East Tester ET4410 ESR Measure

[TimFox]

In the G-R "Digibridges", the analog circuitry was separate from the computation circuitry, which allowed higher currents and lower noise voltage.
The G-R/Quadtech 1659 uses an NE5534 op amp with a high-current buffer to drive the "low" terminal with the high current required from a powerful source through the DUT.
This allows measuring high current at virtual ground (0 V).
An earlier version in my inventory of not-yet-repaired units used a CA3130 CMOS op amp with multiple CMOS inverters to buffer the output, if I remember correctly.

[Martin72]

ET4410:
The signal output stage to H-Cur got a push-pull stage, the output resistance is switchable from 100Ohms to 30Ohms, maximum output voltage is 2Vrms.
Pic shows the actual state of my attempt to draw the circuit, according to the keysight handbook, it´s a auto-balancing bridge.

[mawyatt]

The "low" terminal should be a transimpedance amplifier with a selectable feedback resistor to "force" the low side to ~ 0 volts. At very low impedances the DUT current may reach ~0.1 amp in lab type LCR meters and the transimpedance amp must sink/source this at 0 volts and use as low as 10 ohm feedback resistor. This is likely a buffered Op-Amp type transimpedance amp. The "high" side is likely driven by another buffered op-amp with a selectable output resistance as low as 30 ohm to drive the DUT.

Both the transimpedance amp output and the DUT voltage are sensed with a synchronous technique which imposes a narrow bandpass filter around the injected signal which dramatically improves noise, interference rejection and sensitivity, a tried and true technique employed for well over half a century.

Best,

[Martin72]

Yepp,
Not so bad for a cheap one.
Test a panasonic FM electrolyt cap, 1000µF/25V on 100khz a few minutes ago.
ESR 0.034 Ohm
D 0.128
Xc -0.267  Ohm
Q 7.81
Phase - 82.7°
And cap....5.95µF 

[TimFox]

Quickly approaching self-resonance at that frequency...

[mawyatt]

ET4410:
The signal output stage to H-Cur got a push-pull stage, the output resistance is switchable from 100Ohms to 30Ohms, maximum output voltage is 2Vrms.
Pic shows the actual state of my attempt to draw the circuit, according to the keysight handbook, it´s a auto-balancing bridge.

If you trace the Low Cur terminal on the PCB, you'll probably find another Buffered type amplifier with all the selectable resistors shown or the right. With a 30 ohm output and 2 volt signal the max DUT current is 67ma, that's a lot of current for a small Op-Amp to handle without some aid!!

Best,

[Martin72]

Actually can´t find it in the manual, but there is a "overload" warning I´ve already reached sometimes..

[mawyatt]

Noted a few things in the schematic you've shown. First off there is a simple technique we've used for decades to "assist" with cross-over distortion in buffered outputs. The 200 ohm resistor R210 from the op-amp output to the buffered output serves to supply current during the cross-over around the zero point where the Slew Rate is the highest and the buffer devices may have trouble keeping up since they are lightly biased there. As the signal increases the Slew Rate slows some and output transistors begin to take the current load away from the op-amp and after 4~5ma the op-amp maxes out. If implemented properly this works very well.

Another couple points, it appears the meter is utilizing the +-5V supply as somewhat references since the signal waveform seems to be coming from the controller and would be a 0 to 5V pulse train to create the signal sine wave and the multipole filter is to filter out everything and leave a nice clean sine wave. The result from the filter will be a unipolar sine wave and this is offset to swing about zero by using the -5V and a resistor R215 (also R205) into the summing node op amp configuration. Using the controller to create a PWM pulse train eliminates the need for a sine wave generator DDS & DAC. Nothing wrong with this approach as it saves some $ in the BOM, but doesn't possess the precision of a typical lab type instrument, but doesn't cost as much either, so very cost effective!! Think todays linear regulator chips are much more stable than those from long ago, know the ones we've utilized were pretty stable and almost spot on the desired output voltage.

Best,

[The Electrician]

Yepp,
Not so bad for a cheap one.
Test a panasonic FM electrolyt cap, 1000µF/25V on 100khz a few minutes ago.
ESR 0.034 Ohm
D 0.128
Xc -0.267  Ohm
Q 7.81
Phase - 82.7°
And cap....5.95µF 

Which meter did you use to make this measurement?  Something's wrong with these numbers.  The phase can't be anywhere near -90 degrees for a 1000uF electrolytic at 100 kHz.  Would you please do this over.  Also show the numbers at 100 Hz for comparison, and measure a few caps larger than 1000uF at 100 kHz please.

[mawyatt]

Actually can´t find it in the manual, but there is a "overload" warning I´ve already reached sometimes..

Now get yourself a DIY DC Bias Fixture and see just how bad some of the ceramic caps vary with DC bias, also check out the Mylar film types, and electrolytic, then you're nice Mica and Polystyrene types!! Then try some semiconductor diodes to see how they behave with DC bias, fun stuff indeed!!

Oh, and don't forget the DIY Isolation Transformer for Bode Plots of Open Loop parameters with your superb SDS HD DSO

Edit: And forgot, put that nice Twin T notch Filter that you have in the feedback loop of an op-amp, makes nice narrow bandpass filter. Get a high Q inductor and measure it along with your high Q Mica caps, then create a Peltz oscillator!! Fun stuff ahead 

Best,

[The Electrician]

Here are a couple of plots vs frequency of parameters of a 1000uF 16v Rubycon capacitor.  The first plot shows the impedance magnitude |Z| (green) and ESR (yellow) over a frequency range of 100 Hz to 5 Mhz.  You can see that starting at the left at 100 Hz, |Z| decreases with increasing frequency as we would expect for a capacitor, but somewhere before 100 kHz where marker B is, |Z| begins to increase.  A |Z| that increases with frequency is behavior that we would expect from an inductor, and that increase is due to the ESL (equivalent series inductance) of the capacitor.  The capacitance of the capacitor resonates with the ESL of the capacitor just below (it's at marker A in the second image) the frequency where marker B is; this is called the self resonance frequency of the capacitor.

At frequencies above the SRF (self resonance frequency), an LCR meter which is set to measure capacitance will show a reading which is negative, because above the SRF the impedance of the capacitor is inductive.



The second plot shows the same yellow curve of the ESR vs frequency, but now the green curve is the phase angle of the impedance of the capacitor being measured.  The scale of the linear display is -90 degrees at the bottom, +90 degrees at the top.  The green curve starts at the left (100 Hz) near the bottom, so at low frequencies the phase of the capacitor's impedance is nearly -90 degrees.  As the frequency increases, the phase moves upward (but is still negative--capacitive in other words) until at marker A (65.027 kHz), the phase has reached zero degrees--the impedance is purely resistive.  Continuing up in frequency, the phase angle continues to increase toward +90 degrees, staying positive, which means it's inductive.

Notice the behavior near 100 kHz; the phase is close to zero degrees, and it can't be -82.7 degrees.

[mawyatt]

Nice plots 

Best

[Martin72]

Today a electrolytic cap on my ET4410, same cap then on monday with the lcrs on work.
It´s a low esr type from panasonic, 100µF/35V.

Results with 120hz, 1Vrms, series:

C: 102.96µF
Xc : -12.882 Ohm
D: 0.0208
Q: 47.388
Phase: -88.79°
ESR: 0.2699 Ohm (stable.. )

Results keysight U1733C:
103.65µF
D: 0.018
Q: 59.1
Phase: 89.0°
ESR: 0.273 Ohm
Xc: forgotten

Results ST2830:

103.30µF
D: 0.0198
Q: 50.47
Phase: 88.86°
ESR: 0.255 Ohm
Xc: forgotten

I would say, all three LCRs are close together.
The panasonic 1000µF cap on 100khz:
Keysight : OL
ST2830: -32.57µF, 0.0154 Ohm (ESR), Q 3.16, D 0.317, forgotten the phase (tomorrow)

[The Electrician]

Yepp,
Not so bad for a cheap one.
Test a panasonic FM electrolyt cap, 1000µF/25V on 100khz a few minutes ago.
ESR 0.034 Ohm
D 0.128
Xc -0.267  Ohm
Q 7.81
Phase - 82.7°
And cap....5.95µF 

This is the set of measurements from reply #103 that I think are not right.  In reply #112 today you gave some numbers that are from a Panasonic 1000 uF cap.

"The panasonic 1000µF cap on 100khz:
Keysight : OL
ST2830: -32.57µF, 0.0154 Ohm (ESR), Q 3.16, D 0.317, forgotten the phase (tomorrow)"

 Is this the same cap as in the quote above? These numbers are quite different.  What meter did you use to make the original measurements in the quote from reply #103?  Can you make measurements of the LCR meters at work of this Panasonic 1000 uF cap?

[Martin72]

Hi,
First measurements were taken at home on the ET4410, the last posted were at work on calibrated LCRs like keysight U1733C and Sourcetronic ST2830.
Same caps used like home, the 100 and 1000µF from panasonic.
Forgot the Xc parameter, so repeat it in the next days.

[The Electrician]

There's something odd about that Panasonic 1000 uF cap.  Besides making the measurement on all the meters you have access to at 100 kHz, would you please make a measurement at 100 Hz, 1 kHz, and 10 kHz on the best meter.

Edit: Another thing you might do is find some other 1000 uF electrolytic capacitors and make a few measurements at 100 kHz, to see if they are much different then the Panasonic one.

[mawyatt]

"The panasonic 1000µF cap on 100khz:
Keysight : OL
ST2830: -32.57µF, 0.0154 Ohm (ESR), Q 3.16, D 0.317, forgotten the phase (tomorrow)"

Does the negative capacitance reading at 100KHz of the 1000uF Panasonic indicate an inductive reactance due to ESL, which would be past the series resonate frequency? If so, then the phase should be positive.

Best, 

[mawyatt]

Today a electrolytic cap on my ET4410, same cap then on monday with the lcrs on work.
It´s a low esr type from panasonic, 100µF/35V.

Results with 120hz, 1Vrms, series:

C: 102.96µF
Xc : -12.882 Ohm
D: 0.0208
Q: 47.388
Phase: -88.79°
ESR: 0.2699 Ohm (stable.. )

Results keysight U1733C:
103.65µF
D: 0.018
Q: 59.1
Phase: 89.0°
ESR: 0.273 Ohm
Xc: forgotten

Results ST2830:

103.30µF
D: 0.0198
Q: 50.47
Phase: 88.86°
ESR: 0.255 Ohm
Xc: forgotten

I would say, all three LCRs are close together.
The panasonic 1000µF cap on 100khz:
Keysight : OL
ST2830: -32.57µF, 0.0154 Ohm (ESR), Q 3.16, D 0.317, forgotten the phase (tomorrow)

Shouldn't this 100uF cap measured at 120Hz have a negative phase?

Best,

[The Electrician]

"The panasonic 1000µF cap on 100khz:
Keysight : OL
ST2830: -32.57µF, 0.0154 Ohm (ESR), Q 3.16, D 0.317, forgotten the phase (tomorrow)"

Does the negative capacitance reading at 100KHz of the 1000uF Panasonic indicate an inductive reactance due to ESL, which would be past the series resonate frequency? If so, then the phase should be positive.

Best,

There a number of things that make me think something's wrong, either with the capacitor itself, or the measurement.

The plot of the impedance phase of a 1000 uF cap in reply #110 was intended to show that the self resonance of the typical 1000 uF electrolytic is near 100 kHz, and the phase is near zero degrees.  I don't think it can be vary much from zero, plus or minus.

The measurements in reply #103 gave this for the phase: "Phase - 82.7°".  Is that a minus sign or is it a hyphen?  Whichever it is, I don't think the magnitude of the angle can be as great as 82.7  Also the Q is given as 7.81 (the measurement reported today gives it as 3.16); I measured several 1000 uF electrolytic caps, and I can't find any with Q that high.

Would you measure a few 1000 uF electrolytics and see what you get for Q at 100 kHz?

[mawyatt]

Would you measure a few 1000 uF electrolytics and see what you get for Q at 100 kHz?

The only 1000uF electrolytic cap we have is an inexpensive Beryl 1000uF @ 16V RC(M) 2110 105C type.

https://datasheet.lcsc.com/lcsc/1912111437_BERYL-Electronic-Tech-RC016M102LO10-16TH-2A1E_C357639.pdf

Measured with Tonghui TH2830, Hioki IM3536 and DE5000. Very quickly done without calibration or a proper warmup.

TH2830         100Hz              1KHz              10KHz             100KHz 
C                913uF               874uF             1,117uF         -105uF
ESR             0.109               0.051               0.048              0.057
Angle           -86                   -74                  -16                  +15
D                 0.06                0.28                 3.55                 3.80

IM3536       
C               911uF                878uF              1,243uF         -66uF
ESR           0.120                 0.056               0.053             0.063
Angle          -86                     -73                  -14                +21
D               0.07                   0.31                  4.14              2.56

DE5000
C               923uF                875uF              OL                  -101uF (computed from Ls)
ESR           0                        0.04                                       0.04 (Rs)
Angle           -87                      -77                                     +19
D                0.06                   0.23                                     2.89 

Appears the SRF is ~24KHz (Z magnitude angle ~0). Note the 100KHz measurement with DE5000 was with Ls and Rs, C computed from ~ 1/[(2*pi*100KHz)^2 * Ls]

Anyway, this seems like a typical low end cap, not sure about the Panasonic Martin measured tho, maybe a special Polymer type or something?

Best,       

[The Electrician]

Martin72, do you have a link to a spec sheet for the Panasonic 1000 uF cap you measured in reply #103?

If not, do you know if it's a polymer capacitor, or some other type that would be exceptionally low loss?

[MegaVolt]

Here are a couple of plots vs frequency of parameters of a 1000uF 16v Rubycon capacitor.

With what device these wonderful graphs were taken?

[The Electrician]

This: https://www.hioki.com/us-en/products/lcr-meters/3-ghz/id_5772

Also Google this: IM3570E14-23M.pdf

[mawyatt]

For those of us without a nice Hioki IM3570 we can generate some decent plots of DUT Capacitance Impedance Magnitude and Phase with a scope and signal generator. Simply sweeping the source frequency over the desired range and measuring the amplitude and phase of the voltage across across the DUT and current thru the can yield the result with a little math.

Z = V/I (Magnitude and Phase), where V is voltage across the DUT and I is the current thru the DUT which can be measured by inserting a known resistor R in the signal path to the DUT. Then the DUT current is Voltage across the resistor divided by R.

If you have a modern DSO like the Siglent SDS2104X+ which has a built-in Bode function you can employ this to make the nice plots directly on the screen in dB magnitude and phase of Cap DUT Z vs. frequency.

If folks are interested we could start another thread on this technique, so as not to dilute this thread.

Anyway, here's a plot of a 1000uF 16V electrolytic capacitor, another 10uF, 680uF and 100uF. The scale is in dB 1000 ohms and phase is in degrees.

Best,

[Martin72]

So...
Today I´ve measured the both panasonic caps at work again, the menu of the ST2830 is not the clearest one.
A few minutes ago I did the same again at home on the ET4410, to have it all in one post.

The 100µ cap at 120Hz:

ST2830                  ET4410
103.16µF               103.33µF
Xc -12.86 Ohm      -12.84 Ohm
Q 51.10                  51.80
D 0.0196                0.0192
ESR  0.252 Ohm      0.250 Ohm
Phase -88.88°          -88.89°

At 100Khz:

26.50µF                  15.40µF
-0.071  Ohm            -0.104 Ohm
1.607                      2.427
0.6185                    0.4115
0.038 Ohm              0.043 Ohm
-57.86°                   -67.63°

1000µF at 120Hz

ST2830                  ET4410
945.91µF                946.8µF
Xc -1.402 Ohm       -1.401 Ohm
Q 28.60                  29.04
D 0.035                  0.035
ESR 0.049 Ohm       0.048 Ohm
Phase -87.99°         -88.03°

At 100Khz

72.50µF                  23.80µF
-0.025 Ohm             -0.067 Ohm
1.578                      3.756
0.632                      0.266
0.014 Ohm              0.018 Ohm
-57.31°                  -75.20°

At 120Hz they are like twins..

Datasheets:
1000µF/25V FM Type:
https://www.micro-semiconductor.com/datasheet/37-EEU-FM1E331LB.pdf

100µF/35V FR Type:
https://industrial.panasonic.com/cdbs/www-data/pdf/RDF0000/ABA0000C1259.pdf