Make a very simple ESR meter.

Discussion in 'Test & Measurement Forum' started by bob2, Jul 23, 2019.

  1. bob2

    Thread Starter Member

    Jun 15, 2019
    68
    5
    I measured the frequency. Exactly 90 kHz. The voltage at the measuring tips is 75 mV.
    With such a low measuring voltage, not a single semiconductor should open. Even the most sensitive Schottky diodes.
    It is possible to use a voltmeter.
     
    Last edited: Aug 10, 2019 at 11:45 AM
  2. MrAl

    AAC Fanatic!

    Jun 17, 2014
    6,210
    1,351
    That sounds about right for those component values.
    I calculated 88kHz so that is within about 2 percent.

    As i said before though, there could be an issue with the bandwidth of the op amps because the slower ones that are commonly used dont have the bandwidth to pass that frequency with any appreciable gain and output swing.
    So for example it may work at 90kHz and 1v peak output, but at 2v peak output there probably would be too much distortion.
    For any op amp at 90kHz the slew rate has to be at least 0.57 volts per microsecond per peak output volts. So for 2v peak output you'd need twice that slew rate. So the LM358 and similar types probably wont work very well unless distortion is tolerable.
     
    Last edited: Aug 12, 2019 at 7:22 AM
  3. RamaD

    Senior Member

    Dec 4, 2009
    319
    46
  4. bob2

    Thread Starter Member

    Jun 15, 2019
    68
    5
    Yes, this is not the best solution. MC 33078 is more suitable.
    Here is a similar circuit using LTC6244.
    There are two circuits. With and without a transformer.
    https://www.eevblog.com/forum/beginners/yet-another-esr-meter-project/
     
    Last edited: Aug 13, 2019 at 3:05 PM
  5. MrAl

    AAC Fanatic!

    Jun 17, 2014
    6,210
    1,351
    Oh yes almost the same ideas. 100kHz though.
    When i use the scope i use a lower frequency.
    There is another way we can explore too, using the phase angle of the current to detect the ESR.
     
  6. bob2

    Thread Starter Member

    Jun 15, 2019
    68
    5
    You will not feel the difference between 90 and 100 kHz.
    Found an option on the TLC272.
     
    Last edited: Aug 13, 2019 at 4:42 PM
  7. MrAl

    AAC Fanatic!

    Jun 17, 2014
    6,210
    1,351
    Yeah just mentioning the 100kHz cause that is the more typical.
    Does that opamp have the required bandwith and slew rate?
     
    RamaD likes this.
  8. MrAl

    AAC Fanatic!

    Jun 17, 2014
    6,210
    1,351
    Hello again,

    It looks like the device may work up to around 2v peak output at 100kHz. Maybe a little higher but this should still be tested for distortion at 100kHz.
     
  9. bob2

    Thread Starter Member

    Jun 15, 2019
    68
    5
    At a frequency of 60 kHz it works without problems. Most likely it will work at a higher frequency.
    As for distortions, it is difficult to say about their level in a real circuit and how they will affect the measurement results.
    https://files.elv.com/Assets/Produkte/4/477/47773/Downloads/52699_ESR1_UM.pdf
     
  10. MrAl

    AAC Fanatic!

    Jun 17, 2014
    6,210
    1,351
    Hi,

    If it is difficult then stick close to theory.
    min slew rate = 2*pi*f*A
    where A is peak output voltage,

    The pic shows a slew rate (blue) that is too slow. The red plot is the signal, and everything red above the blue line can not be reproduced properly only the red plot that is under the blue line can be reproduced properly.

    We have another issue here though. The oscillator is a square wave. That means we have to consider harmonics like the 3rd, 5th, 7th, etc. Now the lucky thing is that the amplitude for the 3rd is 1/3 of the fundamental, so we might get lucky because that would mean that if the amplifier could pass the required frequency at 1v peak output (for example) then it can pass the 3rd too because it will be 1/3 down from the 1st and if we look at the min slew rate formula we find that if the frequency goes up by 3 times but the amplitude comes down by 1/3 then it can still be passed, theoretically. What might be harder to figure out is the phase shift for the upper harmonics. The margin It would be close though, so some really good testing is advised. Probably with a scope at least.
     
    Last edited: Aug 14, 2019 at 2:16 PM
  11. bob2

    Thread Starter Member

    Jun 15, 2019
    68
    5
    It is difficult to say how harmonics will affect the final measurement result.
    I have not yet found good information on this subject, backed up by practical results.
    For me, the main thing is that the oscillator works stably. That's why I used good electronic components.
    And the output test signal is close to sinusoidal.
    It already depends on the characteristics of the transformer used.
    This is another reason why I decided to use a transformer.
    Here is what I saw.
    https://passive-components.eu/a-simple-method-for-a-capacitors-esr-measurement/
    https://meettechniek.info/passive/capacitance.html
    https://www.nutsvolts.com/questions-and-answers/capacitor-esr-tester
     
    Last edited: Aug 14, 2019 at 3:46 PM
  12. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    9,923
    2,397
    I don't think you need to worry about harmonics, given the simple circuit and that in general you want only an indication of whether a cap is likely good or bad, rather than a precise measure of ESR.
     
  13. bob2

    Thread Starter Member

    Jun 15, 2019
    68
    5
    I don’t worry about harmonics, so I never talked about them.
    And I do not care about a very accurate ESR measurement either.
    By the way, what accuracy is needed to measure ESR?
     
  14. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    9,923
    2,397
    Depends on why you are measuring it. If it were me, I'd be happy with 20-30% error if it were just a matter of deciding on whether or not to replace a cap suspected of having too high an ESR. But how will you know what accuracy your built circuit actually has?
     
  15. MrAl

    AAC Fanatic!

    Jun 17, 2014
    6,210
    1,351
    Hello,

    For some circuits it is good to know the ESR say within 20 percent plus or minus.
    I have used the second link method for years and did a complete article on it i'll see if i can find it. Ok i posted it here. Notice you can not rely on the value of the capacitor either if it is older because not only the ESR changes but so does the capacitance.
    Note how straight up and down the vertical part of each waveform is. That's because the ESR has no storage capability so the ESR responds immediately to the excitation.

    But how do you assume you are getting a sine wave?
     
    Last edited: Aug 15, 2019 at 5:34 AM
  16. bob2

    Thread Starter Member

    Jun 15, 2019
    68
    5
    Well, with the situation, good-bad capacitor, everything is clear.
    But what to do in such cases?
    1.Check the ESR of the capacitor for compliance with the datasheet.
    2. The choice of new capacitors from different manufacturers for a minimum ESR.
    3. The choice of the best capacitors from those that were in operation.
    4.Determination of ESR values, close to the maximum allowable values.
     
  17. bertus

    Administrator

    Apr 5, 2008
    19,666
    4,039
  18. MrAl

    AAC Fanatic!

    Jun 17, 2014
    6,210
    1,351
    Hi,

    I would look for something close to the spec on the data sheet.

    But i noticed the oscillator in post 44 does not look good or else it is drawn wrong.
    There is no resistor to the positive supply.
     
  19. bob2

    Thread Starter Member

    Jun 15, 2019
    68
    5
    Hi,
    Given the fact that the conditions are fairly close, I want to compare the measurement results with the LCR Pro1 Plus.
    Here is what we already have.
    ESR range-0.02-20 Ohm.
    Frequency -90kHz.
    Test voltage-75mV.
    Test signal is close to a sine wave.
     
    Last edited: Aug 15, 2019 at 6:47 AM
  20. MrAl

    AAC Fanatic!

    Jun 17, 2014
    6,210
    1,351
    That sounds good.

    You can also do this...
    1. Measure a caps ESR, call it R1
    2. Add a known resistance like 0.1 Ohms in series with the cap call it R2
    3. Measure ESR again, the ESR should now read ESR=R1+R2

    You can test with R2=0.01, 0.1, 1, and 10 Ohms.
     
Loading...