How do I measure inductance/capacitance without RLC meter?

[technix]

Here is what I have:
* An old GW-Instek function generator
* Rigol DS1054Z oscilloscope (can measure frequency and true RMS voltage)
* Uni-T UT61E true-RMS DMM
* A reference resistor, measured by the said DMM.

Questions:
1. Can I measure an inductor or a capacitor using the circuit above?
2. If so, how should I calculate the capacitance or inductance from the readings on the oscilloscope and the DMM?
3. Can I get away without the DMM for the measurement, after calibrating the resistor, using the four input channels on the DMM?

[Ian.M]

Assuming a series circuit of  R - {L or C}, driven by the signal generator with the scope probing the voltages across the series circuit and across the capacitor or inductor,  simply adjust the signal generator frequency for 1/sqrt(2) of the input amplitude across the capacitor or inductor  (which is also 45 deg phase shift).  At that point the impedance of the capacitor or inductor is equal to that of the resistor so you can simply calculate the capacitance or inductance from the resistance and the frequency.   It will not work well if the capacitor or inductor have excessive losses. 

[iMo]

You may also use the pulse method. Below the RC shapes for C1=1nF..10nF.
It works for RL the same way (see below for L1=100uH..1mH).
50ohm is the internal impedance of the Generator.

[JS]

Depending on what you try to archive, you can get really precise measurements taking data with the scope and prpcessing in the computer, that way you can model as many parasitic components in the DUT as you like and get the nominal value or the apparent value at any frequency you like.

Also there is the rough way posted in the first reply, or there are other methods useful to do a similar thing using what you have, maybe a few more passives and a calculator. That's useful to get measurements of very extreme values, as it's quite hard to do so with the conventional methods, because frequency would be too extreme, impedance too low or too high, or noise levels become a problem.

JS

[Conrad Hoffman]

There's also the old method of resonating an inductor with a cap or vice versa. You need a known cap, but most junk boxes have some 1% mica caps floating around. Resonant frequency gives you the value. It's good to measure some stable parts and keep them around as lab standards.

[JS]

There's also the old method of resonating an inductor with a cap or vice versa. You need a known cap, but most junk boxes have some 1% mica caps floating around. Resonant frequency gives you the value. It's good to measure some stable parts and keep them around as lab standards.

That's fine for some cases but with inductors, specially with a core different than air you want to know the inductamce or complex impedance at a given frequency, with a sig gen and a scope you can archive that with no problems. Single points on the scope and a calculator would do, for a plot of mag and phase getting data to the pc is the way to go. With modern scopes that's quite easy connecting them to the network and capureing data directly from the PC, I should wtite a code wich could be directly used with the ds1054z I have as I do need it but I havent got around doing so yet. I'm missing a sig gen, so some noise source should do for now, ideally a pc controlled sig gen and a script to make both do it's thing. That's more involved tough and getting into the omnicron bode territory... That's the tool for the job up to 50MHz.

JS

[bd139]

https://youtu.be/74fz9iwZ_sM

[Zero999]

Assuming a series circuit of  R - {L or C}, driven by the signal generator with the scope probing the voltages across the series circuit and across the capacitor or inductor,  simply adjust the signal generator frequency for 1/sqrt(2) of the input amplitude across the capacitor or inductor  (which is also 45 deg phase shift).  At that point the impedance of the capacitor or inductor is equal to that of the resistor so you can simply calculate the capacitance or inductance from the resistance and the frequency.   It will not work well if the capacitor or inductor have excessive losses.

Yes, I've done that before and it works well. The output impedance of most signal generators is 50R, which is convenient for capacitors and most inductors, the exception being some low power transformers, which might have a much higher series resistance, although it's trivial to just add another resistor.

[basinstreetdesign]

Here is what I have:
* An old GW-Instek function generator
* Rigol DS1054Z oscilloscope (can measure frequency and true RMS voltage)
* Uni-T UT61E true-RMS DMM
* A reference resistor, measured by the said DMM.

Questions:
1. Can I measure an inductor or a capacitor using the circuit above?
2. If so, how should I calculate the capacitance or inductance from the readings on the oscilloscope and the DMM?
3. Can I get away without the DMM for the measurement, after calibrating the resistor, using the four input channels on the DMM?

Depends what you want.  If you do the RC/RL time constant method you can get only ok results.  Even with a 1% resistor and using that scope you probably wont get any closer than 5% accuracy.  The scope itself guarantees only +/-3% vertical accuracy even though its horizontal accuracy is supposed to be more like +/-25 ppm.  To do better you would need to add a precision voltage standard (+/-.05% maybe?) to vertically calibrate the scope with.

Doing the tank resonance method will work, too, but as others have said you need at least one of a reference cap or inductor to construct it with.

Don't rely on the function generator to tell you what the frequency is with any precision, rely on the scope for that.

If you want better precision then you need a much better equipment package.

[Richard Crowley]

You seem to be asking for a "classic/historic" method.

But for those of us seeking more modern, practical solutions, how does this compare to those $10 LCR/ESR/Transistor/Diode testers? Would this yield similar (accuracy) results?  Certainly much less fiddling around with a lash-up of several pieces of gear for a simple L or C test it seems to me.

Ref: https://youtu.be/7Br3L1B80ow

[bd139]

Those modern do everything testers are garbage. I’ve had four so far and all of them read way off and inconsistent measuring a 0.5% 100pF cornell silver mica reference which has been tested on a proper calibrated HP LCR meter. Unstable oscillators, terrible stray capacitance and generally shitty design.

The resonance method works well for measuring inductors. I do this even to this day using the above reference cap.

There is also the frequency method which is where you insert the L or C in circuit in a colpitts oscillator, then switch a reference cap in and then record the difference in frequency. You can derive L or C with a bit of algebra from Fout then with only one reference component in circuit, usually a capacitor as they are available to better off the shelf tolerances than inductors. This is the basis for some of the Chinese ones which are poorly knocked up clones of the old AADE LC meters with low tolerance components etc.

[JS]

I do have one and is useful but I wouldn't trust the reading for anything but rough guess. Identify a part, like those old packages you can't tell what it is just by looking at it, check if a cap is still a cap, etc. And it also generates pwm signals to test a few things around. Usrful yes, metrology garbage also yes. It might get lucky once in a while or even more often than not but trusting it's reading would come to bite you eventually. A friend was trying to measure a guitar mic coil and it just reads resistance, a frikking coil of wire around some iron and it reads resistance!  The error is also you usually can't force the thing to read a choosen model, so it reads what it think it is.

Sig gen scope gets you a long way to characterize an impedance but not so much for a precise value reading if you try to match components, for that you should use a bridge, to get a good value a bridge with a reference component. Or if you are doing it a lot get a proper LCR meter but the coice of freq and amplitude is usually limited, with sig gen and external components you control every parameter of the test and can get more meaningful numbers for a given application, even if less precise than a proper meter.

JS

[David Hess]

Another way to do it if you have a reference capacitor or inductor is to duplicate the method used by an impedance bridge with the oscilloscope serving as the AC phase sensitive detector.  With some extra effort, this can also measure the quality factor or dissipation of the capacitance or inductance returning the ESR.

https://en.wikipedia.org/wiki/Bridge_circuit

Not that amplitude accuracy of the oscilloscope is not required for this but time accuracy is which is not a problem for an oscilloscope.