Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

ESR of Film Capacitor?

Status
Not open for further replies.
Datasheet shows -

1600867071786.png



Regards, Dana.
 
  • Like
Reactions: zenerbjt

    Z

    Points: 2
    Helpful Answer Positive Rating
I don't understand the question, the ESR isn't specified in the datasheet. It can be measured with a LCR meter. According to |Z| curve minimum, I would expect an ESR value of 5 to 10 mohm.
 
  • Like
Reactions: zenerbjt

    Z

    Points: 2
    Helpful Answer Positive Rating
I don't understand the question, the ESR isn't specified in the datasheet. It can be measured with a LCR meter. According to |Z| curve minimum, I would expect an ESR value of 5 to 10 mohm.
Thanks, the tan delta is given on page 19 of the datasheet. This is used to calculate ESR = (tan d) x Xc

Thanks danaddak for the Z curve... as you know, that shows "Z vs f"...which does not tell us ESR. The "Z" in that graph is a combination of the C, L and Rseries and Rparallel in the capacitor. At the minimum point, that may equal the ESR, but only at that point.
 
Last edited by a moderator:

Yes i see your point, i suppose they dont give tan_d for higher than 1khz for >1uF because the Lseries makes it impossible to measure accurately.
The thing is, the "z vs f" graph only tells us that ESR for the 22uf cap is approx 10 milliohms at 500khz....whereas our switching frequency is 100khz.

So at 500khz, Xc =XL. Which means the stray series inductance in the cap is 4.62nH (seems very low). So i suppose from this and the curve, we can work out ESR at 100kHz.

From the "Z vs F" curve, Z is approx 0.4 ohms at 100khz, so from that, and the knowledge of C and the stray series L, we should be able to work out the ESR at 100khz....would you agree?.......though this assumes that the series L is the same at 100khz as at 500khz.

....uuum, ive just done mag[R +j(wL - 1/wC)] = 0.4 at 100khz and find ESR = 0.4 Ohms at 100kHz...that surely can't be right.

I think this prooves that a frequency analyser is needed. Would you agree?
 
Last edited by a moderator:

No, that's not possible. Simply too inaccurate. ESR at different frequencies can be well determined with a LCR meter which does a vectorial measurement.
 

A handheld LCR meter typically can't measure 22 uF at 100 kHz. You either need a more expensive desktop instrument with wider impedance range or make a custom measurement circuit.

Many capacitor manufacturers specify ESR over frequency, by the way.
 
Cheap LCR - **broken link removed**

Of course you dont have control over freq in these low end, but I think the
ones more like $ 50 - $ 100 have that capability.


Regards, Dana.
 
Last edited:
  • Like
Reactions: zenerbjt

    Z

    Points: 2
    Helpful Answer Positive Rating
From the graph above ESR ~ 0.01 ohms best case for the 22uF ....
 
  • Like
Reactions: zenerbjt

    Z

    Points: 2
    Helpful Answer Positive Rating
Yes i see your point, i suppose they dont give tan_d for higher than 1khz for >1uF because the Lseries makes it impossible to measure accurately.
No, it's because the simple model of losses in the capacitor doesn't hold over a wide range of capacitances and frequencies. They're only providing dissipation factors for conditions where it will yield accurate results. The impedance plots offer a better idea of ESR vs frequencies for specific parts.
 
  • Like
Reactions: zenerbjt

    Z

    Points: 2
    Helpful Answer Positive Rating
I think this prooves that a frequency analyser is needed. Would you agree?

Here is the result of an impedance analyzer sweep of the impedance magnitude and ESR of a 20 uF MKT film capacitor. The frequency is swept from 100 Hz to 5 MHz. The impedance magnitude is the green curve and ESR is the yellow curve. There is a marker at 1 kHz and at 101.7 kHz. The value of ESR at the marker frequencies is shown in the upper right part of the image. The ESR at 100 kHz is about 9.8 milliohms.

MKT 20uF.png


Here is a sweep of a 20 uF MKP capacitor. Note that the ESR is noticeably better.

MKP 20 uF.png
 
The Electrician.....thanks for the amazing plots.
Very interesting to see how, unlike electrolytics, the ESR for the film cap is remarkably stable over frequency.

So the minima point was a good pointer to the ESR at all frequencies.......why couldnt the datasheet have put the excellent graph in that The Electrician shows here....
(now i'm looking for some way to print this excellent thread...this forum used to allow one to do this.....i presume it still does(?))
 

Hi,

Just in case, the Opera browser (and maybe most other browsers) has a pdf option and usually does a good job of pdf-ing the page, including forum thread pages.
 
The increasing ESR below 100 kHz shown in the plot isn't necessarily real. We usually see further decrease with high performance LCR meter. I admit that it's not simple to measure R/X < 0.001.
--- Updated ---

MKP tan d is said to converge against about 10e-4 in Epcos/TDK general information. MKT plot is more realistic.
 
Last edited:
lead and connection inductance and it's resistance with higher freq's can influence an impedance measurement sweep above 1MHz for even the best impedance analysers ....
 
lead and connection inductance and it's resistance with higher freq's can influence an impedance measurement sweep above 1MHz for even the best impedance analysers ....
Of course lead inductance will have an influence on a measurement. The lead inductance is part of the device under test, and the analyzer measures whatever impedance is presented to its measurement terminals.

The connection inductance and resistance should not cause a problem if a suitable fixture is used and if the parasitic inductance, capacitance and resistance are calibrated out with a open/short calibration.
 

you say " of course " but many appear to overlook this fact .... also cable capacitance of various connections to amplifiers / drivers and receivers ....
 

The increasing ESR below 100 kHz shown in the plot isn't necessarily real. We usually see further decrease with high performance LCR meter. I admit that it's not simple to measure R/X < 0.001.
--- Updated ---

MKP tan d is said to converge against about 10e-4 in Epcos/TDK general information. MKT plot is more realistic.
Do you have a recommendation for a high performance LCR meter?
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top