L-Pot) slightly move and the fringe capacitance to the fixture interior case (ground) slightly changes. Same goes for the calibration "Short", use a Zero Ohm Short of the same spacing as the expected DUT, here we usually use Zero Ohm SMD devices.

I was recently measuring some single digit pF 0402 caps on my CKT5000 LCR meter with Tonghui SMD fixture, and I also realised I needed to space the fixture tips in a consistent manner for the open calibration.
I used a small offcut of 0.030" (0.75mm) thick styrene sheet and got the high precision measurements that I expected for those tiny caps that I subsequently lost... 

I didn't use a 0R SMD part for the short calibration as I thought the imperfections in that part and the extra contact resistances would be more significant than getting the spacing right - but I didn't waste time on testing that theory.   

[TERRA Operative]

I have one of these test fixtures too, the cheapest I could find on Aliexpress, and with some tender love, it works quite well on my HP 4276A LCZ meter.

I improved the grounding inside, cleaned out the overspray from inside with acetone to allow proper contact to the case for the BNC's and the divider panel, tightened everything up, and stuck it all back together with stainless steel screws just for a bit more pretty.

I am wondering though, is it worth using coax (shield connected at the BNC end) for the internal wiring?
Maybe using some RG316 or similar. Would this help improve things, or not worth the effort?