Voltage inverter circuit - Schematic dissection

[Puffie40]

I have been trying to figure out how this high voltage circuit works, but I hit a wall.

The two transistors are set up as a Darlington pair, and the 220 resistor in the middle is used to help speed up the switch times.  The voltage divider circuit calculates to about 2.21v, ignoring n2.

What I don't understand is how the circuit oscillates, especially given the two primary transformer windings.  n2 is used as feedback, though how or which winding (n1 or n3) it comes from is not known. 

Does n1 and the Cap form a tank circuit?

The transformer inductor values:
n1: 180uH
n2: 45uH
n3: 11.25H

[katzohki]

OK, so you're on the right track. What generally happens in oscillators that are similar to this is that current flows into the winding (n1) until either the transistor (common) or the transformer (less common, no more current can flow into it) reaches a point of "saturation."

If you know about transistor saturation you can see that the resistive divider 10k / 2.2k sets the base voltage at a certain level, so at what collector voltage does the transistor(s) reach saturation?

As that happens, energy builds up in the transformer (n2 will have current flowing through it towards ground). Now what I suspect happens at this point (I have most experience with "blocking oscillators" and this looks pretty similar) is that the voltage on n2 reverses and current flows "up" from ground. Remember that inductors want current to stay flowing.

The transistor(s) should then turn back on and the process repeats, hence you have an oscillator. Everything to the right of the transformer (n3) is a Cockroft-Walton multiplier.

http://en.wikipedia.org/wiki/Bipolar_junction_transistor
http://en.wikipedia.org/wiki/Cockcroft%E2%80%93Walton_generator
http://en.wikipedia.org/wiki/Switched-mode_power_supply

[LukeW]

It's sort of a Royer oscillator, but without the centre tapped primary.

http://en.wikipedia.org/wiki/Royer_oscillator

[Puffie40]

Thanks guys, that link was very helpful - I was looking at another HV oscillator setup and that actually explained it.

Looking around some more suggests the transformer is a dc-based flyback transformer-
(http://wiki.4hv.org/index.php/Flyback_transformer)



On the multiplier side, I found a EEV episode that covered it very well:

But one thing that was not completely explained was the values of the capacitors.  How were they chosen? One tutorial site mentioned the value is related to the frequency.

So, on that thought, how is the frequency determined?  Thinking that the cap formed a tank circuit with n1, I calculated a frequency around 1.2khz.

[katzohki]

The caps in the multiplier or in the input? Yeah it's related to frequency. In the multiplier I've experienced that value of the caps isn't very critical.

[Puffie40]

The caps in the multiplier or in the input?

The caps in the multiplier. What I read is you set the caps around 100uF if the frequency is around 60hz, and around 1nF for higher freqs.

[katzohki]

We have a few that I'm working on that use .047uF and .039uF, they run around 200-400kHz. They were sort of a "whatever will fit" choice more than something specifically tuned. If you have the luxury of adding a RC filter after the multiplier do it as it will reduce the ripple greatly.