Hioki IM3570 LCR & Impedance Analyzer

[KungFuJosh]

What is the point of going from 4 BNCs on the instrument, through a box and cable to another box with 4 SMA?, and then out to the probes?  Why not directly connect a set of probes to the instrument BNC?  What is the point of the extra boxes and cables in the signal path here?  I would understand if the probes and BNC box was one unit, but separate probes with SMAs?  The frequency only goes to 5MHz.  BNC is more than fine for that.

I would guess that since they consider the kelvin clips themselves a replaceable "wear & tear" item, they kept them separate. My meter only goes to 5MHz, but the probes are rated to 8MHz.

[mawyatt]

@ Smokey,

Think Hioki uses the double box with cable for other fixtures as well, probably for much higher frequency use. Don't believe those are SMA connectors, they look bigger and being a screw type rather than bayonet should have better signal fidelity than BNC. Note how the cables are attached in the box, without any exposed center conductors, and they look expensive which shouldn't surprise anyone!!

Agree does seem odd tho.

Best,

[mawyatt]

@ Josh, can you identify the type connectors used for the cables with Kelvin Clips?

Best,

[KungFuJosh]

@ Josh, can you identify the type connectors used for the cables with Kelvin Clips?

SMA is correct. I'm not surprised. It makes sense to use something standard, and just pick something high quality.

[KungFuJosh]

Mike, help!

I built your cool SMT fixture, but I don't have a 3D printer, and the rubber clamp tip doesn't work at all for 805 parts.

What do you suggest I use instead to secure the SMT parts to the fixture?

Thanks,
Josh

[tautech]

Mike, help!

I built your cool SMT fixture, but I don't have a 3D printer, and the rubber clamp tip doesn't work at all for 805 parts.

What do you suggest I use instead to secure the SMT parts to the fixture?

Thanks,
Josh

Got a matching bolt, thread tap and a drill ?
Turn a piece of nylon shaft to a blunt point, drill and tap it and fit it to a suitable bolt/threaded shaft.....or some combination of this.

[KungFuJosh]

Got a matching bolt, thread tap and a drill ?
Turn a piece of nylon shaft to a blunt point, drill and tap it and fit it to a suitable bolt/threaded shaft.....or some combination of this.

I was thinking of getting a nylon screw and using that instead, but figured I'd ask Mike first since he uses this fixture setup.

The way the clamp drops down is a little awkward too, so I'm not sure the nylon bolt idea will work great either.

[tautech]

Got a matching bolt, thread tap and a drill ?
Turn a piece of nylon shaft to a blunt point, drill and tap it and fit it to a suitable bolt/threaded shaft.....or some combination of this.

I was thinking of getting a nylon screw and using that instead, but figured I'd ask Mike first since he uses this fixture setup.

The way the clamp drops down is a little awkward too, so I'm not sure the nylon bolt idea will work great either.

Hmmm, or machine the existing rubber foot to something more conducive to working with tiny parts......

[Martin72]

It looks like a nutcracker.
And that should have some advantage over the “normal” SMD fixtures?

[KungFuJosh]

Hmmm, or machine the existing rubber foot to something more conducive to working with tiny parts......

The biggest problem with the existing rubber foot is that it's too soft. It doesn't firmly press well, and tends to move the DUT off the traces. Then the DUT is stuck to the foot and needs to be pulled off.

The traces need precision placement of the DUT, but the rubber foot has moved it in every attempted use. I can't do short correction because the 0Ω resistor fails to short the traces when the foot is applied. At best, I've had 2 of 4 traces shorted. Even using a piece of copper gets pushed out of the way.

[KungFuJosh]

It looks like a nutcracker.
And that should have some advantage over the “normal” SMD fixtures?

The traces are separated so it's 4-wires all the way to the DUT.

[mawyatt]

Mike, help!

I built your cool SMT fixture, but I don't have a 3D printer, and the rubber clamp tip doesn't work at all for 805 parts.

What do you suggest I use instead to secure the SMT parts to the fixture?

Thanks,
Josh

Keep the large rubber tip, you'll need it for larger SMD components like power inductors or large chip caps. On the other bolt end add a tapered plastic tip, just about anything will work. You could get a plastic or wooden rod/dowel and turn down the tip, then drill and tap the other side for the bolt. However you do this, don't engage the bolt too deep into the tip as it distorts the fields more than the plastic tip.

Edit: The nylon screw/bolt is a good idea as you can easily taper the end to a dull point. Also did you add a internal shield within the box? This is worthwhile if you didn't.

Experiment with the Plunger base height above the PCB, we found ~10mm works with the Tip & Plunger fixture we have (there are a few types), and most of the SMD stuff we work with. You can fine tune the height with the bolt adjustment for different similar physical size SMD components.

When you use the fixture don't use too much force, often just a light pressure from the tip is enough. Get yourself some precision (0.5% if possible) 1, 5, 10, 15, 20, 30, 40, 50 and 100 milli-Ω 2512 resistors, start off with these larger SMD components and study the profile of the SMD components, some of the SMD resistors have the contact area recessed and often they make better contact flipped over face down.

We've found the larger more expensive quality SMD components seem to be easier to work with, altho we've tested some 0603 and 0805 components with patience.

Make sure the SMD and Fixture surface area is clean, we often clean with IPA and a cotton swab. Play around with how to engage the contacts (Hc, Hp, Lp, Lc) on the two SDM end conductive surfaces. Sometimes you'll get ridiculous readings, either way too high, low or negative, this is due to not having all 4 contacts engaged.

Don't forget to make a good Zero Ohm Shunt, critical for quality low Z measurements.

Best,

[KungFuJosh]

The nylon screw/bolt is a good idea as you can easily taper the end to a dull point. Also did you add a internal shield within the box? This is worthwhile if you didn't.

I think I will try the nylon screw first, as it's the simplest answer. I have a laser cutter, which I used to cut a 6mm thick piece of acrylic for the spacer.

I ground the paint off at contact points so the whole box is a better shield (and has continuity to the PCB ground plane), but I'm guessing you're asking if I added a shield plate between the H and L sides? I didn't do that yet.

[mawyatt]

It looks like a nutcracker.
And that should have some advantage over the “normal” SMD fixtures?

The traces are separated so it's 4-wires all the way to the DUT.

Correct, the actual SMD component conductive ends make the Hcur to Hpot connections and Lcur to Lpot connections.

This reduces contact measurement variations which are pronounced in low Z DUTs with the normal SMD fixture. If you look at the simple model we made for the normal SMD fixture, the Hcur and Hpot connections are made at the thin brass bolt end which screws into the plunger base, same for Lcur and Lpot on the other plunger.

So from this point to the plunger tip is un-sensed on each plunger by Hpot and Lpot respectively. The plunger tip to DUT  impedance variations are also un-sensed on each DUT end which produces a significant measurement uncertainty for low Z components that can't be compensated by the Zero/Short Calibration. A little experimentation with known precision low Z components like low mill-Ω 2512 precision resistors for example will show this uncertainly, at least it certainly did with us, which is why we developed this Split-Kelvin SMD Fixture.

The two resistors on the PCB are for providing a known resistance between Hcur and Hpot and same for Lcur and Lpot. This allows easier Open Cal but has some effects on such. Think we discussed this elsewhere, but can illustrate again if needed.


Edit: Just noted that you don't have a ground wire from Plunger Fixture to Case. You want Plunger Arm at Ground potential for better HF measurements. Also wrt to internal case shield, just a thin Cu sheet or FR4 between the H and L sides works, if you don't have any Cu sheet or FR4 then cut a section out of a Tin Can and solder to a BNC terminal or terminal wire running along the BNCs.

BTW Fixture looks good, nice work Josh

Best,

[mawyatt]

Hmmm, or machine the existing rubber foot to something more conducive to working with tiny parts......

The biggest problem with the existing rubber foot is that it's too soft. It doesn't firmly press well, and tends to move the DUT off the traces. Then the DUT is stuck to the foot and needs to be pulled off.

The traces need precision placement of the DUT, but the rubber foot has moved it in every attempted use. I can't do short correction because the 0Ω resistor fails to short the traces when the foot is applied. At best, I've had 2 of 4 traces shorted. Even using a piece of copper gets pushed out of the way.

Try raising the Plunger fixture a little higher so the plunger tip engages the SMD at ~90 degrees with the Plunger "Arm" ~parallel to the PCB surface.

It's difficult to see the DUT slide with the larger rubber tip, so a thin tip should help. Adjust the heights until the DUT doesn't slide and is pressured almost vertically. As mentioned in our case ~10mm was about right. A small amount of sliding isn't bad as it "scrubs" the surfaces which helps with contact.

Also, start off with larger SMDs like 2512 resistors to get familiar with the way this fixture works best.

Best,

[KungFuJosh]

If the plunger tip is short, what's the difference with the base height matter? Doesn't one length/height make up for the other? I would think 4mm higher base, or 4mm shorter plunger arm should have the same effect?

[mawyatt]

There's limits in how the bolt can be used for adjustment, especially so if you use both ends of the bolts as we do. Usually you want the plunger tip to engage the SMD DUT almost vertical and still cover most of the expected components. Having the plunger base too short limits the DUT SMD height, having it too high limits thin SMDs.

We found ~10mm standoff for the Plunger base has worked well for us. We played around with spacers and washers, then 3D printed the standoff plate.

BTW we found you don't want the Plunger Fixture to be extremely rigid, some slight flexing is desirable as in "springiness".

Play around with various adjustments and configurations until you find something to your liking

Best,

[Martin72]

I would like the Gerber file of the board...

[mawyatt]

I would like the Gerber file of the board...

We've posted details on the thread dedicated to this Unique LCR SMD Fixture. This includes Gerber files for the latest and previous versions of the PCB, and 3D print files for the Plunger Tip and Support Top and Bottom.

Let's try and keep all the discussions related to such on this over here.

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

BTW we have plans to publish the details of this fixture as well.

Best,

[KungFuJosh]

I posted in the other thread, but I'll post it here also since it's relevant to my IM3570. The nylon screw was a success. No modifications, and the screw is flexible enough I can safely position it on 805 SMT stuff with no trouble.

Thanks,
Josh

[KungFuJosh]

I was curious about how the resistance ranges functioned differently between the IM3570, and the 3532-50 so I popped it open and took a photo. The design of the IM3570 made it so I couldn't answer my question without a lot of effort, so I only took the one photo and closed it back up.

It's clearly a nice design. It looks like the top PCB doesn't need to be unmounted, and that there are separate screws for the top plate to be removed with the PCB still attached. But I didn't have the time or ambition to go further now.

The Xilinx Spartan chip says XC3SD1800A. Some of the others are easier to read, some are not. 😉

Thanks,
Josh

[zrq]

Another question that may worth real money is there significant hardware similarity with the lower end Hioki IM353x that allows an upgrade.

[KungFuJosh]

Another question that may worth real money is there significant hardware similarity with the lower end Hioki IM353x that allows an upgrade.

It's a possibility, but I slightly doubt Hioki is as lazy as Tonghui. 😉

We'd need some teardown photos of both models. I'll eventually take more/better photos of mine, but it will probably be a while before I get around to it.

[KungFuJosh]

I submitted the following suggestions to improve the IM3570:

• PNG format instead of BMP for screen captures (PNG is ideal both for image quality, and to be able to share with peers).
• Custom colors for screen and/or screen captures, or at least an option to invert the screen colors on capture.
• Axis markers for analyzer to make reading and sharing info easier (why is test voltage on the same line as the frequency axis?? A lot of screen space is wasted in this regard).
• Screen capture of any screen, instead of just measurement screen (to share settings, etc. Currently need to take photographs of the meter).
• A simple way to get screen captures remotely, either via software app or a built-in web server on the IM3570 (basic functional control, and/or only remote viewing and screen capturing would be fine, though full control would be great if possible). (The LCR Meter Sample Application is not sufficient, but not terrible either.)
• Screensaver option with options for: Off, 10 minutes, 20 minutes, 30 minutes, 60 minutes.

The service manager responded:

Thank you very much for sharing your valuable feedback and suggestions with us.
We truly appreciate the time and effort you have taken to provide us with your thoughts on how we can improve IM3570.
Your insights are extremely helpful, and we will carefully review your suggestions to see how we can implement them to enhance our offering including our HQ in Japan.
We are committed to continuously improving and your feedback plays a crucial role in that process
If you have any additional comments or further suggestions, please do not hesitate to reach out to us.

Once again, thank you for your continued support and trust in our company.

Hopefully, they'll do at least some of them, but the canned response doesn't give me a lot of hope lol. I guess we'll see.

[Martin72]

Screen capture of any screen, instead of just measurement screen

This seems somehow familiar....