EEVblog 1473 - How Your LCR Meter Works

The measurement doesn't have to be in phase with the source signal, it's only important that the two measurements be 90 degrees apart in phase at exactly the source frequency. This is what the fast Fourier transform does and outputs two vectors 90 degrees apart. Although the phase of the two vectors is random with respect to the source phase, the calculations are exactly the same because it is the ratio of the real and imaginary amplitudes and the relative phase between them that is important.
This simplifies the hardware because there is no need to synchronize with the source as long as the frequency is precise. The A/D sampling has to be frequency precise with regard to the source and then an FFT can use the sampled data to extract the vectors and then calculate component values.
A primitive LCR meter can be made using an Arduino to generate a square wave source with the A/D triggered at four times the square wave frequency to give the proper phasing with half the samples the sine value and the other half the cosine. This is the hardware equivalent of doing an FFT. The first sample is sine +, the second cosine +, the third sine - and the fourth cosine -. The - samples are multiplied by -1 and added to the + samples and averaged to get the amplitude of the vector. The cosine value is the real part of the vector and the sine is the imaginary part.
For higher precision the source voltage should also be measured using the same technique. If this is done, the source voltage value is not important as long as it is stable. The LCR measurement depends only on the ratio of the DUT impedance to the reference resistor impedance and if the same A/D reference is used for all measurements a fairly accurate measurement can be made.
A PC sound card can also be used as the basis of an LCR meter because it can easily source and measure and only requires a simple op-amp input buffering and guarding stage.

Excellent. This was needed for persons whom don't understand impedance.

Great video Dave, very good presentation. I took a lot of information from it. Thank you.

I think you explained this in another video 12 years ago

Sorry,.... Hey how do I convert these Australian-rubbish-units into something I don't need Roo-poo smeared across my brow and a f agin my mouth to calculate correctly?

Hi Dave, thank you for another very informative video. There is something in your notation that seems odd, though. From the equivalent circuit and the complex plane graph, you seem to imply that Xs is a complex number. However, this does not seem to work because Xs is calculated in your formula using Vq, Vp, Iq, and Ip which are all real numbers. I believe, what you were trying to say is that Xdut = Rs + jXs, with Xdut complex and Rs and Xs real numbers. If this is true, it makes sense to say that Xs can be either positive or negative. If this is correct, the reactive part of the equivalent circuit should be jXs instead of Xs, and the red arrow in the complex plane should be Rdut instead of Xs. Xs should be the projection of Rdut on the imaginary axis. Does this make sense?

Dave should design a cheap diy lcr meter. There aren't really any good ones.

Why when I pop in 100 nF caps (ceramic I think), does the capacitance slowly drift down (ESR drifts down and Q drifts up)? My Keithley DMM6500 does the same thing with them.

Thanks Dave! Now you are our teacher!

The face of Dave is scary than the concept themselves....

Dave, it doesn't seem like you're really a RF guy but your phase measurement explanation is a really good "intro to Smith charts" ... minus the chart itself. :)

Back in the olden days when I was in an electronics class at a community college, we were all given a 5H choke, large iron core inductor to measure. We all got some silly number I don't remember but it was very small, like 100uH. Nobody cared, even the teacher, as he didn't know why. I followed through (for weeks) and found our test current was so low, it barely moved the domains and put the BH curve where it almost followed the X axis line. If we had tested with a higher current, the line would have separated from the X axis and we would have got a higher inductance answer. Nobody cared. Sad!

Nice work!! This is pretty cool stuff!

Thanks for this, really!

It left me thinking thou. What happens when you force the meter into c or l or r mode? My first idea during the video was it’s just changing what displays where. But your last sentence left me thinking there’s more than that. Is it?

Oh hoy mate you're kinda speaking too fast for non native English speakers to understand. Could ya speak a littler slower? You already use the British accent and I am not even an American but trying to understand you by listening takes kinda a lot of voltage in my brain which kinda fries it. Just try to speak slower a little. We're not native British speakers as you are dude... You kinda sound as if you're high of meth or speed mate... We're not as speedy as you're mate just chill... Try watching AHOC (buildzoid) and try adjusting to his speed, he is a Brit too but I can understand him. You really sound way too fast... I am sorry if you're an Aussie btw but I have an Australian friend who speaks so much more understandable compared to you as well. It's not that I don't understand what you're talking about. If I listen to you twice or three times I do, but I can't literally determine what you're saying man because you're too damn fast...

now, that's a juicy piece of content

When I learnt trig, |Xs|/Rs = cos-1 (ȹ) not Tan(ȹ)

I'll try and design one myself.. that'd be a very interesting proyect. Nice video..
Greetings from Argentina.

👍

It is a great video. But can you explain the equations of Rs and Xs. Rs = (Vp x Ip + Vq x Iq)/ (Ip2+Iq2) and Xs = (Vq x Ip + Vp x Iq) / (Ip2 + Iq2)

PLZ just to remind you of the lcr tweezer toutrial you pomised @EEVBLOG

Thanks for explaining q and how my old lcr meter works! I'm not sure that all stuck, but the overview really helps me a lot! 👍

If Linus had a Father.

Terimakasih atas penjelasannya dan sub titlenya.god job

got it, very clear explanation

👍👍

AD5933 ?

Wonderful video! One thing that confuses me is the graph of the impedance in the bottom center of your board. Impedance is Z = R + jX. So shouldn't Xs be on the imaginary/vertical axis, Rs be on the real/horizontal axis, and Zs be placed where you have Xs marked? For example, if your impedance was Zs = Rs + jXs = 2 + j3, the absolute value (marked with the | | ) of Xs is 3. But the way you have it drawn with Xs being where Zs should be, the absolutely value of Xs would be ~3.6

If the load is purely inductive (so Xout = jwL), and w and L are such that wL = 1. Then we have Xout = j1. Then we know that the current lags the voltage in an inductor by 90 degrees. So if the voltage placed on the inductor is also 1, then Vp = 0 and Vq = 1. The current would be Ip = -1 (since the current lags the voltage in an indcutor) and Iq = 0.

Then if we use your equation for Rs, we correctly get Rs = 0. But if we use your equation for Xs, we should get +1, but the answer I get is -1. Xs = [ (1)x(-1) + (0)x(0) ] / [ (-1)^2 + (0)^2 ] = -1 / 1 = -1.

If I do the same thing but with a capacitor instead of an inductor, I would except to get Xs = -1, but instead I get Xs = +1.

Am I missing something or is there a sign error somewhere?

So if we are measuring a R and C for a device then we have to connect the device in the parallel mode right?

"American rubbish" - them's fighting words! BTW - Excellent video!

nice

Nice show of calculators at the bottom!

I have a confusion over here: wt = 0 degree and wt = 90 degree are two different instances. How are we getting Vp and Vq by extracting the values at two different instances... The vector V in your diagram is the voltage phasor at a particular instant. How would you get info about V's component in Vo axis at a particular instant if we measured at two different times?

Another question: While measuring the quality factor, shouldn't the Xs be measured at the coil's resonant frequency rather than measuring at any random frequency?

Anybody:. I assume there is a single chip available for purchase that does ALL of this, including the phase angle correction, the imaginary numbers, etc?
If I can buy a kit for ten bucks that can be built into a functional LCR tester, I have to assume it is built around a single dedicated purpose chip?

Merci, one of your best videos I think.

Can you use an LCR meter to directly determine the impedance of an audio transformer both on the secondary side (4, 8, 16 ohm) and in the primary side (for matching to a tube for example). When you acquire an unlabeled transformer it would be very helpful to know this. I assume this would be done at 1Khz but I'm not 100% sure. I measure the outer wires on my secondary to be 3.5mH, @1Khz this translates to 22 ohm. From the center tap on the secondary it measures .9 mH which translates to 5.7 ohms. Thanks

amazing sir. Hoe to measure the R L in series and C in parallel that is the one end of C at R and another is grounded. It will form the RLC lumped model. How to measure RLC of the RLC lumped model of interconnect using LCR meter? Please guide sir

Did they ever let you Australians out of civid camps ?

Corkery Crest

Alia View

40466 Wiza Path

O'Hara Islands

Omeeega or Omaaaga

With AI being what it is today and with so many lectures done, doesn’t it seem like a good time to have AI ghostwrite a book for you?

great course density and interactive it felt like 2 minute watching the 20

A little correction.
At the bottom right on the whiteboard, you have ...
D = |Rs| / Xs
... it should be ...
D = Rs / |Xs|
... as Xs is the value that can be negative.