East Tester ET4410 ESR Measure

[Martin72]

Hi,

I´ve ordered a foilcap assortment to rebuild a notchfilter like the one in the sds2000x+ thread.
In this case it should be 10nF caps, in the box there are 10pcs, so I started to measure them with my ET4410.
While the capacity value is displayed constantly up to the third digit after the comma, the ESR value is floating.
It shows something between 7.XXX and 10.XXX Ohms alternating, never stable.
I thought, ESR will be caluclated so I wonder why the cap will be stable displayed, the ESR not.
And:
Taking a different cap, an Elko with 1000µF, ESR is stable displaying.
Are the caps bad, is it because of the low value of 10nf...
Will investigate it further, with other caps.

Martin

[mawyatt]

Hi Martin,

Can you try different frequencies and see if the ESR reading is still bouncing around? Does the meter have an averaging filter that can help?

For the Twin "T" Filter in the mentioned thread, the capacitance value is more critical than the ESR and the key is to try and match values of capacitance.

Edit: Likely not the 10nF caps. At the notch frequency (~1.6KHz in this case for 10nF and 10K) the capacitive reactance magnitude is equal to the resistance used, so a few ohms compared to 10K isn't going to matter much.

Best,

[Martin72]

Hi,

Will do it "tomorrow".

the capacitance value is more critical than the ESR

Yepp, this here is about the ESR measure behaviour itself- My thoughts viewing your model are, that you´ve measured the ESR to get an more precise model which matches better/can better compared to with the real measured one.

[mawyatt]

Yepp, this here is about the ESR measure behaviour itself- My thoughts viewing your model are, that you´ve measured the ESR to get an more precise model which matches better/can better compared to with the real measured one.

Exactly, and why we measured the capacitance and ESR at the expected frequency of the notch, ~1.6KHz for the 10nF and 10K. This allowed an accurate simulation of the filter behavior as compare to the Bode Plot measured result.

The SDS2104X+ performed admirably IMO, and your HD should really shine

Best,

[Kleinstein]

With foil capacitors the ESR is relatively small. So consider the 10 Ohms more like 0.010 kOhms and you have 2 or 3 stable digits after the decimal.
The capacity and ESR (more the invers of it) are measured togehter as a kind of complex number. With a near ideal, low loss (e.g. 0.1% range) capacitor the noise is relatively large in the loss direction, because the absolute loss is small.

Electrolytic capacitors on the other end have quite high losses (e.g. up to the 10% range) and ESR and this allows a similar number of stable digits for the capacitance and ESR.

[Martin72]

Seems plausible.

Meanwhile I´ve tried different cap-types, different values, but always foil caps - It´s always the same, no stable values.
It´s a pity in this case that I´m not at work for the next weeks for checking it against our other lcrs.

[mawyatt]

Might be an issue with that particular instrument, or the actual design. We just measured a cheap 10nF film cap with the DE-5000 and get reasonably stable ESR readings at 1KHz, 10KHz and 100KHz after a few seconds.

9.9944nF 61.6 Ohms @ 1KHz
9.877nF 9.48 Ohms @ 10KHz
9.710nF 2.16 Ohms @ 100KHz

Also should add, these type instruments measure the magnitude and phase of the Voltage across the DUT and the Current thru the DUT, then compute the Capacitance and Series Resistance, so believe the ESR is not directly measured but computed.

Best,

[Kleinstein]

The 47 nF capacitor at 1 kHz has an impedance in the 3 kOhms range. So 1 Ohms ESR is just 0.03 % of the impedance. Depending on how good the implementation is, this may be just a limitation of the instrument.  If not careful with the wires on may get parasitic effects, that may look a little like ESR, e.g. from inductive coupling.

The resilts from Mawatt don't look reasonable - could be a calibration problem, or an extremely poor old MP capacitor with high loss. So even if the readings are stable this would not help as they are way off for a film capacitor - maybe an X7R THT part that looks like a film cap ?.

[Martin72]

I´ve checked this also with a fixture:

https://www.sourcetronic.com/shop/de/lcr-adapter-26048a.html

To minimize parasetic inductive effects, same result.

[Martin72]

Edit:

Tested all displayable parameters on a 8.2µF MKP.
Stable readings by capacity, phase angle, capacitive resistance, D-Factor(OK, it´s the last digit that is changing a bit).
Unstable, really bouncing: ESR and Q-Factor.
Will test it further with all the caps I can find here...in the next days.. #

Martin

[mawyatt]

The resilts from Mawatt don't look reasonable - could be a calibration problem, or an extremely poor old MP capacitor with high loss. So even if the readings are stable this would not help as they are way off for a film capacitor - maybe an X7R THT part that looks like a film cap ?.

Don't think we have calibration problem, nor bad capacitors. All instruments calibrated and using an unused new stock WIMA Style Yellow 10nF film capacitor.

Freq               Hioki IM3536 with Kelvin Clips           Tonghui TH2830 with Kelvin Clips                      DE-5000 with Alligator Clips             

10KHz               9.99574nF with 15.5464 series R      9.99353nF with 15.6286 series R                   10.007nF with 14.85 series R

Best,

[mawyatt]

Just a thought, can you use a lower output resistance than 100 ohms? Might help stabilize the ESR readings.

Edit: Also try a higher drive voltage if you can change this.

Best,

[Martin72]

Hi,

Can do, will do..

Martin

[TimFox]

Might be an issue with that particular instrument, or the actual design. We just measured a cheap 10nF film cap with the DE-5000 and get reasonably stable ESR readings at 1KHz, 10KHz and 100KHz after a few seconds.

9.9944nF 61.6 Ohms @ 1KHz
9.877nF 9.48 Ohms @ 10KHz
9.710nF 2.16 Ohms @ 100KHz

Also should add, these type instruments measure the magnitude and phase of the Voltage across the DUT and the Current thru the DUT, then compute the Capacitance and Series Resistance, so believe the ESR is not directly measured but computed.

Best,

I believe that this type of instrument, evolved from the classic G-R Digibridge, measure the in-phase (“real part”) and phase-quadrature (“imaginary part”) of the voltage across the DUT, phase reference from the drive voltage (or the current through the DUT?) and divide by the device current to get the real and imaginary components of the DUT’s (complex) impedance, so the ESR is determined reasonably directly, rather than from the sine of a small angle.

[Martin72]

Will test it further with all the caps I can find here...in the next days..

You will always get stable values when testing electrolytic capacitors and always "unstable" vaues when testing foil-caps...
No, not at all.
There is a trend to watch, by many of my foilcaps (MKP, FKP) the ET4410 reads a phase of -90° exactly or really near by (89.95..99).
I got some MKT (audio crossover purposes), it´s the same thing but they have losses when using measure frequencies from 10Khz on.
Then the phase lowers slightly down to say -88...-89° and the ESR value becomes stable.
All electrolytics here got a phase of -72...-87° and the ESR value is stable.

Martin

[Martin72]

Today 10nF glimmer caps arrived with 1% tolerance.
Same as all before, capacitance and Xc values are stable displayed, ESR not.
The unstable values are annoying and at the end useless.

edit: ESR pic titles wrong

[mawyatt]

Today 10nF glimmer caps arrived with 1% tolerance.
Same as all before, capacitance and Xc values are stable displayed, ESR not.
The unstable values are annoying and at the end useless.

edit: ESR pic titles wrong

Those 10nF 1% caps look nice, but can't say much about the Easy Tester. What are the ESR specs on the caps?

Think this is something to do with trying to establish the ESR with an impedance angle of close to 90 degrees, whatever method is utilized. Altho other LCR meters we have like the handheld DE-5000, or Lab type TH2830 and IM3536 don't seem to have this problem.

Maybe others with the same Easy Tester can check with a film 10nF cap??

Best,

[Martin72]

According to the Specs, loss factor is about 0.001 at 1khz.

[Martin72]

Thinking further about...

When tan phi is 0.001 (specs) and capacitiy is 9.939nf at 1khz(measured), ESR is about 16.949 ohms calculated, see pic with formulars below.
Other calculation then:
Tan phi is ESR/Xc, take the calculated ESR before and then take the measured Xc on 1khz of 16.014kOhm, its about 0.00106 - It makes all sense.
The ET4410 should measure an ESR about 16...17 Ohms, but it didnt, it measures way to less and very unstable.
Because:
The ET4410 don´t know what we know, the tan phi according to the specs.
So how do the ET4410 determine the ESR...
And is it only a prob of this puppy ?
Maybe not...

[mawyatt]

Thought you had a DE-5000? If so, how does it behave with these caps?

Best,

[The Electrician]

Your problem is undoubtedly a matter of signal to noise ratio.

LSR meters operate as Timfox explained in post #13.  You can also find a good in depth explanation in the Agilent Impedance measurement handbook.  Google "impedance measurement handbook"

When you measure a capacitor, the meter applies a sine wave of whatever the nominal voltage is (1 volt for your measurement) and extracts the in phase and quadrature components of the resulting current.  The ratio of the two components is equal to the value of D.

So if your capacitor is low loss, the real part of the current is going to be very small.  For example, if the cap has a D of .001, the meter will be trying to measure a very small current/voltage (microvolts).  When the measured value of D < .001 (or equivalently, Q>1000), low cost meters are going to have problems.

A good meter will know when it's measurement is in the noise, and will not show very many digits for the result.  Watch out for that clue that the measurement is "iffy".  For example in reply #9 your measurement shows 4 digits for the capacitance, but only 2 digits for the ESR.

In reply #15, the 10 kHz measurement of ESR shows a value of .0085 ohms.  This is impossible.  The value of Q is Xc/ESR; that measurement would imply a Q of about 188000.

The Hioki IM3570 can be set to show 7 digits.  This is well beyond the accuracy of the instrument but it's handy for matching components, and watching drift with temperature.  Here's what I get when I measure a 10 nF silver mica capacitor, which is a very low loss capacitor, at 10 kHz:



The capacitance is shown as 7 digits, but only about 5 of them are stable.  ESR (Rs) is shown with 6 digits, but only the leading 2 digits are stable.  Note that the Q is about 6000; this is approaching the capability of the instrument.  When the Q>10000, the Hioki won't even show a value; it just indicates an out of range value.

I think that your East Tester just can't cope with very low loss capacitors.  I've noticed that some of the hand held meters I have exhibit the same limitations.

[Martin72]

@mawyatt
You take the knife, hit it in my heart and turning it around....
I´ve sold it not long time ago and I could bite in my a*** I did it...

[The Electrician]

@mawyatt
You take the knife, hit it in my heart and turning it around....
I´ve sold it not long time ago and I could bite in my a*** I did it...

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 

These two ESR values for this very low loss capacitor are completely bogus.  It's too bad the DE5000 doesn't recognize that it's bogus and refuse to show a value.  This is very misleading behavior.  A user of low cost LCR meters needs to know that as soon as the value of D is less than .001, the measurement will be noisy.  The Gold standard LCR meter is the Keysight E4980A (@ $20000) which can get measurements down to a D of about .0001 with maybe 50% accuracy at the limit.

Your East Tester is better than the DE5000.

I found a 10 nF polystyrene capacitor and measured it with the Hioki IM3570:



This capacitor is somewhat beyond what even the top of the line Hioki can do.  Notice that the capacitance is still given with 7 digits, but the ESR (Rs) has only one stable digit, and it is suspect.  The DE5000 and other handhelds and even low cost bench meters won't be able to measure ESR of such low loss capacitors; but they can still correctly measure the capacitance.

[The Electrician]

Hi,


Taking a different cap, an Elko with 1000µF, ESR is stable displaying.
Are the caps bad, is it because of the low value of 10nf...
Will investigate it further, with other caps.

Martin

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.

[Martin72]

Your East Tester is better than the DE5000.

Yap, it´s not the badest one as it got 16 selectable frequencies up to 100khz, output up to 2Vrms, Bias, 30 or 100Ohm selectable output resistance and finally a basic accuracy of 0.2% .
And the basic RCL measures it performs very well, but the "secondary parameter" thing...
A look into the ET4410 offers a good design of the circuits, but you can´t get schematics for it.
As I´ve asked for it, they (East-Tester) immediately send me the SCPI commands.
Now I´ve asked of the ESR thing at the beginning of the week and nothing happens.