|
|
~High Voltage Power Supply v2~
(Codename: Advanced Flyback Driver)
High Voltage PSU circuit:
The idea for this circuit came from another project I was doing at the time. I was asked to build a function generator that produced a sine, square and triangular waveforms without using any specialist IC's, so as you can imagine this was quite a challenge if you never built one before. So after a little research I came to the conclusion I would have to use an existing design and modify it for my needs. The circuit is based around a TL084 quad op-amp which was choosen as it works a lot faster than the LM317 ic's.
Overview: Produces near perfect square wave and triangle waves for driving mosfets/transistors. Input voltage required is 12V-18V at about 200mA no load. Frequency range 16-80Khz adjustable (or others depending on just capacitor values used)
The function generator workings:
The circuit first produces a triangle wave (op-amp one) and then intergrates that into a square wave using a second op-amp. To produce a sine wave, we shape the original triangle wave into a smoothed top and bottom using a diode and capacitor around a mosfet transistor. The circuit operated nicely at the various frequency ranges 20hz - 20khz and produced extremely sharp square waves and triangle waves.
Circuit: Click to view
The sine wave that was produced was not very good but probably ok for testing equipment. Please note that this circuit uses a floating ground point to allow the op-amps to swing below a set 0 point and so that they do not require a negative voltage power supply.
The "Advanced Flyback Driver":
The HV driver circuit modification: Since we are not interested in the sine wave we can discard this part of the circuit (you could use it if you wanted but I didn't). Also we cannot have the floating ground point as observed at op-amp 2 output as this will keep the transistor on when it should be off. To do this we use the forward voltage drop of diodes to decrease the value to near 0 (eg below 0.7v). Three 1N4148 diodes are used to subtract 2.1v from the output. We also want to isolate the op-amps from the output as best we can, so a transistor driving the mosfet gate is used. As for frequency range, I decided to have the widest range availble for tranformer compatibility. A range selctor is still used but the capactior values have been tweaked to produce range of 16-25khz, 30-47khz and 60-80khz. So it covers all flyback tranformer types (usually 18khz).
Circuit: Click to view (I do appologise about the quality, but have not re-drawn the diagrams yet)
The components:
Outcome: The advantage to using two mosfets to drive the tranformer is that they don't get too hot when running for extendend periods of time as they share the load. Also the amount of current that is availible to the tranformer at any one time is significantly higher; perfect for arcs!
There aren't too many components needed and I've found that by producing a near perfect square wave and driving a flyback transformer with this, it perform alot better than I ever got with a 555 circuit. Also you have the capability of using triangle waves and sine waves to drive the mosfet with! Lots more to experiment with.
Construction: I also tried to keep the flying leads to the pot's as short as possible to avoid noise being introduced to the circuit. Inside my case, the circuit board is raised up from the metal case to prevent any spikes/static from destroying the circuit. The case is also grounded to the negative input to reduce circuit desruption from noise.
Cooling: I came up with a clever way for powering the fan and protecting it from noise (I have lost a few pc fans before when using them on HV equipment). The circuit arrangement is known as a pie filter arrangement (Found in an RF book). Basically it's very simply and you will prbably have come across it before. It consists of two capacitors and two chokes (coils) wound on ferrite bars. The circuit looks like this:
It has managed to protect the 12v fan from harsh rf spikes on my psu. If I create a continuous arc as seen in the photos below, it just slows or stops the fan rotating due to the current draw.
My HV PSU:
2 top photos are the prototypes when I was designing the circuit, The other 4 are the finished unit
My custom made flyback transformer:
I built this because I was fed up with trying to fiddle around with tv flybacks to get them working well enough to use for experiments. I tried everything from different driver 555/opamps to winding new primary coils on the outside of the flyback's ferrite core.
I have to say this was extremely hard to build and not something that took 5 minutes so be warned if you attempt this!
It's built around an old flyback ferrite cored striped from an old tv (1990 something) that had a large flyback. I started by winding about 5 layers of electrical tape around the top of the rectangular core. I then wrapped a piece of double sided tape around over the top of the tape and stuck the first wind of wire to the tape with super glue for added security. I then proceeded to wind the wire as close as possible to each other until i got to the end. After laying another layer of double sided tape, I preceeded to wind in the opposite direction so as to go up and back down.
Click to Enlarge
After about 55 layers with around 20 turns on each layer, I sealed the top layer off with a few tightly wound layers of sellotape (The best I could come up with!). Total turns on my flyback, around 1100.
The two halves of the core were then put back together using the retaining clip. The unit was fixed to a peice of expanded foam plastic (window trim) with a few rubber feet glued to the bottom to make a base.
The primary for the flyback is simply a couple of turns, 6 in my case, of heavy gauge copper wire (household earthing wire) with some heatshrink tubing shrunk around the wire. The two ends are soldered to a length of 3 core flex (2 used) that run to the hv psu.
The outter end turn of the secondary coil goes to a HV lead and crocodile clip (+ lead). The return of the secondary coil, the very first turn is left un-grounded and is used as the return lead (- lead) for AC projects. This could be grounded to the negative supply but I have observed no benefit in doing this. If HV DC is required you can add some HV rectifier diodes to the + lead output, so for my coil at 1.5Kv AC, it would become 800V DC (not too impressive though).
Finshed transformer:
Since I built this the flyback has worked faultlessly and has alot of current availble. Can't say how much but enough to draw arcs at 12v in to over 3.5cm. I could run this off alot more input voltage but don't really want to burn it out, as internal arcing may occur due to my "sellotape" insulation :-)
Operational frequency for maximum output is around the 30-50Khz region, this is higher than tv flybacks are normally run.
The reason I wound the wire around the short side is because I was interested to see if increasing the width of the winds means increasing the current and increasing the length overlapping the core is what determines the voltage. I can't say that I've found that answer but it sure does put out alot of current at a low voltage. Output voltage around 1.5Kv
In action:
Top left, Gas Discharge tube from a disposable camera. Top right, plasma globe arc through glass. Middle left, Plasma globe with flyback return lead brought close, Middle Right, Jaccobs ladder max gap 4cm Bottom left, Plasma globe, notive the halo at the top of the bulb, wire leading to flyback return.
Videos:
|
||||||||||||||||||
Please do not reproduce anything contained within my website, as it maybe hazardous to your health unless you fully understand what you are doing. I cannot be held responsible. This website is copyright. © Oliver Hunt 2006-2011
|
