High Voltage Bias for DIY Electrostatics

I have built a pair of electrostatics using Sanders' book. I'm having trouble building the permanent high voltage power supply. Right now I'm using a professional, loaned power supply which costs about $2K more than I'm willing to spend.
I plan on using a step up transformer (400-500Volts) and using capacitors and diodes to retify/multiply the voltage up to 3KV DC. The plans are all laid out in Sanders' book, but I cannot find a step up transformer that is sufficiently cheap. The bias draws almost no current, so power is not a problem. So, can anyone help me find a cheap (<$30) transformer that steps the voltage up to 400-500 Volts (given 110V at 60hz, of course). The current rating can be less than 10mA.
Thank you very much.
Thanks for replying, folks.
I have seen Barry's bias supply, but I'd like to try and build my own, and perhaps save some money. If it looks like it will be too expensive then I will swallow my pride and order a bias supply.
I'll check up on the neon signs though. I plan on using a scavenged variac to control the final voltage of the bias. Neon transformers seem like the most viable option.
Try hooking up a regular step-down transformer backwards. this is a very inexpensive way to use it as a step-up instead. The voltage ratio will be the same. Just be careful not to step up too far!!! The transformer should be hi-pot tested to a voltage well above what you intend for the output voltage.

Then, use your voltage multiplier to take it up to the level you need... Easy as pie!
After looking at some neon sign transformers, I ran across an obvious problem. How do I rectify 3-4KV in a way that is not cost-prohibitive? 1KV Diodes are no problem. Can I just put four of them in series, or do I have to worry about possible manufacturing differences in the diodes?

hifiZen, I've tried hooking up a transformer backwards. As I turned up the voltage on my variac the output did not climb linearly, and then I noticed smoke rising from the transformer. Obviously, it wasn't rated to take such high voltages. Finding a tranformer that is hi-pot tested above 500V is my problem. Do you know where I could get a cheap one?

Jake, thank you very much for this site. $19 doesn't seem like too much to ask for a 440Volt Transformer. I may just buy it and be on my way.
Thanks again folks.
No, I hadn't forgotten you, just been busy.
Yes, diodes can be run in series. You will need to run resistors in parallel with each one to ensure even division of the voltage. The rule of thumb for the resistance is to multiply the PIV of the diode by 500 ohms, i.e. for 1000PIV, use a 500k resistor.
It's also a good idea to run a small cap (~.01uF will do) in parallel with the diode and resistor pair.
Be very sure that you allow *plenty* of PIV for whatever rectifier scheme you decide to use. Diodes, resistors, and small caps are cheap. An out of control power supply could be very expensive, indeed.

After looking at my options I decided to do what seemed like the cheapest and simplist. I bought a couple of step down transformers off of Ebay and plan on using them backwards, as step up transformers. The model is Hammond 160H24, at http://www.hammondmfg.com/160.htm . I used a suicide cord and plugged it into the wall to see what would happen. It began to violently buzz and smoke. I am almost positive that there was no cross connections in my wiring. I did not have anything hooked up to the output, as I was about to measure the voltage across the leads. Is there anything obvious that I did to deserve such a fate? Should I have had a load across the output? The specs of the transformer say that it has been hipot tested at 2000V, which is well above the 440 that I was expecting. Any input on this is welcome.
Thank goodness I bought more than I needed.
Just thinking out loud (or at least in type), maybe you should've had a load across that sucker. Just imagining the primary by itself, its essentially a wire shorting the suicide cord. The built up EM field isnt being drained into the secondaries to dissipate energy. Perhaps this is what destroyed your transformer.
Then again, something doesnt sound quite right about this diagnosis, so I think I've just successfully talked out of my ***. But you do have several more of those things. Maybe we should try sticking a fuse in series with the suicide cord (making it a safety cord) and then putting one of our scavenged power resistors across the secondaries (in series, I assume you're running.). If you dont want to risk one of your heatsunk 2ohm power resistors, we can use that resitor bank that i ripped out of that power supply. Thats my input, anyway.
- Jonathan
People claim that using a step-down transformer backwards is a solution. Not just HifiZen, but Roger Sanders. Are they using a different type of transformer? Larger, more robust ones? My transformer looks about the size of the one that Barry Waldron sells on the Electrostatic Exchange. I haven't heard of the problem of saturation of the BH curve (whatever that is) before. The only thing that I do know is that my transformer couldn't take line voltage. Perhaps I am witness to this problem first hand.
Once again, I am left in the dark without a high voltage power supply.
Any help will be greatly appreciated.
I don't normally disagree with Petter, so let me try to catch up on this thread (been writing and working on tube phono stage--not enough hours in the day to do all this fun stuff)...
Let me verify what's going on.
You're running a smallish power transformer "backwards." It has dual primaries and dual secondaries (now being reversed). When you hook it up, it starts grunting and huffing and smoking.
Two things occur to me:
1) Are these transformers new or used? If new, that's one thing, but if used...
2) I'm thinking that there's a fairly strong possibility that you've got the windings hooked up funny, and that's what's giving you fits.
Petter's point about the flux saturating the core doesn't ring true for me (based on what I think I know--somebody correct me if I've misunderstood). The flux in a transformer is going to be proportional to the load. Since, in this case, the secondary (erstwhile primary) is open (Yes? A meter doesn't count, since it's going to be a skillion ohms going in the front end.), the reflected impedance back though the primary (formerly the secondary) is also going to be nigh onto infinite. Hence there's not a significant current drawn and the flux is going to be very, very low, although not zero. I find it difficult to envision the transformer core saturating under these conditions.
Besides, I've run transformers backwards before.

These transformers came out of apparently unopened boxes. I got them off of Ebay though, so one never knows...

Looking at the website specs : http://www.hammondmfg.com/160.htm , I see a possible problem. I shorted 6 and 7 and hooked up 5 and 8 to line. Simply because it was easier to do.
Should I hook up the hot lead to 5 and 7 and the neutral to 6 and 8? Or is this the wrong connection. I am pretty baffled here.
Well, I still have four to play with.
I think that I've found a solution.
I consulted the local electronics expert. He guessed that the reason that the transformer blew was because of shorting inside of the transformer. The side that was designed for 24 volts may not enjoy 110. Likewise for 110 and ~450. Hence, shorting and current draw, resulting in violent buzzing and smoke. He did, however, offer a solution to my problem. He suggested that I hookup all four transformers in series, but still as step up transformers. That way, each transformer only sees 28 volts, well within spec. Then, I merely hook up two transformer's secondaries in series, and I can graciously receive my 450+ volts.
The only problem that I have now, is that I would like to build 2 bias supplies. Since I have 4 transformers and I'm only using the ouput of two of them, it seems to me that I can work it out. The other two transformers are acting only as inductors. Perhaps, I can replace them with an inductor and save myself from buying 4 more transformers for the second supply.
Basically, I need to drop the voltage out of the wall by half and a resistor might get too hot. I am trying to do this without buying another step down transformer, but I may have to. My local expert has dissappeared for a couple of days before I could ask him about it, so any input is appreciated.
Don't mean to be a pain but I think you could save yourself alot of trouble just getting one of Barry's transformers, and hang your diode/cap ladder off it. Shouldn't cost much, has a primary rated for AC line voltage, gives you the ~500V steps, small size, etc. (I originally wanted to just hook one up backwards too). To fine tune the voltage, I use a simple voltage divider on the primary with a couple of high wattage Ohmite resistors, one has the adjustable tap. Not elegant, but works great!
I am very close to completing the high voltage bias supply with my own transformers. I have plenty of time to play with this stuff while I'm waiting on parts for my next set of electrostatics. I'm also making a pair as a gift. I'm pretty stubborn and want to try and finish this stuff with the things that I already have. I don't want to sink any more money than I have to into this project. Besides, it's a learning experience for me. But, if I cant get it to work I'll just buy Barry's supply.
Thanks everyone for your input.
I am very close to completing the high voltage bias supply with my own transformers. I have plenty of time to play with this stuff while I'm waiting on parts for my next set of electrostatics. I'm also making a pair as a gift. I'm pretty stubborn and want to try and finish this stuff with the things that I already have. I don't want to sink any more money than I have to into this project. Besides, it's a learning experience for me. But, if I cant get it to work I'll just buy Barry's supply.
Thanks everyone for your input.

I guess I'm a little late for the dance here but if you're still looking for a cheap, effective power supply, you could build a pair (2) like mine for about $50 in parts. They're configured for 1/16" stator-to-diaphragm spacing so you would need a higher output supply for larger d/s spacings. There is a parts list and schematic posted on my blog page here: The Jazzman's Electrostatic Loudspeaker Page
It's an old thread, but allow me to present my solution to you all.

I bought a 'cold cathode fluorescent lamp', or CCFL, the type of colored tubular lamp that kids put inside their computers to show off. They are dirt cheap, maybe $8. Those lamps need a high AC voltage, so each set contains a little resonant inverter, a small 1x3" box you can tuck away anywhere.

All you need then is just a short voltage multiplication ladder, and a series resistor. The inverters I have run at around 52kHz (43kHz loaded), so you can use tiny capacitors in the ladder -- I used 4.7pF 2kV, which are waaaay cheaper than what you need for 100 or 120Hz.

The inverters run off 12VDC, of course. Want to adjust the high voltage? No problem, just put a LM317 in front! My panels start to hiss at a supply voltage of around 5V.

Small, light, cheap, buy anywhere, adjustable, ho hum, ... It just doesn't get any better than this! And you can even pimp your speaker by putting the lamps in! :D