I would like to build a simple LC oscillator (Colpitts, maybe) that uses a coil that I can wind using magnet wire. I don't have any winding equipment, so my preference would be for a coil with an air core wound by hand. I would like the oscillator to operate in the human hearing range, say about 2 kHz, and be able to drive a speaker. I have some experience with the LM386, and can build an amplifier if someone will supply a schematic for the oscillator and the instructions for winding the coil. If possible, I would prefer the coil to be wound on an air core of about 1" diameter.

Thanks for any help.

 

Start by picking values for L and C. Maybe pick your capacitor from what you already have or can get, then from that, calculate the L you need to resonate at that frequency.

Then go online and find a coil-winding formula that will tell you how many turns you will need for a given length and diameter to get that value of L.

Don't worry about being too exact with all the dimensions because your capacitance value itself may not be given to better than 10% or so.

But get an estimate of how many turns you will need, and the length of wire you'll need. That would be a good start.

There are lots of ways to wind a coil, and some are pretty simple. I've wound coils by clamping a form in an electric screwdriver and feeding the wire onto a plastic form by hand.

If your oscillator needs a tapped coil, it's nice to wind your own because you can build in any tap you want, or even several if you want some tuning ability.

That much should get you started.

 

Unless I've screwed up badly: coil form 1 in dia, 1 in long, with 1.5 in dia end plates, filled with # 26 enameled wire, about 841 turns to give 14.5 mH tuned with .5 μF to about 2 kHz. Try a 6 μF cap in series with .5 for feed back tap. Schematic escaps me for now.

 

Unless I've screwed up badly: coil form 1 in dia, 1 in long, with 1.5 in dia end plates, filled with # 26 enameled wire, about 841 turns to give 14.5 mH tuned with .5 μF to about 2 kHz. Try a 6 μF cap in series with .5 for feed back tap. Schematic escaps me for now.

That sounds about right, You wont be able to make a 2 Khz oscillator using an air core coil. You cant get enough inductance. Air core coils are only practicle at VHF frequencys.

 

This version of a Colpitts Osc was derived from one by Peter G Sulzer W3HFW, around 1953, using CK722 & mercury batteries. L was .3H torid. I used the origional version for a 1 kcps frequency standard- around 1960.
Adjust R1 for minimum clipping. +V maybe 5V. Output into a high impedance.

 

Thanks for the help. I think I will build the oscillator, but try to buy an inductor and get the oscillator working. Then, I'll try winding my own.

 

Might try for a more standard size, like 10 mH or 20 mH.
Found a 15mH, # G 18739 @ Electronic Goldmine.

 

LC oscillators aren't practical for frequencies much below 50KHz.

 

'Looked in junk box & found a nice 20 mH, 2 Ω, torid inductor, so hooked it up on solderless bread board. Changer some values to what I had.
C 1, C 3 = 5 μF, electro.
C 2 = .47 μ F, film
L = 20 mH
R 1 = 2.5 m pot, set @ 118 k
R 2 = 68k, non critical
Output 2 k Hz, adjusted to 5 V PP via R 1
It would operate at low out without C 1, @ reduced distortion.
Power = 5V , 4 AA N-Mh batteries.
Sorry to hear that LC osc are not pratical at low Hz, guess I've been wrong for last 53 years. I have a 1 H choke- should we try it??

 

Sorry to hear that LC osc are not pratical at low Hz, guess I've been wrong for last 53 years. I have a 1 H choke- should we try it??

Thanks, Bernard.

I haven't done anything on this project except put that .15mH coil at EG in my basket.

I am replacing my old desktop computer with a new machine that I assembled, and have been busy moving files from the old to the new. I still have a lot of programs to install, but the new machine is now functional, and the old one is just on standby to see if there is anything else I need from it.

However, I still plan to build this oscillator when I get back from a trip. This is a new tack on my copper vs zinc penny discriminator. The theory is that placing a penny near the coil will change the frequency, and the change will be different for the cu vs. zn. I wanted the output to be in the audible range just for easy confirmation of the frequency change, if any. I guess if that is successful, I will then try to use a μC to read the frequency and control some mechanical means to sort the pennies. Obviously, I haven't figured that part out yet.

What would the 1H coil do...just change the frequency?

Thanks again for your valued input.

 

Here are some typical large inductors used for speaker crossovers that may work for you.

 

Thanks, Bernard.

I haven't done anything on this project except put that .15mH coil at EG in my basket.

I hope that you really are going to order a 15mH inductor. You won't get 2kHz oscillation with .15mH.

 

I hope that you really are going to order a 15mH inductor. You won't get 2kHz oscillation with .15mH.

Yes, it's 15mH. pesky fly left a speck on my post.

 

Could I interest you in 2 oscillators? The method is that a small change in one oscillator causes a beat frequency in the audio range. With this method, you can use higher frequencies, easier to build, and the resulting circuit will be more sensitive to small changes. 4 instance, a couple of 40KHz oscillators. If the penny pulls one oscillator 1% off its base frequency, the result is 400Hz difference. This could turn out so sensitive that you could hear the difference in the pennies with only your ears, let alone electronic filters.

 

Could I interest you in 2 oscillators? The method is that a small change in one oscillator causes a beat frequency in the audio range. With this method, you can use higher frequencies, easier to build, and the resulting circuit will be more sensitive to small changes. 4 instance, a couple of 40KHz oscillators. If the penny pulls one oscillator 1% off its base frequency, the result is 400Hz difference. This could turn out so sensitive that you could hear the difference in the pennies with only your ears, let alone electronic filters.

A BFO metal detector.

 

Could I interest you in 2 oscillators? The method is that a small change in one oscillator causes a beat frequency in the audio range. With this method, you can use higher frequencies, easier to build, and the resulting circuit will be more sensitive to small changes. 4 instance, a couple of 40KHz oscillators. If the penny pulls one oscillator 1% off its base frequency, the result is 400Hz difference. This could turn out so sensitive that you could hear the difference in the pennies with only your ears, let alone electronic filters.

I am interested, but I don't fully understand it. Would there be a single audio circuit, somehow connected to only amplify the difference between the two oscillators? Ron_H had provided some information about metal detector circuits in another thread, but after some investigation, I went back to the eddy current method. As you guys know, I am just a tinkerer, and am often in over my head.

 

Simply adding the output of the two oscillators will give you the difference between their frequencies, so if you tune them properly close you will be able to hear the change in the difference of frequencies. Google gives a lot of different schematics.

 

Try this LC oscillator

And you can use any type PNP BJT's.

 

Try this LC oscillator

And you can use any type NPN BJT's.

The schematic shows PNP, but you say NPN. Please clarify. Thanks.

 

Yes I meant PNP BJT's. Sorry for stupid mistake.