East Tester ET4410 ESR Measure

[Martin72]

The measurement is stable with electrolytics because they are lossy; their D is never going to be as low as .001, or even close to that.

Yap, see https://www.eevblog.com/forum/testgear/east-tester-et4410-esr-measure/msg4297858/#msg4297858

[mawyatt]

Likely our ESR readings are also stable due to capacitor loss. These cheap film types we have are nowhere near the Q of the Silver Mica or Polystyrene types. Here's what the Tonghui TH2830 and Hioki IM3536 show at 10KHz for a 10nF cheap film type.

Best,

[TimFox]

A comparison of results with my DE-5000 using a home-built four-terminal sense of two binding posts (0.75 in spacing).
DUT:  SBE 715 (now CDE), 10 nF/5%/600 V film-foil polypropylene ("orange drop") radial leads.
Test mode:  series capacitance Cs
Manufacturer's spec:  Q > 3400 at 20 kHz, Q > 2500 at 100 kHz.

100 Hz:  9.932 nF  Q = "OL"  ESR cycles from 10 to 130 ohms
120 Hz:  9.929 nF  Q = "OL"  ESR cycles from 5 to 56 ohms
1 kHz:    9.927 nF  Q = "OL"  ESR cycles from 0.02 to 0.3 ohms
10 kHz:  9.925 nF  Q = "OL"  ESR stable at 0.01 ohms
100 kHz: 9.921 nF  Q = 1911  ESR stable at 0.08 ohms

The "cycling" ESR behaves like a relaxation oscillation between the limits.
The last (finite) Q display settled to that value after maybe a dozen readings.
The stable ESR values settled after only a few readings.
Obviously, this high-Q part is at the limit of loss reading for this meter.
Phase angle is always 90.0 deg.

[indman]

I measured the silver mica cap from reply #20 on my DE5000.  Here's what I got.
At 1 kHz, the value of the cap is good. D is 0.000, Q is OL, ESR is 1.86
At 10 kHz, the value of the cap is correct.  D is 0.000, Q is OL, ESR is 0.18 
I found a 10 nF polystyrene capacitor and measured it with the Hioki IM3570:

Now I'm starting to understand why the DE-5000 can't correctly measure high quality film capacitors at different frequencies. Can you measure similar capacitors on your Hioki IM3570 and confirm my findings in this thread?
https://www.eevblog.com/forum/testgear/any-one-mod-the-de-5000-for-1mhz-frequency/msg2676075/#msg2676075

[Martin72]

Come on, not again...

[indman]

Martin72,Why are you so annoyed? No one has ever given me a reasonable explanation for this device behavior.

[Martin72]

I don´t think this is really an issue, it goes more in the direction couldn´t get any better.
The DE5000 is an fantastic measure device for its price - when you´re basically interested on the main values like L,C,R...
I´ve did "tons" of measures comparing to calibrated LCRs, it never fails.
The ET4410 got better accuracy, is 3-times more expensive, did some things better . but seems to have "problems" similar to the DE5000 and other cheap LCRs.
But let us see what this thread will bring more on informations, which could eventually "used" for understanding the DE5000.

[TimFox]

For the quite reasonable price, my only problem with the DE-5000 is that when measuring a capacitor or inductor above its self-resonant frequency, it does not indicate negative capacitance or inductance.  For that application, I wish it had R and X (with sign) direct readout.  I learned decades ago not to trust the “minority” component’s accuracy (e.g. ESR for good capacitor) when Q > 1000 on not-super-expensive equipment.

[Martin72]

Yep,
And a closer look to the "chipset" of the DE5000 offers some informations, how it works and where the limitations are.
BTW, this chipset is used by several brands, even on those who cost more than double.
When you want "more", you must pay more.

[indman]

The DE5000 is an fantastic measure device for its price - when you´re basically interested on the main values like L,C,R...

Your explanation is not convincing to me, I'm sorry. A device that claims to measure secondary parameters Q, D, ESR, tg on a par with the primary parameters and cannot be trusted to make these measurements is NOT A MEASURING DEVICE! And I don't care about the cost(price) if the manufacturer can't guarantee the real specifications.

[Martin72]

Therefore I said not again...
We can discuss about in one of the DE5000 threads, here it´s off-topic.

[indman]

Then eliminate the mention of the DE-5000 from every 2nd post in this thread and ban others from mentioning it in comparative measurements! What difference does it make what is the name of the LCR measuring device if we are discussing the principle of its operation? We speak different languages, alas.

[Martin72]

Meanwhile I´ve "studied" some methods of ESR measuring....interesting. 
This brings me to the point to know how the ET4410 measures ESR and why it´s so "bouncing".
Last can depend on the noise, no, better, will depend.
Therefore I begin to draw a schematic for the ET4410, starting with the output stage.
Maybe on this way I find also a "solution" also for the famous DE 5000 and can make people happy, who are expecting everything from a 100 bucks instrument like having one for 1000 or more.

[The Electrician]

Meanwhile I´ve "studied" some methods of ESR measuring....interesting. 
This brings me to the point to know how the ET4410 measures ESR and why it´s so "bouncing".
Last can depend on the noise, no, better, will depend.
Therefore I begin to draw a schematic for the ET4410, starting with the output stage.
Maybe on this way I find also a "solution" also for the famous DE 5000 and can make people happy, who are expecting everything from a 100 bucks instrument like having one for 1000 or more.

The phrase "methods of ESR measuring" spans several kinds of instruments.

Meters designated "ESR meter" are very specialized and generally only measure the ESR of capacitors at 100 kHz, and typically electrolytic capacitors of more than 1 uF.  They can't measure inductance, or ESR at frequencies other than (typically) 100 kHz.  For information on those types of meters see this website:
http://kripton2035.free.fr/esr-repository.html

Meters designated "LCR meter" or "impedance analyzer", if they are  modern digital instruments, make measurements using the methods described here: http://kripton2035.free.fr/LCR%20meters/lcr-theory.html
There is more information on these meters here: http://kripton2035.free.fr/lcr-repository.html
Within this repository there is info on so-called "LC meters" which use yet another technique.

Your ET4410 is a full-fledged "LCR meter" and the reason for the "bouncing" ESR reading is what I already explained in reply #20.  Even though it is low cost, the DE5000 uses the same technique described in the references in: http://kripton2035.free.fr/LCR%20meters/lcr-theory.html, and its limitations are mainly due to signal to noise ratio.

[Martin72]

Thanks.
Nevertheless I want to try it out making a schematic, today I finished the output stage.
(Resistor values are not correct, except the output)

Edit: Added signals from PA4 and PA5 of the controller.

[Martin72]

Man, it´s getting more and more difficult...
Problems identifying sot23, is it a diode, a transistor...
Signal out path to the Hi-pot/cur socket was easy, but the way back....
6 pcs CMOS switcher circuits, several OPs, tons of resistors/caps...
It´s much more complicated than on DE5000, therefore it wonders me a little bit having the problems.
It´s a pity holidays are going to end, otherwise I can test on monday with the lcrs at work.

[The Electrician]

Man, it´s getting more and more difficult...
Problems identifying sot23, is it a diode, a transistor...
Signal out path to the Hi-pot/cur socket was easy, but the way back....
6 pcs CMOS switcher circuits, several OPs, tons of resistors/caps...
It´s much more complicated than on DE5000, therefore it wonders me a little bit having the problems.
It´s a pity holidays are going to end, otherwise I can test on monday with the lcrs at work.

Are you doing all this tracing circuitry in the ET4410 to try to figure out how it works?

To know how a particular meter calling itself an "LCR meter" works, only a few clues are needed.  If it applies a sine wave to the DUT, which you showed the ET4410 does in reply #39, and if it allows the user to choose the frequency used to test the DUT, and if the frequency range covered is fairly large, then the method used is almost certainly the ones shown here: http://kripton2035.free.fr/LCR%20meters/lcr-theory.html

The commercial so-called "ESR meters" all (with one exception as far as I know, the Capacitor wizard) use pulse or square waves of 100 kHz (some DIY "ESR" meter projects may use a slightly different frequency, such as 40 kHz) .  They don't allow the user to choose the frequency of measurement, and they can't make a measurement on a capacitor of less than about 1uF.

[Martin72]

Are you doing all this tracing circuitry in the ET4410 to try to figure out how it works?

Yes and no.
How it works I know, but I want to know more exactly to find out where the problem lies - And maybe "correct" it.
You can supply up to 2Vrms into the DUT, that´s way more the most LCRs offers and we´re talked about signal to noise ratio why the ESR thing "sucks".
Also you can change output resistance, bias level..The ET can do things other cheapos can´t and you probably got an idea why the ESR measure´s annoying me.
Some posts before, the ET measure the most things right and if it would "know" the tan-phi, it would display the ESR correct, because all other measured values are correct.

What the cap wizard concerns I agree that this is not a good example because of the fixed frequency which make comparisons useless under the 100khz and because the capacitor limitations.

[mawyatt]

It´s much more complicated than on DE5000, therefore it wonders me a little bit having the problems.

Please remember the DE-5000 utilizes a highly integrated chip set, apparently the ET4410 doesn't. This chip set integrates much of the circuitry required to perform the LCR measurements, although don't know the details since we've never designed with this chip set.

Best,

[Martin72]

I hoped someone come to argue this way..
I know about the 2-Chipset on the DE (and other brands), chances are high that in the end there´s nothing really different except output and bias.
We´ll see.

[Martin72]

Hi,

Today I took one of the glimmer caps to work and measure it with our calibrated LCRs..
A little surprise (for me):
The sourcetronic showed a value at 1khz lower as the ET4410 did (appx 3 Ohm vs 4.9) and a little bit unstable...
The keysight U1733C you can completely forget, it "freaks out" by the attempt to show the value(Display blinking).
The ET4410 cost about 350€, the keysight 700€ and the sourcetronic 1200€.
All seems to have in common that they can´t measure the ESR of foilcaps correctly.
Interesting in a negative way..
In the next days I´ll measure a electrolytic cap on all the three, if there are remarkable differences between them.

[TimFox]

 Some more quantitative values for your "glimmer capacitors" at 1 kHz.
10 nF at 1 kHz has a reactance of 15,900 ohms.
With an ESR value of 4 ohms, that is a very high Q = 4000.
That corresponds to a phase difference (from the ideal 90 deg) of only 0.014 deg, which is obviously difficult to measure given your measurements (and mine on the DE-5000).
I once discussed this problem at a trade show with an engineer at SBE, who worked with pulse-rated polypropylene capacitors that have absurdly high Q values for high-current applications.
He said they had to go to calorimetry to measure the loss of their units:  run AC current through them and measure the temperature rise.

[The Electrician]

Hi,

Today I took one of the glimmer caps to work and measure it with our calibrated LCRs..
A little surprise (for me):
The sourcetronic showed a value at 1khz lower as the ET4410 did (appx 3 Ohm vs 4.9) and a little bit unstable...
The keysight U1733C you can completely forget, it "freaks out" by the attempt to show the value(Display blinking).
The ET4410 cost about 350€, the keysight 700€ and the sourcetronic 1200€.
All seems to have in common that they can´t measure the ESR of foilcaps correctly.
Interesting in a negative way..
In the next days I´ll measure a electrolytic cap on all the three, if there are remarkable differences between them.

What is a "glimmer" capacitor?

It's not that they can't measure the ESR of foil caps generally; they just can't measure the ESR of very low loss foil caps.

With all the measurements of ESR, also measure D or Q.  If D<.001, or Q>1000 you can expect lower performance meters to have trouble.

The meters you are comparing can probably get better correspondence among measurements if you were to use foil caps with more loss.  Find a foil cap with d>.01 or Q<100 and measure that cap on all three meters.

See if you have any mylar capacitors on hand; they are usually fairly lossy.

[The Electrician]

I found a couple of capacitors, one of ordinary loss, and the other is low loss.

Those measured values I'm relying on to be accurate were made on a Hioki IM3570.

Capacitor A has a D value of .00135 at 1 kHz, and a D value of .0025 at 10 kHz.
Capacitor B has a D value of .000184 at 1 kHz, and a D value of .000256 at 10 kHz.

Capacitor A was chosen because its D value of .00135 is slightly larger than .001, which is where the low performance (cost) meters will begin to have trouble getting a good measurement.  What I mean by "have trouble" is that the measurement of secondary parameters D, ESR, Q will be incorrect or noisy (fluctuations in the measurement digits).

I would expect that low performance meters will be able make reasonably accurate measurements of Capacitor A, but they will get bogus results for capacitor B, whose D value of .000184 is substantially less that the threshold of .001

Here are some measurements of these two capacitors made with a high performance meter (Hioki IM3570), and a popular low cost meter (DE5000):

Code: [Select]

                                             Hioki IM3570                                 DE5000
                                   @1kHz                 @10kHz                    @1kHz            @10kHz

                           C     10.435                 10.409                   10.446             10.419
                         ESR      20.4                    3.90                    19.3               3.51
Capacitor A               D      .00135                 .0025                    .001                .002
                          Q        735                     393                    785                  434

                                              Hioki IM3570                                DE5000
                                  @1kHz                @10kHz                     @1kHz            @10kHz

                           C     10.268                 10.267                   10.269             10.274
                         ESR      2.84                   .396                    1.03                .16
Capacitor B               D     .000184               .000256                   .000                .000
                          Q       5450                  3920                      OL                   OL


Capacitor A has a 1 kHz value of D of .00135, which is somewhat larger than the threshold value of .001, and we see that the values measured by the DE5000 are quite comparable to those obtained by the Hioki.

Capacitor B has a 1 kHz D value of .000184, which is quite a lot smaller than the threshold; we should expect that the values obtained by the low performance DE5000 would be  considerably in error, which they are.

These experimental values show that low loss capacitors can't be correctly measured by low performance (low cost) meters.  The user should always measure D or Q as well as ESR.  If the D or Q values are beyond the threshold values of D<.001 or Q>1000, expect that the measurements will be substantially in error.  IF the capacitor being measured is very low loss, low performance meters may not be able to make any measurement at all, or will return totally bogus results.

Even if the value of D is not checked, the fact that the ESR value is unstable means that the capacitor being measured is very low loss, and if the meter is a low performance unit, the measurements may be wrong. The front end noise of the meter is contaminating the measurement.  You just can't correctly measure secondary parameters of a very low loss capacitor with a low performance meter, although the capacitance measurement will be good.

[MegaVolt]

Introducing a small additional series resistor can help stabilize the result.