Find Large Inductance Without Meter

sonnect it in series with a known resistor and a known (or measurable) AC source. measure voltages, frequency etc.

I would use a time-domain approach. Apply a pulsed voltage and measure di/dt.

Find or make the largest inductor your meter will measure. Put your unknown and known inductances in parallel and measure the value. Here's a calculator to give you an idea where you stand.

Mikek

PS. This kinda depends on how high your meter will measure, and how big your unknown is.

Here's a wacky idea: what about using a "negative power inductor"? Make a grounded negative inductance circuit using an op amp and an audio power amp, with appropriate local and global feedback (see figure 9 here:

Connect the unknown inductance in series, sweep the function gen and watch for the current spike.

Connect a random capacitor across it and measure the resonant frequency.

Or connect the scope and fun gen both across it. Plot ac amplitude vs frequency. Thet will tell you a lot more than inductance.

Do you really need to resonate it? Does it need a core?

It's a lot of kHz too!

On Nov 24, 2016, Kevin Foster wrote (in article ):

I assume that you mean a DC resistance of 20 Kohms, not 20 MHz.

This kind of sensing coil is a very low Q inductor, where the DC resistance well exceeds the inductive reactance in the operating band.

Most $200 and lower LCR meters will yield nonsense when attempting to measure such an inductor, which can easily be many tens of Henries. What can work are the Extech 380193 and the DerEE DE-5000 LCR meters, which are widely used to measure electric guitar pickups. Pickups are typically 2 H to 10 H, with many thousands of ohms of series resistance, plus much added series AC resistance due to eddy-current loading.

The problem is that the inductance of cored inductors varies with frequency, and even a 100 Hz test signal may cause errors. If the mu metal core is properly laminated, this may not cause too much trouble.

The most general way to make the needed measurement is to cobble a Maxwell-Wein Impedance Bridge together. This will yield answers to the accuracy with which you can measure the resistors and capacitors. The null equation does not involve the test frequency, but the answer will vary with frequency to the degree that the inductor varies with frequency. A low-distortion sine wave test signal is required, or it will be hard to find the null.

The self-resonant frequency of the coil may well be in the audio range - such large coils have a lot of self-capacitance. The easy way to tell is to connect the coil to a high-impedance scope input, drive the coil with a nearby coil driven by a variable-frequency oscillator, and find the amplitude peak. Add a few capacitors in parallel with the coil and find the peaks. Solve for inductance and self-capacitance. This method is documented in Terman?s Radio Engineer?s Handbook.

If one can do X-Y plots on the scope, the more precise approach is to find the zero-phase frequencies (phase nulls), not the amplitude peaks (which are not in the same place with low-Q components).

Win Hill?s suggestion (apply voltage, measure the rate of rise of current)is the best approach to get in the general area.

For very large inductances, it takes forever to balance an AC bridge. What people did in the old days was to use a pushbutton to apply a voltage surge to the bridge and observe the jump in a meter connected to the bridge output

- adjust bridge for minimum jump on surge.

Terman?s procedure fits right in to this.

Joe Gwinn

Yep, that's probably the best bet. You'll also want to measure DC resistance of the coil, and connect your function generator through a resistance of higher value. Those resistances go into the resonant-frequency formula (instead of R, use the parallel sum of the resistances, Reffective :== (R_inductor * R_generator)/(R_inductor + R_generator).

The DMM is presumably a meter, so if you want to measure without a meter, it cannot be used.

The easiest way is probably to find a known capacitor value, connect it across the coil and measure the resonant frequency with your function generator and oscilloscope. However, if that's too easy, there are some alternative methods.

1. Reverse engineering. Unwind the rather large number of turns of copper wire, measure the length, calculate the number of turns, and use an online calculator for determining the inductance. Rewind the original coil when done measuring. If you know the wire gauge accurately, you can calculate the wire length, and therefore the number of turns, from the resistance.
2. Balanced bridge. Build a Wheatstone bridge using your coil as one leg, and a gyrator as the balancing leg. The gyrator can simulate an inductance. Apply any frequency with your function generator, and check for balance with a dual trace oscilloscope using two scope probes and the A-B feature. Once a balance is achieved, you can use the component values of the gyrator to calculate the inductance.
3. Phase shift. Place a resistor in series with the coil. Apply a sine wave at any frequency with your function generator. Place one scope probe across the resistor and the other across the inductor. If the function generator output is grounded, then you'll either need a 4 channel oscilloscope, or two instrumentation amplifiers. A transformer might work. Measure the phase angle between the two channels and calculate the inductance.
4. Low pass filter: Function generator connected to coil to resistor to ground. Place scope across the resistor. Sweep the frequency range until you get a response that resembles a low pass filter. Measure the -3dB (0.707*V) corner frequency. At that point, XL = R. Subtract the coil resistance from the value of R as they both appear to be in series. L(Henry) = R(ohms) / (2*Pi*freq)
5. Time constant. Place a resistor in series with the rather large coil. Place an oscilloscope across the resistor and time how long it takes for the current to hit zero. T(sec) = L(Henrys)/R(ohms) Subtract the coil resistance from the value of R as they both appear to be in series. Also, don't put the scope across the inductor. With what I would guess(tm) might be a rather large inductor, you can easily generate a rather high back EMF voltage, which might arc over your scope probe or vaporize your scope input amplifier. Don't ask how I know this.
6. Transformer. If the coil is wound on an open bobbin, apply a few turns of wires and build a transformer. From the voltage ratio, you can calculate the number of turns. Using the physical dimensions and AL (inductance per turn) core properties, you can calculate the inductance.
7. TDR. Time domain reflectometer. If you own a decent scope, you should consider building a TDR. Insert your coil between the TDR and a suitable load to form a rather crude delay line. A TDR is usually used to measure coax cable inductance and capacitance, but should work for lumped elements. L = Td * Zo
8. Ringing frequency. Use your function generator in pulse mode.

Thanks Jeff. I will try the "easy way" first and then proceed from there.

I was wondering about another approach. The coil is bound to pick up a lot of 60Hz. How about I find a cap that maximizes the amplitude as viewed on a scope and work out the inductance from that and the coil resistance?

Keven Foster

Thanks for your detailed explanation.

I have downloaded the Terman book from here:

Kevin Foster

It

read.

easure

ency,

What the heck is an 'earth sensing magnetic coil'? Are you talking about a seismometer? Seismometer resonance is NOT electrical alone, there's moving parts.

Right. If the coil has a DCR of 20K ohms, the generator impedance needs to be higher. 20K won't load a 50 ohm gen enough to notice.

The Q of that inductor may be really really low around 10 Hz. Below 1 likely. It may not be worth resonating.

The easy way often becomes an obstacle course of unexpected surprises. John Larkin mentioned that the Q of the coil might be less than 1 resulting in a very broad resonant peak, which you're likely to miss. If your function generator has a frequency sweep mode, it might be easier to see. Take your best shot at picking a series resonant capacitor. Put the coil and capacitor between the sweep generator output and the oscilloscope probe. Connect the external horizontal sweep of the scope to the sawtooth sweep output on the function generator. Play with the controls until you see a peak.

You have a function generator which will allow you use a much higher frequency for resonating the inductor. Then doing the series (or parallel) resonant trick, it doesn't need to be done at 10Hz, 60Hz, or other low frequency. It's easier at higher frequencies. It's also a bad idea to do any such testing at 60Hz, where stray pickup will cause nothing but trouble. In any case, you'll need some kind of capacitor decade box or variable capacitor, which will difficult for large values. Also, a starting value of the approximate inductance, so that you don't get bogged down in trial an error exercises.

Now, some questions.

1. What do you mean by "resistance of 20,000KHz"? Is it ohms or milliohms as it's certainly not 20MHz.
2. If 20K ohms, your coil must have an enormous number of turns, look more like an open circuit, or you blew the measurement. If it was really fine gauge wire such as #40AWG, with a resistance of about 1 ohm per foot, the winding would be 20,000 ft long. Post a photo. I want to see this thing.
3. What is an "earth sensing magnetic coil"? Are you building a seismometer? Are you trying to pickup low frequency emissions allegedly used to predict impending earthquakes? Are you listening for "whistlers"? ELF receiver (below 22KHz) for monitoring the magnetosphere? ELF emissions from TLE (transient luminous events) such as sprites, jets, halos, elves?

Yes. In fact, you can sample the current using a scope, and pulse generator and measure the time where the current reaches the 63% point then use t = L/R. Knowing, of course, the coil plus sample resistors' values.

I hate to tell you this, but your title says "Without Meter" and a DMM is a meter, and by extension, so is a scope.

One could use an adjustable square wave generator in series with the inductor and a resistor, using the scope to determine L/R time constant. The advantage is that one cold determine if the inductor is being saturated by the signal across it.

An alternate is to use a variable sine wave and "tune" for equal voltage drops across the resistor and inductor. One might use either a scope or a DMM (provided the frequency is inside the BW of the DVM).

One might be able to L-C resonate for determine inductance, but that may not be possible if the self-resistance of the inductor is father large. In that case the second method will also be not useful.

I leave it to the student to appropriately modify and enhance the first method to accomidate a large internal R inductor.

I thought that inductors were made of wire that were not perfect conductors...

Hell, the inductor may have so much internal/self resistance that it CANNOT be resonated. Worse,have a "Q" well under one.