How do you find a coils natural frequency?

[blueroomelectronics]

I'm looking for a method of testing very large coils and trying to figure out their natural frequency and quality. Pumping a square wave through them is the method I've been using with a scope and the method on this page.
http://www.vlf.it/matteobruna/ULF_Induction-Coil.htm

 

[Nigel Goodwin]

'Ring' it with a single pulse, and measure the frequency you get as it decays.

If you're going to use a squarewave it needs to be considerably lower than you're expecting.

If it's any help, we used to do this YEARS ago with old scopes, by connecting a pulse from the timebase flyback to the coil - common practice for checking LOPTX's for shorted turns.

 

[Willbe]

trying to figure out their . . .quality.

If takes the coil ringing 10 cycles to go down to 1/e of the original amplitude the Q factor is 10.

 

[blueroomelectronics]

Ahh the magic of inductors. So the more a coil oscillates the higher the Q?

 

[Speakerguy]

Yep, ringing = high Q.

 

[Hero999]

If you just want to measure the resonant frequency an LC oscillator would do.

 

[Willbe]

D'you ever see or hear a bell made of lead?

But a crankshaft can ring and have resonances at several freqs. and that's why engines have a harmonic balancer (a rotational mechanical system is analogous to an RLC tuned circuit and a balancer is an "R."). This unlikely gadget is a small flywheel coupled to the crankshaft end through a viscous rubber plate.

A hardware store customer who spoke Espanol wanted to make windchimes from the tubing we sold. Through a translator I tried to tell her that it would sound terrible.
She bought the tubes anyway.
I suppose she could have gotten the pitch right but the Q was way too low.

 

[Speakerguy]

We've got some wind chimes made out of steel pipe that sound nice. Was she buying PVC or what kind of pipe?

 

[Space Varmint]

I agree with Hero999. Just make an oscillator where you can plug in the coils interchangeably. Since you know the capacitor's value all you have to do calculate coil's inductance. Well that will get you the inductance anyway. As far as it's natural freq it would depend on things like spacing between the turns, wire diameter and so forth.

I would think that using a square would be a very poor method because a square is composed of many odd order harmonics so which harmonic would the coil be resonating to? You will get a higher Q with a high permeability core. Air core type coils are very low Q and will have a wide range of frequencies or bandwidth. I would think it would be much more difficult to identify. Especially with a digital readout type of freq counter or whatever device you are using.

 

[kchriste]

You could also build or buy a "grid dip meter".

 

[Willbe]

We've got some wind chimes made out of steel pipe that sound nice. Was she buying PVC or what kind of pipe?

I think brass, alum, maybe SS. Are you sure your chimes are not some alloy of steel?

 

[Roff]

What is the range of inductance?
What measurement tools do you have? An oscilloscope? A network analyzer? A Q meter? A...

 

[jpanhalt]

Bill,

Check out **broken link removed**'s site.

John

 

[Hero999]

I would think that using a square would be a very poor method because a square is composed of many odd order harmonics so which harmonic would the coil be resonating to? You will get a higher Q with a high permeability core. Air core type coils are very low Q and will have a wide range of frequencies or bandwidth. I would think it would be much more difficult to identify. Especially with a digital readout type of freq counter or whatever device you are using.

Actually a squarewave does work as long as its frequency is much lower than the coil's self resonant frequency.

All you need is a step function to ring the coil. A square wave is normally used because it generates a repetitive waveform which can be viewed on a CRT scope.

You could use a debounced push button to ring the the coil once if you have a digital storage scope.

 

[Space Varmint]

Actually a squarewave does work as long as its frequency is much lower than the coil's self resonant frequency.

All you need is a step function to ring the coil. A square wave is normally used because it generates a repetitive waveform which can be viewed on a CRT scope.

You could use a debounced push button to ring the the coil once if you have a digital storage scope.

Huh? I didn't know that. It would seem like a pure sign-wave note would be best. I'm sure somebody did allot of work to figure it out.

 

[Nigel Goodwin]

Huh? I didn't know that. It would seem like a pure sign-wave note would be best. I'm sure somebody did allot of work to figure it out.

It's simple enough - to ring a bell you use a single short strike, the bell then rings at it's resonant frequency. A coil is exactly the same, you feed it a single pulse, and it rings at it's resonant frequency.

Like I mentioned earlier, a common technique years ago was to use a flyback pulse from a scope to ring a coil.

Applying a sinewave wouldn't ring the coil at all.

 

[Space Varmint]

It's simple enough - to ring a bell you use a single short strike, the bell then rings at it's resonant frequency. A coil is exactly the same, you feed it a single pulse, and it rings at it's resonant frequency.

Like I mentioned earlier, a common technique years ago was to use a flyback pulse from a scope to ring a coil.

Applying a sinewave wouldn't ring the coil at all.

Yeah but I figured the harmonics would mess up the reading.

 

[MrAl]

Hi there,

Im pretty sure a square wave pulse works too, but you can't just connect
the output of a signal generator to the coil and expect to get the right
wave in most cases because the signal generator output impedance
characteristic will affect the overall waveform seen during and after
the pulse. A diode also has to be incorporated, between the signal
generator and the coil, so that when the sig gen wave goes low
the coil gets totally disconnected from the sig gen output.
Of course a relay would work too.
Probably have to use some damping too, like a 100k resistor or
something like that to prevent extremely high voltage spikes
from damaging the test equipment.
Just some ideas.

Of course another way is to use a sweeping sine wave
and look at the current response.

 

[Space Varmint]

Hi there,

Im pretty sure a square wave pulse works too, but you can't just connect
the output of a signal generator to the coil and expect to get the right
wave in most cases because the signal generator output impedance
characteristic will affect the overall waveform seen during and after
the pulse. A diode also has to be incorporated, between the signal
generator and the coil, so that when the sig gen wave goes low
the coil gets totally disconnected from the sig gen output.
Of course a relay would work too.
Probably have to use some damping too, like a 100k resistor or
something like that to prevent extremely high voltage spikes
from damaging the test equipment.
Just some ideas.

Of course another way is to use a sweeping sine wave
and look at the current response.

There ya go. That's what I'm talking about.

 

[MrAl]

There ya go. That's what I'm talking about.

Hi again,


Ok

I forgot to mention though that if the coil voltage goes negative
the diode conducts (with that technique). A few diodes connected
in series helps here, with the sig gen pulse amplitude chosen
accordingly.
Another thing to do is to do a simulation of this in a circuit simulator
program. Use an idea inductor in parallel with an ideal capacitor
(be sure to put some resistance in series with the inductor too)
and try the pulse technique and then the sweeping sine wave
technique. Measure the voltage across the coil and the current
through the coil.
This will help with knowing what to look for as well
as what to avoid.