East Tester ET3502 LCR-Bridge

[Martin72]

I had actually had a connection earlier...
The device is not displayed in the home network, but Nmap finds it and it also finds a port (4321).
I was able to get the ET35 to talk via Putty using *IDN?
Then I tried a number of other SCPI commands and e.g. with all of them concerning LAN, either the ET35 fails or Putty closes.
I will describe this in more detail later, but it should be noted that the ET3503 at least responds somehow.

[w.v.s.]

Looks good. SCPI over TCP seems to work generally. As far as I remember, serial communication with the ET4XXX was a little special, too. So I think you can now try to experiment with timing, CR/LF termination and such aspects. 

[Martin72]

As long as you avoid the communication commands as much as possible, you will usually get an answer.
However, as soon as you enter system:communicate:lan:self: and then what you want to know, either there is no response or putty quits.
I currently have a 100Ohm resistor connected, I don't yet know how to display its value directly.

[w.v.s.]

What happens with a simple

Code: [Select]

FETCH??

[Martin72]

Oh look, with "fetch?" you get the value.
It's nice that it's not in their own document.

[w.v.s.]

I think it is basically described in 2.4.10 with the defaults in [].

[Martin72]

I have tested a few more commands, most of them work well, the Ethernet interface works albeit very rudimentary - Somehow East Tester reminds me of Rigol...
Good hardware, good ideas, but the software then....
For now, it's time to focus on the actual features of the ET35.
There are many things to adjust, cable length is just one of them.
You can also make this component-specific and save/call it up, under these circumstances I believe that the device can definitely achieve the 0.05% basic accuracy.
But how to test, I don't have such components here...
I think I'll limit myself to values that are known and then compare them.

[mawyatt]

I have tested a few more commands, most of them work well, the Ethernet interface works albeit very rudimentary - Somehow East Tester reminds me of Rigol...
Good hardware, good ideas, but the software then....
For now, it's time to focus on the actual features of the ET35.
There are many things to adjust, cable length is just one of them.
You can also make this component-specific and save/call it up, under these circumstances I believe that the device can definitely achieve the 0.05% basic accuracy.
But how to test, I don't have such components here...
I think I'll limit myself to values that are known and then compare them.

Think you have a TH2830 available, this could be a good reference to use. Would consider forgoing the Kelvin Clips and use the Direct Junction Box types for the Leaded and SMD type components. Might consider something like discussed here:

https://www.eevblog.com/forum/testgear/sensitive-lcr-meter-measurements/

For reference components consider quality leaded Polystyrene, Polypropylene, Mica Capacitors, with some C0G SMD types. For DC or very low frequency using the Kelvin Clips for leaded precision resistors should be acceptable, and you can also use your precision DMM to verify results.

Inductors are another story tho, and we don't place much faith in any typical inductors we have where precision is considered. Precision reference inductors may be available tho, but likely expensive.

If you consider the actual LCR measurement taking place with an inductor, it's the same as for a capacitor. The DUT Complex Voltage and Current are measured, then the Complex Impedance is computed. The only difference is the Impedance Angle is + for Inductance and - for Capacitance, then the test frequency is taken into account to display the Inductance/Capacitance Value. So unless the LCR is misbehaving the impedance levels measured, computed and displayed should be similar in magnitude but opposite in phase for Inductors and Capacitors. Sometimes meters display negative Capacitance for Inductance and visa-versa as the component goes thru resonance, or the wrong parameter is selected.

Anyway, you should be able to get some good components for reference measurements at a reasonable $. If you decide to use the mentioned direct connected "boxes" for various DUT types, you should use the same "box" for the reference measurement on the TH2830 and then on your ET35, that should help reduce another "variable uncertainty" in your evaluations.

Best,

[Martin72]

Hi,

SMD Test Fixure Aliexpress

SMD Test Fixture Sourcetronic

I've looked at it three times, I don't see any difference....

[KungFuJosh]

Hi,

SMD Test Fixure Aliexpress

SMD Test Fixture Sourcetronic

I've looked at it three times, I don't see any difference....

The generic ones on Ali are okay with some extra work like tightening fasters internally, and maybe needing to shorten wires. They also often don't send you what's pictured, which is why I got my generic one for under $20. 😉

The DUT holders are also different and effect how things actually fit between the pins. Not easy to compare in photos, especially if one of the photos is a lie.

I'm not saying not to get one from Ali, just be aware you'd be lucky if you receive what you hope to.

[Martin72]

I will see in about 2 weeks.

[mawyatt]

Hi,

SMD Test Fixure Aliexpress

SMD Test Fixture Sourcetronic

I've looked at it three times, I don't see any difference....

Here's something that might help, follow the details on how to improve the fixture.

https://www.eevblog.com/forum/testgear/cheap-smd-lcr-fixture-the-good-bad-and-ugly%21/msg4539458/#msg4539458

We don't have a quality SMD fixture to compare with, but after some rework the cheap ones we have seem OK. Note on one, we had to 3D print another plastic SMD DUT holder as the supplied one was too shallow and allowed smaller SMD DUTs to slip under when the plunger fingers were engaged.

Note that with the ET type LCR Meters reversed Terminal order, the Hcur and Hpot will be connected to the movable plunger rather than the fixed plunger on other LCR Meters. Might become an issue since the fields are higher with the H terminals which will now move with the plunger on ET types, whereas the Lcur and Lpot are close to ground potential and thus the field strength is lower wrt ground potential and should produce less field movement induced errors with other LCR Meters.

Best,

[KungFuJosh]

https://www.eevblog.com/forum/testgear/cheap-smd-lcr-fixture-the-good-bad-and-ugly!/msg4539458/#msg4539458

URLs with an exclamation mark like that won't work. You need to replace the ! with %21:

https://www.eevblog.com/forum/testgear/cheap-smd-lcr-fixture-the-good-bad-and-ugly%21/msg4539458/#msg4539458

[Martin72]

Hi,

Note that with the ET type LCR Meters reversed Terminal order, the Hcur and Hpot will be connected to the movable plunger rather than the fixed plunger on other LCR Meters.

I consider this to be negligible with an AC voltage as the measuring voltage.
"Fun" fact:
The Kelvin clip terminal and cables supplied are reversed.
In other words, the red-marked wires are connected to L and vice versa.

[mawyatt]

Hi,

Note that with the ET type LCR Meters reversed Terminal order, the Hcur and Hpot will be connected to the movable plunger rather than the fixed plunger on other LCR Meters.

I consider this to be negligible with an AC voltage as the measuring voltage.
"Fun" fact:
The Kelvin clip terminal and cables supplied are reversed.
In other words, the red-marked wires are connected to L and vice versa.

Kelvin Clips won't matter, SMD fixtures will!! Even more so with AC at higher frequencies. The modded SMD fixtures we have consistently produce a few fF uncertainties, don't think that would be the case on reversed H and L Terminals.

At one point we were considering an ET type LCR Meter but then discovered the reversed order. Now won't consider them for that reason, even if we needed another Bench LCR Meter (we don't) and with a big discount.

Anyway, of course YMMV.

Best,

[Martin72]

We will find out whether that plays any role.

[Martin72]

Think you have a TH2830 available, this could be a good reference to use.

Yes, at work.
If you want to know "exactly", either the ST2830 has to go to me or my ET35 has to go to work, because of the same ambient temperature/humidity.
However, this should only be of an academic nature, although with 0.05% basic accuracy with an LCR measuring bridge, we are already close to the academic question.

For reference components consider quality leaded Polystyrene, Polypropylene, Mica Capacitors, with some C0G SMD types. For DC or very low frequency using the Kelvin Clips for leaded precision resistors should be acceptable, and you can also use your precision DMM to verify results.

That is clear, or I hope it is clear to everyone.

Inductors are another story tho, and we don't place much faith in any typical inductors we have where precision is considered. Precision reference inductors may be available tho, but likely expensive.

Inductors are the "bitches" among complex components.
If they have a core, things get quite adventurous.
The core causes them to saturate (well, you won't usually achieve this with an LCR bridge).
The core has its own inductance, Al, which plays a role.
A core has a large tolerance range, easily around 30%.
If Q is negligible, and this is the case with a pure measurement, the tolerance values can be considerably reduced by adjusting the air gap.
Nevertheless, I don't really "like" coil measurements.

If you consider the actual LCR measurement taking place with an inductor, it's the same as for a capacitor. The DUT Complex Voltage and Current are measured, then the Complex Impedance is computed. The only difference is the Impedance Angle is + for Inductance and - for Capacitance, then the test frequency is taken into account to display the Inductance/Capacitance Value. So unless the LCR is misbehaving the impedance levels measured, computed and displayed should be similar in magnitude but opposite in phase for Inductors and Capacitors. Sometimes meters display negative Capacitance for Inductance and visa-versa as the component goes thru resonance, or the wrong parameter is selected.

If you have ever done arithmetic with complex values, you should be familiar with this.
In principle, we can cover the majority of inductance measurements with a measuring bridge such as the ST2830/TH2830.
The situation is different when calibrating power chokes, as the measuring voltages and currents of a standard measuring bridge are not even close to being sufficient.
Or to put it the other way round, you are scratching at the "beginning" of an inductance.
For such cases, there is then the voltage-time area measurement, with real voltages and currents.

[mawyatt]

We will find out whether that plays any role.

Answer is, It does make a slight difference where the moving Plunger section is active with the H or L terminals, favoring the latter just as we suspected   

Curiosity got the best of us, so we used our SDM fixture with the TH2830 at 100KHz.

First we did an Open Cal with the spacing ~ 0805 spacing, then measured an "Open" at same spacing with average of 64 and saw a variation of ~3fF total.

Then we flipped over the SMD fixture so the H terminal is driving the SMD Movable Plunger, here we saw a Open variation of ~7fF. We returned to upright position and mounted a 0805 C0G 0.3pF capacitor and witnessed a ~6fF variation, then flipped over the SMD fixture upside down and had ~13fF variation, all with 64 averages at 100KHz.

Next first we did an Open Cal upside down,  and repeated the above with 0805 C0G 0.3pF results showing ~7fF variation upside down and ~5fF normal.

These tests were done as quick tests just to "see" if terminal orientation had any effect on High Z Small Capacitance measurements.

Very subtle effect but favoring the normal terminal orientation for our LCR Meters when using our modded SMD fixture as we suspected.

BTW agree that Inductors are more of a measurement issue than capacitors, since we all can get quality, stable, somewhat precision capacitors in reasonable values, for reasonable prices. Not so to the best of our knowledge with inductors.

We can send you a quality 10, 100, or 1000pF 0805 C0G chip cap from US to Germany and expect you'll get a result comparable to what we measure. Not sure what values would appropriate for the inductor tho, and what the results would be??

Anyway, you can do the same tests we did with your ET35 by flipping over the SMD fixture, or do so with your SH2830 at work.

Edit: Another simple test is to vary the plunger spacing between two fixed distances, use a spacer on the "Arm" so as not to disturb the fields around the active plunger and compare the results when flipped over and normal, we found the normal position has ~28% less variation in capacitance readings.

As always, YMMV.

Best,

[Martin72]

Hi,

We used our SDM fixture with the TH2830 at 100KHz.

First we did an Open Cal with the spacing ~ 0805 spacing, then measured an "Open" at same spacing with average of 64 and saw a variation of ~3fF total.

Then we flipped over the SMD fixture so the H terminal is driving the SMD Movable Plunger, here we saw a Open variation of ~7fF. We returned to upright position and mounted a 0805 C0G 0.3pF capacitor and witnessed a ~6fF variation, then flipped over the SMD fixture upside down and had ~13fF variation, all with 64 averages at 100KHz.

Next first we did an Open Cal upside down,  and repeated the above with 0805 C0G 0.3pF results showing ~7fF variation upside down and ~5fF normal.

Don't be angry with me, but I very much doubt that a TH2830 can still resolve values in the femto range accurately, I consider that to be "background noise".

[mawyatt]

Hi,

We used our SDM fixture with the TH2830 at 100KHz.

First we did an Open Cal with the spacing ~ 0805 spacing, then measured an "Open" at same spacing with average of 64 and saw a variation of ~3fF total.

Then we flipped over the SMD fixture so the H terminal is driving the SMD Movable Plunger, here we saw a Open variation of ~7fF. We returned to upright position and mounted a 0805 C0G 0.3pF capacitor and witnessed a ~6fF variation, then flipped over the SMD fixture upside down and had ~13fF variation, all with 64 averages at 100KHz.

Next first we did an Open Cal upside down,  and repeated the above with 0805 C0G 0.3pF results showing ~7fF variation upside down and ~5fF normal.

Don't be angry with me, but I very much doubt that a TH2830 can still resolve values in the femto range accurately, I consider that to be "background noise".

Agree, it can't!!

However it can resolve with good repeatability to 10~20fF with carefully setup.

For this level one must have an extremely stable mechanically and electrically setup. We prefer the modified LCR Meter SMD Fixture with the 4 cover attachment bolts. This has proven more stable due to the superior mechanical construction over the 2 bolt types. Also for the "Open Calibration" use a Plunger Spacing similar to the expected DUT spacing, the best way we've found is to wedge the Fixture Lever Arm so the Plunger Spacing is appropriate.

Here's a measurement of a quality 0603 C0G "Ref Cap" we use. Measured 4.68440pF with TH2830 & SMD Fixture @ 100KHz on 12/23/22 and just now 4.68808pF shown. Over the years we made similar measurements with similar results, with the IM3536 even better

Anyway, hope this shows what we've been talking about and the better of the two cheap SMD Fixtures we have, expect the quality Tonghui types should be even better. If anyone wants to ship us one we'll be happy to evaluate such

Edit: Added images of SMD Fixture Interior, and lastly what we got putting the fixture back together and remounting on TH2830 with no recalibration (using previous Cals)!!!

Best, 

[Martin72]

Things have been quiet with the guys from ET for 3 days now, I've also given them a few nice "tasks" along the way.
Before that, they always replied within 12 hours - that's outstanding.
This is a nice device, but it could be even nicer, so I suggested the following to the guys:
-Implementation of an average function
-Implementation of a DCR function
-Implementation of an ECAP measurement function
-Screenshot not only for the measurement screen
-More than one screenshot can be saved
-Reduction of the file size to a realistic size, at the moment it is over 1Mb for a 640x800 pixel image, something is wrong.

The first three points are available on the ET44 and ET45 models, so should not be a major problem.
Actually...
The guys have thanked me politely, let's see if it stays that way.

[Martin72]

Unfortunately I have a cold, but at least there is one good thing, the package with the SMD Fixture arrived today...
Makes a very solid impression for the fact that it doesn't even cost half of what sourcetronic wants for it (let alone other brands).
I also tried it out on a 1Ohm resistor, then I remembered that the ET35 can also screenshots...
Hence a screenshot of the smallest inductance I have here, a 10nH choke.
Measured at 300kHz and of course open/short calibrated beforehand.
That looks good and plausible.
I would also like to have a test fixture like this for our ST2830 for work, so I will order one - but then one from Sourcetronic.
This has several advantages, as you can imagine...

[KungFuJosh]

Looks like your generic SMT fixture is nicer than mine. Mine only had 2 mounting lock tabs (though the pictures did show 4).

[Martin72]

I took a few more pictures.
It doesn't look too bad.

[egonotto]

Hello,

I don't quite understand the image 10nH.png. Shouldn't VAC be around 0.2 mV or IAC around 100 mA?

Best regards
egonotto