Simple Power oscillator

Low RF, like 100-200kHz. MHz are for dielectric heating.

Slap together a work coil, tank capacitor (look for polypropylene capacitors) and a power BJT. Stay away from MOSFETs, at least for now. Add bias resistor, bypass capacitor and feedback coil (making this a tuned-collector Hartley or so) and apply power.

This is the sort of thing people would use 2N3055s for, but only in the 10s of kHz range. Try something faster, like fT > 5MHz.

Tim

-- Deep Fryer: A very philosophical monk. Website @

Internet is great! I've asked a question from Italy and in moment I had an useful answer from the Wisconsin!

Thank you very much Tim for your fast reply.

I've also given a fast glance at your very interesting web page. In particular I've seen your solid state Tesla coil. Probably this is the closest approach of the oscillator that I've in mind. In fact, when I spoke about "experiment on RF induction" I had in mind something like this:

Best regards, Antonio

What you want to do is simple, but...........

You need to go search FCC records to find the frequency allocation for dielectric heating. If you select a random frequency, you could find yourself in bad FCC trouble.

I think there may be frequencies around 13 MHz in use for inductive heating (plywood mills for example), but not for RFID.

What will be the load device for the MOPA (master oscillator power amplifier)? I always favored a Pi-net for random impedances, and a balun of some variety when the load was balanced.

Hi Don Bowey, I try to describe what I've in mind. Suppose a RF oscillator (e.g. Colpitts) which a LC resonator working at few MHz. The inductance L should be a simple loop of copper wire (one turn with diameter of about 0.7 feet). Placing near this loop another little LC resonating at the same frequency the RF energy will be efficiently transferred in this last one. If, for example, the power oscillator is about 5 watt, the transferred energy is able to light up a little bulb (for example). While I've experience with antennas, high Q LC etc. I've not a scheme for that type of oscillator. In my modest opinion it has to be as much as possible easy, just to begin some experiments. Thank you for your attention, Antonio

I would strongly suggest you using a tube unless your father is a manufacturer of RF transistors. An uncontrolled transistor or FET power RF oscillator can go wild and blow itself in a fraction of a second. You need a carefully designed and very fast acting protection. A tube is much more forgiving. Also, since the gain of transistor is less compared to a tube, the mode of the generation will be strongly dependent upon the loading.

Vladimir Vassilevsky DSP and Mixed Signal Consultant

Thanks Vladimir, I will take in account your advice too.

Antonio

The last induction heater I was around used a pair of EIMAC 304TL triodes.

If you build a power oscillator, the frequency is likely to be unstable when the load impedance or power supply voltage changes. This means that you are likely to pollute an unnecessarily wide range of the RF spectrum and you might find that this causes you to receive complaints from nearby users of the RF spectrum, who may call in the authorities.

I strongly recommend that you get a crystal oscillator and then feed the output of this into a power amplifier. The power amplifier could be built with transistors or valves (vacuum tubes).

Chris

OK, thanks to Michael and Chris too.

Now the design requirements are more clear.

- crystal oscillator stage in an allowed frequency

- buffer and PA stage

Cheers Antonio

This is the right, and probably only legal way to do it. The maximum power that can be put through a crystal is on the order of 1 mw; so you need several stages to get appreciable power out.

Several people have mentioned using vacuum tubes. My offhand guess is that the cost crossover point between transistors and tubes is around 200W output.

Tam