Cheap SMD LCR Fixture, the Good, Bad and Ugly

[mawyatt]

That look good, just make sure the cover makes good contact with the main case. Also make sure the main case is making good contact with the copper surfaces.

Actually you don't need to go to this extreme, just making sure everything has a good electrical ground connection is sufficient, this effectively creates a somewhat like Faraday Shield around everything inside the case. You want all the parasitic capacitances to remain fixed, and a good mechanical structure keeps things from moving around.

You folks with these SMD fixtures should consider the DC Bias Adapter, kinda interesting to "see" how the DC Bias capacitance changes with ceramic caps, diodes and such!!

Good work and keep us up to date on your progress

Best,

[mawyatt]

Also should mention, when doing the "Open Cal", use something to space the fixture plungers the same distant apart as they will experience when engaged with the DUT. We just use something that keeps the Lever Arm at the proper spacing, a pencil works!!

I was just about to write this myself.  But a pencil??? 

Yeah, works well. It just sits behind the Lever arm between the arm and the low side brass plated (silver) plunger housing. Anything will work as a spacer, since it's not electrically involved.

Best,

[Kean]

Yeah, works well. It just sits behind the Lever arm between the arm and the low side brass plated (silver) plunger housing. Anything will work as a spacer, since it's not electrically involved.

OK, I was imagining the graphite being placed between the contacts... 
I don't know if your original description made it clear otherwise.

On the topic of your DC bias adapter, I've got some of your PCBs on the bench (5pcs added onto a JLC order) but I've not yet reviewed the BOM or ordered the parts to make one up.

[mawyatt]

Yeah, works well. It just sits behind the Lever arm between the arm and the low side brass plated (silver) plunger housing. Anything will work as a spacer, since it's not electrically involved.

OK, I was imagining the graphite being placed between the contacts... 
I don't know if your original description made it clear otherwise.

On the topic of your DC bias adapter, I've got some of your PCBs on the bench (5pcs added onto a JLC order) but I've not yet reviewed the BOM or ordered the parts to make one up.

The DC Bias PCBs were designed to fit inside this type 65x110x50 enclosure (sawed in half). Use the latest files for the PCB, they have the BNC terminals correctly spaced and added the fast charge capability (and a few other minor changes) mentioned earlier. We haven't used the latest PCB yet, waiting on a PCB order to piggyback on, which should be soon!!

Anyway, hopefully this DC Bias Adapter works as well for you as it has for us!!

Best,

https://www.aliexpress.us/item/3256801465819165.html?spm=a2g0o.order_list.order_list_main.638.27e31802vJOM1n&gatewayAdapt=glo2usa4itemAdapt

[hpw]

Good work and keep us up to date on your progress

So on second tweak as  while the connections wires are too thick, also shows the common 2mm diameter connection connector as one was loosely too.

Finally, last picture, RG178 inner silver wire as more flexible and in addition with torx screws for better alignment. Enlarged also inner holes to 4mm diameter as better to adjust.

May at last, shorten or use different connection cables 

BTW: The Hantek uses for cap measurements only 2 wires and each time switched on, the open calibration is a requirement.

Hp

[KungFuJosh]

It seems like those screws are always loose. I tightened them by unscrewing the parts so they could be turned without damaging the wires.

[mawyatt]

Yes, one of ours was also loose and causing erratic readings. These look different than what we have, ours are just long thin brass bolts without the insulator. These look better with the insulator!!

BTW this image illustrates why getting repeatable low Z measurements is difficult with these type fixtures. The effective Force and Sense point is the top of the brass screw insert where the two wires are soldered, this leaves the entire electrical path from the solder connections to the plunger tip not sensed.

We developed another inexpensive DIY type SMD LCR fixture for this reason where the Force and Sense electrical connection is made by the DUT SMD terminal, this has shown promise with repeatable low milli-ohm SMD DUTs readings.

Best,

[hpw]

It seems like those screws are always loose. I tightened them by unscrewing the parts so they could be turned without damaging the wires.

The loose part(s) are the cause of the not flexible and more harder cable (even after longer time as SMD size movements). In addition caused by the "Heat shrink tubing" as cable would not brake at the solder points.

The other question about the used Chinese Gold as may thin as it gets??

That's why I used the more flexible inner RG178 coax cable.

Hp

[gitm]

@mawyatt

I would be interested in purchasing a PCB from you, especially with the gold contacts, if you move forward with the design.  I'm in the US.  Please let me know if you make them available.

Thanks,
gitm

Anyway, just an update, hope some folks find this interesting as we've found getting good repeatable low Z measurements with SMD is not trivial.

Edit: These are DCR measurements. Also no disrespect for the standard SMD fixture, these aren't intended for low Z measurements. If one studies the design the Kelvin type connections are terminated on a long thin brass bolt and screws into the bottom of the brass plated plungers, so everything after the thin brass bolt head is not part of the Kelvin measurement. This includes the plungers, and the plunger to DUT contact surfaces, which can only be compensated for by the Short Cal. Whereas the Lever Arm structure with the split contact area, the DUT contact "ends" make the Kelvin termination, and thus should improve results for low Z components, which apparently works is some cases as we've demonstrated.

BTW, if some folks want to participate we might roll the next PCB with gold contacts, which should improve contact area performance.

Best,

                     

[mawyatt]

@mawyatt

I would be interested in purchasing a PCB from you, especially with the gold contacts, if you move forward with the design.  I'm in the US.  Please let me know if you make them available.

Thanks,
gitm

Anyway, just an update, hope some folks find this interesting as we've found getting good repeatable low Z measurements with SMD is not trivial.

Edit: These are DCR measurements. Also no disrespect for the standard SMD fixture, these aren't intended for low Z measurements. If one studies the design the Kelvin type connections are terminated on a long thin brass bolt and screws into the bottom of the brass plated plungers, so everything after the thin brass bolt head is not part of the Kelvin measurement. This includes the plungers, and the plunger to DUT contact surfaces, which can only be compensated for by the Short Cal. Whereas the Lever Arm structure with the split contact area, the DUT contact "ends" make the Kelvin termination, and thus should improve results for low Z components, which apparently works is some cases as we've demonstrated.

BTW, if some folks want to participate we might roll the next PCB with gold contacts, which should improve contact area performance.

Best,

                     

More details over here for those interested.

https://www.eevblog.com/forum/projects/different-type-lcr-smd-fixture/msg5061028/#msg5061028

If we can get a few folks in US interested (shipping overseas is too expensive and better for those folks to just direct order PCBs), we'll consider ordering a small batch of the PCBs.

Best,

[TERRA Operative]

I wouldn't mind a PCB or two but I'm in Japan. Have you released design files or Gerbers etc?

[mawyatt]

Here's some additional work we've done with these cheap SMD LCR Fixtures.

The DC Bias Adapter is a good addition and works well with all our Bench Type Lab Grade LCR Meter fixtures, however it is somewhat bulky, especially for use with SMD Fixtures.

Since we often work with SMD components, we decided to modify one of our 2 SMD Fixtures to allow direct External DC DUT Bias, similar to what the DC Bias Adapter does. To make this more useful want all the components inside the Fixture case.

Since this Bias function doesn't have to work with all the various LCR Attachments like Kelvin Clips, Tweezers, Leaded Component Fixture, and so on, and we aren't interested in working with very large value capacitors for example (this is a SMD fixture after all!!), keeping things simple and small was a priority.

The SMD fixtures are structured so the Force and Sense signals are terminated at the plunger midpoint, thus there is always a Force and Sense connection, whether a DUT is installed or not. This simplifies the DC bias setup and LCR Meter protection scheme.

We used a 10uF Mylar Film (250V) for the series Hcur decoupling and a 1uF Mylar Film (400V) for the Hpot decoupling. Meter protection is utilized on the Hcur and Hpot with back-to-back Zeners (24V) for Hcur and back-to-back diode isolated Zeners for Hpot. The diode isolation of the Zener reduces the apparent shunt capacitance as "seen" by signal and why utilized for Hpot Sense terminal. Hcur isn't affected by the Zener shunt capacitance, and so just back-to-back used.

One of our Lab Grade LCR Meters the TH2830 is rated for a Maximum Input Voltage at any terminal of +-42 DC Volts, so using 24V Zeners provides some safety margin. When doing a Short Cal you must bypass the DC obviously as the DC coupling capacitors provide DC isolation!!

After some quick tests we can verify this setup works very nicely with the TH2830, however the IM3536 is having an issue with Calibration which we'll investigate when time permits.
Edit: Problem solved, see #64 below.

Anyway, hope this is of interest to some folks.

Best,

[mawyatt]

Few more images.

Best,

[mawyatt]

After some quick tests we can verify this setup works very nicely with the TH2830, however the IM3536 is having an issue with Calibration which we'll investigate when time permits. 

After some thought and more testing figured out the issue with the IM3536. Because it operates differently than the TH2830, the IM3536 requires a low impedance DC coupling on the Hpot terminal as well as the Hcur terminal. Using a 10uF for the Hcur works fine, however the 1uF on the Hpot terminal has too much impedance difference wrt to the Hcur and flags an error failure during Calibration.

Solution was to replace the 1uF with a 10uF on the Hpot terminal and problem solved

BTW one of the issues with these cheap SMD fixtures with the "U" shaped bottom cover attached at the ends with just two bolts, the cover is too flexible and isn't mechanically secure as the ones with the cover attached on the end sides with 4 bolts. Why this matters is when measuring small capacitors (High Z) any physical change in the configuration and/or fixture (flex of the cover) can change the measurement. If the case flexes, then the interior fields change as the case/cover are the effective ground shield for the fixture interior and effect the measurement. These interior field effects are removed during Calibration but must remain stable during measurements.

Best,

[dazz1]

After some quick tests we can verify this setup works very nicely with the TH2830, however the IM3536 is having an issue with Calibration which we'll investigate when time permits. 

After some thought and more testing figured out the issue with the IM3536. Because it operates differently than the TH2830, the IM3536 requires a low impedance DC coupling on the Hpot terminal as well as the Hcur terminal. Using a 10uF for the Hcur works fine, however the 1uF on the Hpot terminal has too much impedance difference wrt to the Hcur and flags an error failure during Calibration.

Solution was to replace the 1uF with a 10uF on the Hpot terminal and problem solved

BTW one of the issues with these cheap SMD fixtures with the "U" shaped bottom cover attached at the ends with just two bolts, the cover is too flexible and isn't mechanically secure as the ones with the cover attached on the end sides with 4 bolts. Why this matters is when measuring small capacitors (High Z) any physical change in the configuration and/or fixture (flex of the cover) can change the measurement. If the case flexes, then the interior fields change as the case/cover are the effective ground shield for the fixture interior and effect the measurement. These interior field effects are removed during Calibration but must remain stable during measurements.

Best,

I arrived at this topic because I was looking for a SMD fixture for my super cheap Chinese component tester.  Having read the thread it looks like the target resolution is  zepto Farads with 4-wire sensors. 
Rather than attempting to use split probes, which I suspect will be unreliable in practice, I propose coaxial probes.   The sense wire would be old wire-wrap.  The good-ole stuff is silver plated, single strand.  It has a tough thin insulation.  I think a better option than enamel wire.

The drawings show a probe tip drilled to allow wire wrap to be placed inside.  The tip of the wire wrap would be soldered to just behind the face of the probe tip.  This would place the sense wire very close to the smd component with very little opportunity for error.   This would allow the cheap-as SMD fixture to be used by simply replacing the probes with the coaxial versions shown in the drawings. 

I could make the coaxial probes but this modification would be wasted on me.  I don't have the cheap-as smd fixture and I am waiting for the next firmware release for the Chinese component tester to measure down to yocto Farads.  In the mean time it would be interesting to see someone make/try/test coaxial probes.

[mawyatt]

That's an interesting idea, as it does move the connection point for the Force and Sense just inside the probe tip. This would remove most the impedance effects of the plunger, altho the resistance of plunger is very low as it's a solid chunk of plated brass. 

However, it doesn't help with the SMD DUT contact effects from the plunger tip to the DUT SMD end terminals. We've found this effect (DUT Terminals contact point) to be the most problematic in low Z SMD measurements and why we developed the Split-Kelvin SMD Fixture shown here:

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

Here the Force and Sense connections are made by the DUT end terminals, on the DUT side of the connection. This Split technique works very well down to milliohm Z levels with excellent repeatability as we've done extensive testing with thousands of measurements, also holds up for higher frequency measurements as we've just built another using shielded coax from the BNC connectors to the PCB connection as shown in the attached image.

We don't have the capability to machine a plunger like you've shown, would be interesting to see how well it works 

Best,

[KungFuJosh]

Is the clamp the Adafruit 2459?

Thanks,
Josh

[mawyatt]

Got it off AliEx, there are variations. Get the one like shown in the image, note how the plunger is attached.

Best,

[KungFuJosh]

I just measured your PCB, it looks like the $4 Adafruit 2459 should fit well.

[tautech]

After some quick tests we can verify this setup works very nicely with the TH2830, however the IM3536 is having an issue with Calibration which we'll investigate when time permits. 

After some thought and more testing figured out the issue with the IM3536. Because it operates differently than the TH2830, the IM3536 requires a low impedance DC coupling on the Hpot terminal as well as the Hcur terminal. Using a 10uF for the Hcur works fine, however the 1uF on the Hpot terminal has too much impedance difference wrt to the Hcur and flags an error failure during Calibration.

Solution was to replace the 1uF with a 10uF on the Hpot terminal and problem solved

BTW one of the issues with these cheap SMD fixtures with the "U" shaped bottom cover attached at the ends with just two bolts, the cover is too flexible and isn't mechanically secure as the ones with the cover attached on the end sides with 4 bolts. Why this matters is when measuring small capacitors (High Z) any physical change in the configuration and/or fixture (flex of the cover) can change the measurement. If the case flexes, then the interior fields change as the case/cover are the effective ground shield for the fixture interior and effect the measurement. These interior field effects are removed during Calibration but must remain stable during measurements.

Best,

I arrived at this topic because I was looking for a SMD fixture for my super cheap Chinese component tester.  Having read the thread it looks like the target resolution is  zepto Farads with 4-wire sensors. 
Rather than attempting to use split probes, which I suspect will be unreliable in practice, I propose coaxial probes.   The sense wire would be old wire-wrap.  The good-ole stuff is silver plated, single strand.  It has a tough thin insulation.  I think a better option than enamel wire.

The drawings show a probe tip drilled to allow wire wrap to be placed inside.  The tip of the wire wrap would be soldered to just behind the face of the probe tip.  This would place the sense wire very close to the smd component with very little opportunity for error.   This would allow the cheap-as SMD fixture to be used by simply replacing the probes with the coaxial versions shown in the drawings. 

I could make the coaxial probes but this modification would be wasted on me.  I don't have the cheap-as smd fixture and I am waiting for the next firmware release for the Chinese component tester to measure down to yocto Farads.  In the mean time it would be interesting to see someone make/try/test coaxial probes.

Let's see what you think of ST42 tweezers when I visit next weekend.
 

[mawyatt]

I just measured your PCB, it looks like the $4 Adafruit 2459 should fit well.

We've got a few of those shown in your image, they were Ok but the one like we've shown performed better. Note how the plunger attaches to the Lever Arm frame. Maybe get few of each and try them out, they are really cheap.

You need to mount the PCB on a BNC hosting fixture like shown. This was harvested from a cheap LCR Meter Kelvin Clip set also from AliEx. You want a rigid mechanical structure for the PCB to mount too. The PCB you have is an earlier version, doesn't have the "Guard" under the active DUT area, however that doesn't seem to matter much with lower Z measurements.

Think we also sent you some of the very early version PCBs (can't remember) which have the 4 BNCs directly mounted to the PCB, these worked well but we wanted a more rigid mechanical structure and prefer the latter version as shown.

Get yourself some low Z High Precision SMDs for checking things out, 2512 Resistors are a good start, and relatively easy to handle being large.

Best,

[KungFuJosh]

We've got a few of those shown in your image, they were Ok but the one like we've shown performed better. Note how the plunger attaches to the Lever Arm frame. Maybe get few of each and try them out, they are really cheap.

Do you have any links to stuff you liked?

Get yourself some low Z High Precision SMDs for checking things out, 2512 Resistors are a good start, and relatively easy to handle being large.

I love 2512 resistors. That's what I usually use on my SMT guitar amp PCBs. Much less likely to go flying when I check them with a tweezers, and much easier to find when they do. 😉

[mawyatt]

Do you have any links to stuff you liked?

Cheap Kelvin Clips for LCR Meter, toss everything except the case and BNCs, which we used.

https://www.aliexpress.us/item/3256804400699838.html?spm=a2g0o.order_list.order_list_main.90.503618020zePkv&gatewayAdapt=glo2usa

Preferred type Lever Arm & Plunger.

https://www.aliexpress.us/item/3256803387853373.html?spm=a2g0o.order_list.order_list_main.149.503618020zePkv&gatewayAdapt=glo2usa

Different type.

https://www.aliexpress.us/item/3256801104879812.html?spm=a2g0o.order_list.order_list_main.255.503618020zePkv&gatewayAdapt=glo2usa

You need to get a few of the plungers and go through them to find the better ones, being so cheap they are highly variable!! Try out both types, they both work, we just like the version shown slightly better.

Be sure and use an internal shield with the BNC case, especially if you don't use shielded wires to the PCB (like we did originally). We made one up with the shielded wires for higher frequency use, however the unshielded wires worked fine, OCD

Note how you can flip over the plunger to use the standard supplied tip (black) for larger SDM DUTs.

Best,

[KungFuJosh]

Thanks!

Your second link is the same as the first though.

[mawyatt]

Fixed now!!

Best,