Author Topic: Different type LCR SMD Fixture  (Read 12557 times)

alpelectronics, Vgkid, georgd and 1 Guest are viewing this topic.

Offline gitm

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #50 on: January 23, 2025, 04:20:18 pm »
Where did you get the fine Cu powder?

Amazon.

https://www.amazon.com/gp/product/B0CZ9SGDG7/ref=ppx_yo_dt_b_asin_title_o00_s01?ie=UTF8&th=1

Thought about gold too, but it would have to be hard gold and not ENIG.
« Last Edit: January 23, 2025, 04:24:36 pm by gitm »
 
The following users thanked this post: mawyatt

Offline ZGoode

  • Regular Contributor
  • *
  • Posts: 216
  • Country: us
  • Grad student by day, equipment nerd also by day
Re: Different type LCR SMD Fixture
« Reply #51 on: January 23, 2025, 06:09:44 pm »
I've been trying out copper powder from our LPBF printer for this.  Currently, I'm waiting on a new batch of PCB's to apply this to before I take some measurements.  If anyone is interested I could try and send some to people for testing (just pay for shipping (and probably USA only, not sure how customs would handle this)).  This stuff has pretty tight size tolerances and is nearly pure.  I don't remember the exact metallurgical compound, but I do know it is 40um particle size +/- 5um.
 

Online mawyattTopic starter

  • Super Contributor
  • ***
  • Posts: 4196
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #52 on: January 23, 2025, 08:21:49 pm »
This is getting interesting indeed!!

How these various Copper filings and powders improve surface connectivity to the SMD end terminals vs the standard PCB surface prep should be of interest too many folks.

For the Short Calibration we use a cut piece of thin Copper plate same size as 2512 Resistor, this has worked well and has good repeatability.

Best 
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 

Online Kean

  • Supporter
  • ****
  • Posts: 2565
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Different type LCR SMD Fixture
« Reply #53 on: January 24, 2025, 06:29:23 am »
The copper powder seems like it would just get "absorbed" into the solder, but I guess some gets left behind after the wicking to leave a rougher surface.

Another option that might be worth a try is gold plated SMD pads used for probing.  I have some, but not yet ordered any of Mikes SMD fixture PCBs.

Look on Digikey under Spint Loaded Contact Pad.  This link might work for you, then filter by type Contact Pad.  Many different sizes.
https://www.digikey.com/en/products/filter/contacts-spring-loaded-pogo-pins-and-pressure/311
 

Online mawyattTopic starter

  • Super Contributor
  • ***
  • Posts: 4196
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #54 on: January 24, 2025, 03:02:33 pm »
The copper powder seems like it would just get "absorbed" into the solder, but I guess some gets left behind after the wicking to leave a rougher surface.

Another option that might be worth a try is gold plated SMD pads used for probing.  I have some, but not yet ordered any of Mikes SMD fixture PCBs.

Look on Digikey under Spint Loaded Contact Pad.  This link might work for you, then filter by type Contact Pad.  Many different sizes.
https://www.digikey.com/en/products/filter/contacts-spring-loaded-pogo-pins-and-pressure/311

Those gold plated contact pads look interesting, seems one could use such for the Zero Ohm Calibration DUT!!

Best,
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7333
  • Country: fi
    • My home page and email address
Re: Different type LCR SMD Fixture
« Reply #55 on: January 24, 2025, 11:20:04 pm »
Harwin S70 series of contact pads for pogo pins are rated at 0.15mΩ max contact resistance.  Mouser lists them under Passive Components > EMI Filters / EMI Suppression > EMI Gaskets, Sheets, Absorbers & Shielding; for example S70-125161545R (2.5×1.6×0.15mm, 6.0A max, 0.050µm gold over 2µm nickel over beryllium copper; Harwin product page).

I wonder if these are too hard?  The gold layer is definitely shallow; a thicker layer of pure gold with a fine pattern would be quite soft and malleable as a contact.  Gold leaf is too thin also (typically on the order of 0.1µm).

Here's an utterly crazy idea: Use short lengths of pure gold (99.9%, 24 karat) #18 to #28 gauge wire.  These are sold in jewelers' supplies, and aren't as expensive as one might think (about 2€ per mm, depending on the gauge).
 

Offline gitm

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #56 on: January 28, 2025, 01:44:56 am »
Just closing off the copper screen comment I made above.  For those wanting a more evenly distributed, finer grain but rough finish on the PCB SMD pads , I tested the following modification to the technique described above, producing pretty good results.

  • Changed from lead solder to ChipQuick NC191LTA10 Sn42/Bi57/Ag1 No-Clean T4 and ChipQuick CQLF4 No-Clean liquid solder flux.
  • Make a powder dispenser using an empty bottle, e.g. old pill bottle, partially fill with #320 mesh size copper powder and put a #200 mesh size copper screen over the top, secure with a rubber band.  You could use a finer screen for more control at the expense of time.
  • Wet pads with solder flux.  My PCB was HASL finished.
  • Lightly dust pads with copper.
  • Touch soldering tip (400C) to pad letting heat move through the pad.
  • Add a tiny dab of solder paste, melt with iron and let it spread across the tab.  If it looks like there’s excess solder, use a wetted solder wick to remove excess.
  • Wipe with IPA to clean loose copper and flux before repeating.
  • Repeat the above steps to get the desired surface finish you’re looking for.

Picture of finish after 4 iterations is attached.  EZ-Peasy.

With the above said, I haven’t tested whether this finish will produce more repeatable measurements than my first board, which produced excellent results (I’ll post those once I type them up.)


 
The following users thanked this post: Kean, RoGeorge, Nominal Animal

Offline RoGeorge

  • Super Contributor
  • ***
  • Posts: 7094
  • Country: ro
Re: Different type LCR SMD Fixture
« Reply #57 on: January 28, 2025, 07:50:37 am »
Would be interesting to see if the surface ruggedness can survive after repeatedly pressing components against it.

I wonder if self-adhesive foil, copper or aluminium, laid on a rubber substrate (instead of FR4) would work.  ::)
The hope with that would be to get the equivalent of a "rubbery" metal, but I didn't try.

Online mawyattTopic starter

  • Super Contributor
  • ***
  • Posts: 4196
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #58 on: January 28, 2025, 03:08:14 pm »
The pressure involved is very slight since the end terminals of the DUT make the Force and Sense connections on each end of DUT, thus the DUT Contact pressure/resistance for the Force current signal becomes much less critical. The Sense terminal has such a high relative impedance that the DUT Contact pressure/resistance is also much less critical.

These effects make the wear on the physical DUT contact area on the PCB much less critical than a normal measurement. If wear does take a toll on the measurements then one can just resurface the area again, but haven't had to do this yet after many thousands of engagements/measurements.

The PCB design is such that one could use a connector (it's laid out for this), but we just soldered the 4 wires directly. This appears the same for gitm and Josh uses as well. The only component on the PCB is just the connector which is a cheap JST type, so one could just swap out the PCB after surface wear out as they are throw away cost-wise.

Josh used a gold plated process for the PCB which could wear thru and require PCB replacement but suspect this would be after many many uses and then the PCB could simply be replaced as mentioned, the PCB would cost more but still in the throw way price range.

For those that have a quality bench type LCR Meter or quality DMM and want highly repeatable measurements with SMD DUTs, this "Split-Kelvin Technique" just works, and even more so with low Z milli Ω DUTs.

Best
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 
The following users thanked this post: RoGeorge

Offline gitm

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #59 on: January 28, 2025, 09:15:34 pm »
Repeatability with low Z resistors:
=====================
I wanted to know, in general, how repeatable are measurements using the fixture in both time and position of the DUT on the SMD pads.  That is, knowing what to expect and how much attention is needed to what details?  Of the things I could control were measurement setup, warm up time, and the environment as found.  Although tested in a “controlled” environment, I only have limited, indirect control of instrument repeatability through it’s configuration and warmup.

I imagined a big part of repeatability is consistent contact resistance which would vary depending on SMD pad flatness/contact area with the DUT and where the DUT ended up on the pads.   So the testing was setup to explore this point.

Procedure:
  • Use 4 “precision” low Z resistors: 1 milliohm, 10 milliohm, 100 milliohm, 2.4 ohm, all in 2512 size.  Note: the 100 milliohm “metal element” resistor’s tappered contacts required a bead of solder to be layed down on the resistor’s underside; without this bead, there was no contact made with the SMD pads at all.
  • Use a “precision” low resistance meter CKT515A and SDM3065X (a.k.a., use what you got ;) )  The 3065X is not really suitable for measurements 100 milliohms and below due to a +/- 0.005% of range specification, but still informative for measurement variations; it’s 200 ohm minimum range translating to +/- 10 milliohms deviation and still in specification.  Due to this, measurements were grouped only once for each milliohm resistor on the 3065X, while CKT515A had many more measurements.
  • Measurement “groups” would be made alternating between the CKT515A and 3065X to require the fixture be repeatedly detached and attached from one to the other.  After all, this would happen in the real world.
  • 3 hours warmup with both machines powered on during testing across 48 hours.
  • Each instrument was configured for high(er) accuracy, e.g. 16 averages, 100 NPLC, statistics turned on with 25+ measurements, relative on, etc., ensuring stable readings and no drift.
  • The 1, 10, 100 milliohm resistors would be measured at various times over two days in groups of 6 measurements, with each measurement in a differing location from the last.  The locations were named “center”, “right”, “down”, “up”, and “left”.  “Center” placed the 2512 resistor centered directly over where the four corners of the pads meet so that equal resistor dimension were over all four pads.  “Up” and “down” were horizontally centered but 2/3 of their height placed above or below, respectively, of the horizontal center line with the 1/3 on the other side.  Similarly, “right” and “left” were 2/3 to the right or left, respectively, of the vertical center line.
  • The sequence of resistor placement was center, right, down, up, left, and back to center, for 6 measurements.  This gives some insight into location dependance on the reading value and direct comparison for the center to center measurement variance.  For CKT515A measurements, each resistor had the group of 6 measurements repeated twice on the same day but 6 hours later.
  • 2.4 ohm readings were performed immediately before and after each of the milliohm resistor groupings across the 2 days.  No significant attention was placed on placement other than what one might when casually inserting into the fixture.  This 2.4 ohm resistor has been measured many times before this test and it’s value well known, averaging 2.4078 ohms previously measured overnight with +15k readings.
  • Compile measurements and compute basic stats.

Results:
To me, the most telling metric on repeatability is the extreme spread (span) on the milliohm resistors since contact resistance variation would be a more significant value when compared to the actual resistor value and be more readily apparent.  The tables below show the span and % of the mean for the milliohm measurements.  The attached screenshot also shows each’s position readings one above the other which helps in showing (or not) of any patterns due to position.
———————————
1 milliohm CKT515A 20 mΩ range => 0.1%*Rd + 0.0025%*Rg => +/- 0.0015 mΩ
CKT515A 1 mΩ max span:0.0060 mΩ
CKT515A 1 mΩ min reading:0.9809 mΩ
CKT515A 1 mΩ max reading:0.9869 mΩ
CKT515A 1 mΩ mean reading:0.9830 mΩ
CKT515A 1 mΩ span as % of mean:0.6104 %
———————————
10 milliohm CKT515A 20 mΩ range => 0.1%*Rd + 0.0025%*Rg => +/- 0.0105 mΩ
CKT515A 10 mΩ max span:0.0210 mΩ
CKT515A 10 mΩ min reading:9.9792 mΩ
CKT515A 10 mΩ max reading:10.0002 mΩ
CKT515A 10 mΩ mean reading:9.9900 mΩ
span as % of mean0.2102 %
———————————
100 milliohm CKT515A 200 mΩ range => 0.1%*Rd + 0.002%*Rg => +/- 0.1040 mΩ
CKT515A 100 mΩ max span:0.0530 mΩ
CKT515A 100 mΩ min reading:100.0010 mΩ
CKT515A 100 mΩ max reading:100.0540 mΩ
CKT515A 100 mΩ mean reading:100.0197mΩ
span as % of mean0.0530 %
———————————

For the 2.4 ohm resistor, the 3065X was not nearly as hindered by it's accuracy specification of 0.005% * 200 ohm range; it performed very well with an extreme spread of only 0.5 milliohms for 7 measurements over 2 days. 

CKT515A readings on 2.4 ohms had a span of 2.3 milliohms or +/- 1.15 milliohms.   It’s accuracy spec for the 20 ohm range is 0.02% of reading + 0.002% of the 20 ohm range, or +/- 0.880 milliohms, putting it some 270 micro ohms above it’s spec.

To be honest, I think this test was a worst case scenario but still relevant to understand how well the fixture works.  And I am very pleased with it.  Great design.

Attaching screenshot of measurements taken and resistor specifics along with two pictures.  One picture being the testing pads after some 100+ measurements.  There doesn’t appear to be signs of wear, not that I was expecting any.  The second picture is a more oblique picture of the new PCB fixture pads to get a better perspective of the new rougher texture; I feel the close up posted above in my other post doesn’t convey how subtle yet rough it actually is — very grippy on the finger.
« Last Edit: January 29, 2025, 03:36:45 pm by gitm »
 
The following users thanked this post: Kean, RoGeorge

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7333
  • Country: fi
    • My home page and email address
Re: Different type LCR SMD Fixture
« Reply #60 on: January 29, 2025, 12:44:53 am »
The tables below show the span and % of the mean for the milliohm measurements.
Note that it would be more useful to use median (50th percentile; the value where half the measurements are below, and half above) here, as usual: mean is unduly affected by outliers.
 
The following users thanked this post: gitm

Offline gitm

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #61 on: January 29, 2025, 02:05:24 am »
Note that it would be more useful to use median (50th percentile; the value where half the measurements are below, and half above) here, as usual: mean is unduly affected by outliers.

Good thought which I had too.   Between using median and using mean, I chose mean because the distribution on the data was pretty symmetrical.  Median would have been my choice if the distribution was skewed.  But being pretty symmetrical, using mean would highlight a small number of outliers, if present, which, to my mind, is more descriptive.

ETA: The reason it’s more descriptive is that the testing is looking for some noticeable change in a reading that should be fixed and might show up only in a minority of or even just 1 sample.  I.e., the sample that gives information about the fixture’s repeatability, like where a location had inconsistent contact resistance.  Mean, as you note, is unduly affected by outliers, but median ignores those few or even 1 outlier that I’m looking for.  Hope this clears up my point.
See next post.

As a practical matter, it’s moot as the calculation using either mean or median return essentially the same results with the distribution being as symmetrical as it was. 
For example:
  • 1 milliohm: mean=0.9830, median=0.9826, and span as % = 0.6104% and 0.6107% respectively.
  • 10 milliohm: mean=9.9900, median=9.9904, and span as % = 0.2102% and 0.2102% respectively.
  • 100 milliohm: mean=100.0197, median=100.0310, and span as % = 0.0530% and 0.0530% respectively.

I also considered comparing the span to the anticipated span (I won’t use “expected” due to its mathematical definition) based on the accuracy specifications for each range, but the calculations were a bust with 2 out the 3 results having essentially the same span either way.  The third was only 3 micro ohms over with the measured span of 6 micro ohms.  I might have had to make many, many more measurements to pull out any difference.  In the end though, I think it all points to the fixture enabling very repeatable measurements.
« Last Edit: January 29, 2025, 01:49:34 pm by gitm »
 
The following users thanked this post: Nominal Animal

Offline gitm

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #62 on: January 29, 2025, 02:26:06 pm »
Note that it would be more useful to use median (50th percentile; the value where half the measurements are below, and half above) here, as usual: mean is unduly affected by outliers.

Thank you Nominal Animal.

Having slept on this, I’m going to reverse myself and agree with you.  Median would be more useful.  The outliers I’m looking for are captured in the extreme spread (span).  Ideally, I want to compare this value against the resistor’s “true” value.  But we don’t have a “true” value, we have only mean and median values.  In this testing, the median does appear to be closer to the “true” value than the mean. 

Why? All measurements would have contact resistance included.  The key aha moment was to realise there’s a practical lower limit on contact resistance, but the upper limit is essentially unbounded.  In a perfect world with perfect contact on all measurements, mean and median would be equal and the measured resistance would be at a minimum within the uncertainty of the instrument’s capability.  But as reality creeps in and contact resistance inexorably increases in an almost random pattern, it still can’t go below the lower limit.  The measurements would thus be skewed above resulting in the median trending lower than the mean and closer to the “true” value.

It boils down to wanting the outliers in the extreme spread but not in the estimate of the “true” value.  In this context, median is more useful.  Thus outliers would be more apparent using median over mean.

Took me a while to see it.  Thanks.
« Last Edit: January 29, 2025, 02:44:51 pm by gitm »
 
The following users thanked this post: Nominal Animal

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7333
  • Country: fi
    • My home page and email address
Re: Different type LCR SMD Fixture
« Reply #63 on: January 30, 2025, 05:08:04 am »
Well, I can accept both arguments myself, as it really is about what information you consider most useful and important to convey to others.

I mostly deal with noisy datasets (with obviously erroneous measurements), especially microbenchmarking results for various algorithms (where events external to the benchmark will occasionally affect the measurement, and such errors are always relatively large, as there is a minimum duration for any kind of interruption in the benchmark), and have found the most useful report to be of k'th percentiles; as in, 97% of all runs completed within time T.  The percentile I choose reflects my trust in the measurement setup; the higher, the more repeatable/reproducible (and thus better) the setup.

For other kinds of measurements, giving the actual distribution (or histogram) of measurement is best, of course; that way one can best compare reproduced measurements, and any asymmetry and deviation from standard/Gaussian distribution tells something about the measurements and the values being measured.  The next best thing, in my opinion, is form \$m^{+a}_{-b} ~ (p\%)\$, where \$m\$ is the median, and \$a\$ and \$b\$ are the symmetric \$p\$'th percentile error bounds.  That is, \$p\%\$ of all measurements were between \$m-b\$ and \$m+a\$, inclusive; with \$(100-p)/2\$ percent of measurements below \$m-b\$, and \$(100-p)/2\$ percent of measurements above \$m+a\$.  While the standard 68.3% error bars are well suited for standard/Gaussian distributions, one can choose a much larger \$p\$ to describe ones confidence in the testing methodology: the higher the \$p\$, the fewer measurements are rejected, and therefore the better the confidence in the implementation of the measurements.

I'm not a metrologist, I only watch an occasional Oxtools and Joe Pie videos on Youtube.  In all my posts in this thread, by "useful" I have meant "useful to those like myself, who use the reported numbers to build up an intuitive understanding of the properties and limits of this method/apparatus/approach".  If I've understood correctly, the numbers you/gitm have reported, \$p = 100\%\$, with \$\text{span} = 100\% \cdot (a - b) / m\$, and \$m\$ also reported.  To get a correct mental picture of the limits of the distribution, leaving its shape unspecified, and what kind of measurements one would "expect" if constructing similar setup, it would be intuitively easier if \$m\$ refers to median, and not the mean.  If you think about it, for that purpose, using mean for \$m\$ makes it quite difficult to wrap ones mind around what it means to the entire distribution of measurements.

Furthermore, if \$m\$ is mean, then all comparisons to the result someone else got from duplicating the measurements, must be based on the mean also.  (Having distributions/histograms, and knowing the median, is then useless information and not worth gathering.)  Which is okay, if that is the way the equipment works, or how it has to be done in practice (by convention, or for some other reason).

Mean differs from median only when the distribution is asymmetric (wrt. median).  For all symmetric distributions, mean = median.  The difference between mean and median depends on the asymmetry and width (deviation) of the distribution.  Thus, if you repeat an experiment (getting a set of results you wish to compare), and your errors have a different distribution than the one you are comparing to, the two mean will differ from the median by different amounts.  Then, the error bars will also include different portions of the two distributions, neither centered on the middle of the distribution!  Essentially, you are comparing apples to oranges.  And without knowing the median on both, we cannot even tell we're comparing apples to oranges, because we don't know how much each mean differs from the median.  Essentially, the entire distribution becomes un-useful information, and we can only compare the two means.

Which, again, is acceptable if metrologists always work with mean/average, and never even look at the actual measurement distributions/histograms and consider those non-useful.  If everything you can compare to is mean/average-based, then having a histogram or distribution –– or thinking in such terms! –– is useless and extra unneeded work for nothing.
« Last Edit: January 30, 2025, 05:10:07 am by Nominal Animal »
 
The following users thanked this post: gitm

Offline gitm

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #64 on: January 30, 2025, 03:01:44 pm »
Your point is well taken.
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 7333
  • Country: fi
    • My home page and email address
Re: Different type LCR SMD Fixture
« Reply #65 on: January 31, 2025, 07:03:20 am »
Of course, when using \$p = 100 \%\$ as you have, my "only part of the distribution" argument does not hold.  Then, if mean is what is typically used, then it makes more sense to use that, to keep results comparable.  Median does have the benefit of describing the "center" of the distribution, unlike mean, so median-based results from completely different distributions are still directly comparable.

Don't let my excess verbosity look like certainty!  It's just an affliction I'm working on.
 

Offline gitm

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #66 on: January 31, 2025, 01:39:10 pm »
Yes, but it did help understand the distinction between how you were using “useful” and how I used it.  It’s good to have these discussions otherwise I wouldn’t have thought twice about mean vs. median. 

For my purposes, I believe I have an understanding of how the fixture helps with repeatable measurements.  Did I mention I like it very much?  ;D
 
The following users thanked this post: Nominal Animal

Online mawyattTopic starter

  • Super Contributor
  • ***
  • Posts: 4196
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #67 on: January 31, 2025, 04:17:41 pm »
We found the Vishay WSL2512 Zero Ohm (Mouser 71-WSL251200000ZEA9) as a good Zero Ohm Reference with our Split-Kelvin Fixture. These have shown as good as the solid Copper thin plate cut to a 2512 size that we normally used in the past.

The WSL routinely produces <10uΩ (SD ~6uΩ) residual with the Fixture on our DMM6500 without any meter compensation.

What Zero Ohm Reference are you using, and what uncorrected residual do you see?

Best
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 

Offline gitm

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #68 on: January 31, 2025, 09:30:30 pm »
Neat part!  Didn’t know it came in “0”.  :)

My “zero ohm reference” is solid 16 AWG copper wire, flattened in a press, trimmed to 2512 size, and then finished on both sides with 2000 grit diamond stones at 0.0227” thickness.  This was to match the 2.4 ohm 2512 resistor (0.0238”).  Final dimension is 0.250”x0.120”x0.0227”.  Picture attached with it next to 2.4 ohm resistor.

As for uncorrected measurements (CKT515A again): N=20, median=7.7µΩ (I’ve seen the light), mean=8.2µΩ, ES (span)=3.3µΩ, sigma=1.1137µΩ, min=6.9µΩ, max=10.2µΩ.

Screenshot of actual measurements attached.  Order of measurements is left-to-right, top-to-Bottom.  Each measurement was taken by removing the 0 ohm reference and re-placing in a slightly different position.
« Last Edit: January 31, 2025, 09:45:07 pm by gitm »
 

Online mawyattTopic starter

  • Super Contributor
  • ***
  • Posts: 4196
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #69 on: February 15, 2025, 09:32:31 pm »
Here's our latest rendition of the Split-Klevin Adapter tailored for 4-Wire Bench DMM use. Initial test results are good, altho we are considering direct wiring to the Banana Jack rather than to a Spade Lug.

Best
« Last Edit: February 15, 2025, 09:35:13 pm by mawyatt »
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 
The following users thanked this post: RoGeorge

Online mawyattTopic starter

  • Super Contributor
  • ***
  • Posts: 4196
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #70 on: February 16, 2025, 01:36:56 am »
After thinking about using the spade lugs to attach to the Banana Jacks we decided to consider soldering wires directly to the Jacks rather than use the lugs. Reasoning was that the lugs only made electrical contact to the Jacks on one side since the Jack "Bolt" snugs up against the PLA plastic of the 3D Printed case, whereas normally the lugs make contact on both sides. The cheap Banana Jacks accept solder (claim Gold plated ??) and this seemed an easy improvement since it should also reduce the electrical interfaces of wire to lug to Bolt to Banana Jack to simply wire to Bolt to Banana Jack.

Anyway, here's what it looks like, note the Green Wire Ground with Black Banana Plug plugged into LF. Too early to tell if any of this is worthwhile since the best solution for low Z components is high test currents like the specialized Lab Grade Bench type Milliohm Meters. Higher is better because the sense voltage is higher with better SNR and less uncertainty.

Edit: After a few measurements last night and this morning of a Precision 1.00mΩ 2512 SMD Resistor we see repeatability within ~4uΩ on the DMM6500 on 1Ω Scale. We also see good repeatability on our 3 KS34465As on 100Ω Scale within ~10uΩ, and the 3 agree within ~50uΩ with each other and ~55uΩ with the DMM6500. If the older KS (Tan Case) is removed, then the 2 other KS (Black Case) are within ~15uΩ readings, and within ~25uΩ of the DMM6500.  Not scientifically rigid measurements but good enough to show how this SMD fixture helps with low Z DMM measurement repeatably, even with DMMs not known for good low Z measurements.   

Best
« Last Edit: February 16, 2025, 11:35:19 pm by mawyatt »
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 

Online Kean

  • Supporter
  • ****
  • Posts: 2565
  • Country: au
  • Embedded systems & IT consultant
    • Kean Electronics
Re: Different type LCR SMD Fixture
« Reply #71 on: February 17, 2025, 05:05:14 am »
Nice work.

I was going to ask for a link to those banana posts/plugs, but I grabbed a snippet of your image of just the "gold plated" parts, and pasted it into AliExpress image search to get some useful results.

It is actually amazing the accuracy and speed of the search results, both listing near identical items and some very similar brass refrigeration/AC fittings.  It is not like they can index things like with a word search.
 

Online mawyattTopic starter

  • Super Contributor
  • ***
  • Posts: 4196
  • Country: us
Re: Different type LCR SMD Fixture
« Reply #72 on: February 17, 2025, 01:48:11 pm »
That's interesting about the AliX Image search, did not know about that...thanks!!

WRT to the Banana Plugs, only the blades are magnetic. Not sure what metals or plating are used on the plugs but they are solderable. The threaded parts seems to be cast rather than machined threads, sort of sloppy, but intended to engage stranded wires. Also you can engage another 4mm Banana plug into the threaded "Bolt" open end. Not Metrology grade, but good enough for our needs :-+

If you decide to construct one of the Split-Kelvin Type Fixtures, we're sure you'll be impressed with the results.

Best
Curiosity killed the cat, also depleted my wallet!
~Wyatt Labs by Mike~
 
The following users thanked this post: Kean, Betty-testleadsworld


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf