Sine wave oscillator tech, and prototype circuit / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Sine wave oscillator tech, and prototype circuit

Move Thread

LAN_403

Inducktion

Fri Feb 11 2011, 02:59AM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

Hello all, I created a Sine wave oscillator circuit on Falstad's famous circuit simulator. I have all the necessary components in the circuit to keep oscillations going, and works for a wide variety of frequencies.

I don't know if it'd work in real life, though, I'm doubting it would due to nearly all circuits created on it don't work for some reason.

Anyways, I would just like to know your opinion on the circuit, and if you think it'd work, and all that good stuff.

The transistor can be any NPN one, I've done numbers from 8 gain to 1000 gain, all work. The transformer is a center tapped transformer, changing the inductance changes the frequency. Now, the most important components in keeping this working are the resonant capacitor across the transformer, and the zener diode. The resonant cap can be used for a wide variety of frequencies, I've gone from 70 hz to over 50 khz, just by changing that value around. The zener diode is what keeps the oscillations going, the value needs to be generally around 3-6 volts below the input voltage. If this worked, and you used it for an inverter, 12 volts in would require a 8-10 volt zener diode there. The potentiometer on the lower part of the circuit needs to be in a specific range for oscillations to occur, a 20 ohm pot works best, no idea why. Too low a value causes clipping, too high, and the oscillations stop.

Thoughts, improvements, comments, cheap shots, whatever, thank you guys for being patient with me. :)

Live simulation ^^

Frosty90

Fri Feb 11 2011, 11:15AM

Registered Member #1617 Joined: Fri Aug 01 2008, 07:31AM
Location: Adelaide, South Australia
Posts: 139

Looks a bit like a hartley oscillator. The general idea is you have an amplifier with some gain and say a 180 degree phase shift, with feedback which provides another 180 degree phase shift, and some attenuation. If the total gain around the loop is greater than 1, oscillations will start from random noise. As the oscillation amplitude gets larger, some nonlinearity in the circuit causes gain compression, and the gain falls. If it drops below 1, there is not so much gain compression so the gain can increase again, and the amplkitude can build. So the amplitude settles at a value so that the gain is 1, and you have steady oscillation.

There are several different kinds of LC oscillators like this, the main difference between them is just the feedback network: so if your amplifing element gives a 180 degree phase shift, so should your feed back network. If it gives no phase shift (like maybe a common base amplifier), the feedback network should also give no phase shift at resonance.

In your circuit, I think the 100pf and the transformer form the feedback network (a pi network, the two halfs of the transformer form the legs and the cap across the top.) which feeds back by the 22nF cap, the transistor forms basically a common emmiter amplifier with a emmitter resistor.

Mattski

Fri Feb 11 2011, 05:11PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

Given the low oscillation frequency (70Hz apparently) I don't think this is operating like an LC oscillator, it might be working like an RC relaxation oscillator.

22nF cap charges through the transformer (current limited by 108R resistor) when 22nF cap is charged near 24V current flows through the zener turns the BJT on. BJT discharges 22nF cap, limited by the pot and the resistor on the transformer. Not sure exactly what the 100pF cap does. But I'm not sure exactly how to get the numbers to work out to 70Hz oscillation frequency.

Conundrum

Fri Feb 11 2011, 05:58PM

Registered Member #96 Joined: Thu Feb 09 2006, 05:37PM
Location: CI, Earth
Posts: 4061

sounds good, i've made an oscillator with a backwards transistor before (google "negistor") as well as a homemade tunnel diode using J310 and BC212 transistors with resistors that I had to hand, was amazed that they worked as well as an opamp.
also found a neat little dual tunnel diode transmitter using the ebay ones,

I have some real tunnel diodes here circa the mid 1970's

sent a few to other 4HV'ers.

A related device can be made from a germanium point contact diode salvaged from a defunct radio, and carefully "cooked" by overcurrent until the current begins to drop. This causes it to behave a bit like a tunnel diode, and can oscillate with some combinations of L and C and series constant current from JFET with G-S which gives about 2-5mA with a 2N3819.

-A

#include "Not another lost parcel.h"

Inducktion

Fri Feb 11 2011, 09:01PM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

Hm, I might have to actually build this, be interesting to see working. :)
and that's not a 100 pf cap, thats the resonant cap set at 100 uf, it works with different waveforms for all cap values. Sine waves only appear at 1uf and higher. Square waves and inductive spikes occur below that range, unless you change the inductance value of the transformer, in which case sine wave oscillations will occur. This is kind of interesting to see.

Inducktion

Sat Feb 12 2011, 04:37AM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

Woo! I built it, and guess what?

It actually works! I had to change out the transistor for a darlington though, but it does indeed oscillate in a nice sine wave pattern!
PICS

There's the scope trace, a nice sine wave with a little clipping on the bottom half, though it smooths out under load, like a flyback transformer, which works, but barely.

There's the circuit itself, not the prettiest thing in the world but it works quite to my surprise and enjoyment. I have a question though, how could i push more current through this, because the resistors at the bottom get hot, and the flyback hardly works, but does work with my circuit. If i take the resistors out, the zener diode dies, and I already had to replace one. Do I just need to put a resistor in series with the zener to prevent it from dying?

Thanks!

I'm so happy to design something that works IMO, so sorry for my giddyness ^^

Edit; Okay, changed a few more things, figured out it works with all transistors now, including only 8 gain ones, and I put a 5k pot in series with the zener, allows for amplitude control, and let me get rid of the emitter resistors. Getting small, like 2mm sparks from flyback, but my batteries are dead, and I'm letting them charge overnight to see how well it works on full batteries.

Edit edit,
Let batteries charge, no difference in arc length unfortunately.
With some minor modifications it works with FETS and IGBT's as well, but the arc length still remains the same for all versions. I don't know, not enough current? Under heavy load the oscillations are heavily damped, and heavy enough they stop altogether, or skip, like oscillate, then die, oscillate, then die, etc.

I don't know what use this could be, maybe a radio transmitter with a fet? That'd be interesting. Wireless power would work, maybe, under really light loads im sure. Does NOT work to drive a flyback, the sparks are extremely small. Oh well, it was fun playing around with something I designed myself. ^^

Inducktion

Sat Feb 12 2011, 07:19PM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

So, any ideas what to do with the circuit now?? Or give up on it? :P

Martin King

Sat Feb 12 2011, 09:26PM

Registered Member #3040 Joined: Tue Jul 27 2010, 03:15PM
Location: South of London. UK
Posts: 237

Good start, the general rule in electronics seems to be if you want to build an oscillator it won't and if you want to build an amplifier it will

Martin.

Inducktion

Sun Feb 13 2011, 10:59PM

Registered Member #3637 Joined: Fri Jan 21 2011, 11:07PM
Location: Buffalo, NY
Posts: 1068

Wow! This circuit of mine works CRAZY well for induction heating!
I took out the ferrite core transformer, and all it uses is a tank circuit and some mosfets, heated my screwdriver red hot in less than 5 seconds.

Wow. Using fets, its amazing...

I'll post pics and traces in a little bit. It's so easy a caveman could do it.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob