Page 1 of 1

Flyback Transformer

Posted: Sun May 20, 2012 4:51 pm
by Osi
I've been reading a lot about fusor design, and because my school year has just finished I thought that this would be a cool summer project. I've been looking into the different materials and it seems as if one the most difficult piece of equipment to find is the power supply. For fusion the FAQ says I should have about 30kv and 20ma which is quite a lot of power. I found this website on a home built flyback transformer ( http://www.teravolt.org/fryback-transformer/ ) and it made it seem to be relatively easy to get the power requirements. Because of this a few questions popped up:

1. Why are power supplies expensive when it seems relatively easy to produce high voltages and a few milli amps?

2. Has anyone attempted using a homemade flyback transformer before (a quick search through the forum search engine wasn't too helpful) for a fusor?

3. Because flyback transformers (at least how I understand them) requires a high frequency AC is it difficult to rectify it to DC?

4. the website I mentioned previously used a ZVS driver ( http://www.teravolt.org/zvs-driver/ ) to convert his DC batteries into AC for the transformer. I believe I would have to step down the 120v mains to 36 volts, convert to DC (full-wave rectifier with caps) , apply the ZVS driver, apply to the flyback transformer, and then re-rectify it for negative voltage into the fusor. Is this thinking correct?

5. Is there an easier way to do this?

Re: Flyback Transformer

Posted: Mon May 21, 2012 10:51 am
by Richard Hull
There have been numerous discussions noting that a standard flyback is just not suitable in any instance to make a real neutron producing fusor.

In the real high power, $4,000 supplies they use a flyback. It is a special flyback though. Often, the cores are 4-5 times the size of the one you are looking at in the video. Rarely do these large cores put out much over 10kv as is the case of this custom wound flyback at "teravolt". They do put out significant current, much as seen on the video.

The key is getting the needed 40 kv or more at useful current output levels. This is done with very expensive voltage multiplier diodes that are fast enough and at a high enough PIV rating to even assemble a voltage quadrupling multiplier stack.

With the proper parts in hand and a lot of custom work, and good insulation techniques, a large flyback and multiplier system can be readily assembled to rival any $5,000 HV supply. It might not be as regulated or have any current or voltage stability or sensing controls, but it will work well in a fusion system.

How handy are you? Do you have a source for the special diodes?

I would hold out for an x-ray transformer or a good used surplus supply on e-bay unless you are fairly well skilled in HV system assembly and comforatable with electronic assembly work.

Richard Hull

Re: Flyback Transformer

Posted: Mon May 21, 2012 11:21 am
by Osi
I've worked with electronics before (printed circuit boards and smd soldering) but never had much experience with high voltage electronics.

I was thinking of waiting for a good find on ebay or other surplus stores for the power supply as that would be the easiest method, but the problem is that its so hard to find a decent and suitable system and when I do its usually too expensive. The reason why the flyback transformer is intriguing is because it would be fairly simple to assemble with the only time consuming part being the coils.

What I don't understand is why would you use a multiplier, which would require expensive diodes and caps to handle the current, when a transformer could do most of the work? Is it physically impossible to design a flyback with 40 volts input at 40khz for a 30000v 25ma output, or is it just impractical? Although standard flybacks in tvs and crts would be insufficient it seems perhaps it is possible with a homemade one.

Re: Flyback Transformer

Posted: Mon May 21, 2012 12:53 pm
by Carl Willis
HV power supplies are expensive because they usually do more than just convert low voltage to high voltage. There are feedback and control circuits, filters, and protection systems. The engineering is not trivial. High voltage is a special niche in the power supply world, rather than a massive market such as exists for computer power supplies.

I have built my own ~2 kW high frequency ferrite-cored step-up transformers. Jon Rosenstiel has done it also. I think Tyler Christensen has done it. Details can be found with the Search feature. Steve Ward over at the 4HV forums has extensive experience and I recommend you read his discussions. None of these HV transformers are technically "flyback" transformers. Instead they were designed to be as nearly ideal transformers as possible, as favored by the SLR driver being used.

The commonly-referenced "ZVS flyback driver" is a Royer oscillator made with mosfets or IGBTs in a push-pull configuration. This circuit as I understand it does require a transformer with considerable leakage inductance, like a gapped-core flyback transformer. Designing a suitable transformer involves some different considerations from ideal transformers, although I think the practical aspects of construction would be no different.

It is not necessary to reduce the input voltage to 36V, and in fact for high-power designs it's preferable to just drive them right off the 120/240V power line (suitably rectified and filtered to provide DC). The reason many people use low-voltage DC with their "ZVS flyback drivers" is because they are typically driving low-power horizontal-output transformers taken from TV sets, transformers where there is little real-estate for making new primary windings on the core. If you wanted to drive these small transformers off the line, you'd need to put a hundred turns or something like that on the primary, which is pretty impractical. The Royer circuit's feedback mechanism to the gates of the IGBTs is also prone to HV gate breakdown, something that's easier to avoid if the supply voltage is low. This design also requires a high-Vds transistor relative to, say, the SLR design because of the oscillatory high voltages developed in the Royer primary circuit. That implies more expensive parts at high power. However, the Royer design works fine for driving tremendous flybacks at the kW level. I have done that with a Royer-powered induction heater and a ~10 kW flyback from an ion implanter supply. This situation resulted in foot-long electrical arcs and it is going to be unusable unless the transformer is potted in oil.

-Carl

Re: Flyback Transformer

Posted: Mon May 21, 2012 8:12 pm
by Osi
@Chris Willis
Would you happen to have a link to your power supply design? I tried searching your name through the forum, but can't find the specific post.

I see that both Steve and Jon ( download_thread.php?site=fusor&bn=fusor ... 1173980351 ) use a Cockroft Walton multiplier approach. Where do they find the components for such beefy devices, I've seen plenty of small power multipliers, but nothing that can handle 30kv and 20 ma. Could you put several small power multipliers (that go to 30kv) in parallel to gain the needed current?

For the multiplier I suppose you would use a high voltage transformer (flyback again?) running at a relative high frequency to power the multiplier?

Re: Flyback Transformer

Posted: Mon May 21, 2012 8:25 pm
by Tyler Christensen
I posted a controller here a while ago: viewtopic.php?f=11&t=4846#p27684

It has feedback for both voltage and current, and ended up working pretty well on my fusor, although there are a few design changes I've made. Since I work on my fusor for a grand total of about a week and a half a year now, it's not likely to be updated any time soon or have any further results.

There is a controller without feedback on Steve Ward's website.


You can put multipliers in parallel, and it is done in industry to allow for multiple rack supplies to power one load. You need significant effort in load balancing, though.


It is worth noting that not many retail supplies actually use SLR's. Glassman runs a resonant secondary and hard switches half the transitions.


As for why we use multipliers, it's mainly because it's not a practical thing to insulate a small ferrite transformer for only a few kilowatts to more than 10-15kV.

Re: Flyback Transformer

Posted: Tue May 22, 2012 7:12 pm
by Osi
Intresting things to think about. I've been trying to do a little bit more research and I see that finding an x-ray transformer is usually the way to go and I suppose this is because it offers an easy way to connect directly to the mains?

I was also looking into other flyback transformers, as I think winding my own is a bad idea for safety matters as well as its time consuming, and I found this nice site: http://www.amazing1.com/transformers.htm . Scroll down a quarter of the way and there is a flyback with a 30kv peak and 20 ma which seems perfect for a fusor. They even provide a nice schematic, which is a bit overkill in my opinion. Anyone think that this transformer could become a potential power supply?

Also a question about component selecting. Which statistics do I need too look for when choosing diodes and capacitors? What i find confusing is that, in Steve's high voltage CW ( http://www.stevehv.4hv.org/5STCW.htm ) he uses 40kv 2nf doorknob capcitors but gets a 150kV output? And after stringing diodes together to form a 64kV diode he also gets a 150kV output? It would seem that the components would be the limiting factor of what a CW multiplier can do, but it appears as if the values don't make much difference when I know they do.

Re: Flyback Transformer

Posted: Tue May 22, 2012 7:18 pm
by Tyler Christensen
It looks like they're doing ZVS on that amazing1 driver. That could also work, I've always personally preferred ZCS.

The multiplier is using 40kv capacitors with 5 in series (5 multiplier steps) making it effectively a 200kv capacitor, and the diodes are 64kv*5 with respect to the full output voltage. This puts everything well within spec.

Re: Flyback Transformer

Posted: Wed May 23, 2012 9:57 am
by Dennis P Brown
OsiVD (??? Thats your name?)

The link/site for HV x-formers is really great. I located the perfect battery driven system to power the ionizer for my deuteron gun (I needed at least 10 KeV.) Thank you very much! That saves me the trouble/time to make mine. That was the last item I need to get the Linear accelerator up and running!

Thanks again!

Re: Flyback Transformer

Posted: Wed May 23, 2012 10:13 am
by Osi
Yea, my real name is Osi, (VD is just my last name initials).

I was having second thoughts about the flyback transformer now as I was asking some people over at the 4hv.org forums and one post said the site was a bit sketchy. Here's a link to my post: http://4hv.org/e107_plugins/forum/forum ... php?138850 . I'm not sure what to do right now, either buy the flyback and build an adequate driver (which isn't the problem), make my own transformer (I don't think i know enough of HV to attempt this), or wait for the impossible x-ray transformer.

Has anyone purchased an item form amazing1.com ? (now that I think about it, the url is a little sketchy....). It would be nice if their product description was accurate, any insights on this?

Re: Flyback Transformer

Posted: Wed May 23, 2012 12:02 pm
by Tyler Christensen
Building your own is all that easy (although it's not trivial either). It doesn't take much electrical knowledge, just some craft in insulating everything. You can order cores very cheap from TSC international.

As for amazing1, they are a legitimate company and will send you what you order. As to whether the products are legitimately quality is another question.

Re: Flyback Transformer

Posted: Wed May 23, 2012 12:14 pm
by Dennis P Brown
I am very unknowledgeable on HV above 10 KeV but I listen to people who are and some of what "bwang" says is correct (about diodes, least.) Maybe his overall approach is correct. Maybe use some of bwang's ideas as a starting point to search both here and for google. Building a very HV, 30+ ma system is something that will pay off both in savings and safety later on. Can’t say for sure but that is my limited take on what they wrote.

Re: Flyback Transformer

Posted: Wed May 23, 2012 12:49 pm
by Carl Willis
Amazing1.com is Bob Iannini's site. He was formerly a rather prolific author of DIY articles for Popular Electronics, Radio Electronics, and some other publications, as well as his own books. His e-business is legitimate; I've bought some little things from the site in the past for the hands-on volunteer activities I do with a local museum, and it's decent stuff (although understandably quite a bit more expensive than the equivalent found through opportunistic scrounging or rolling your own).

The voltage / current capabilities quoted on Amazing1 for the flyback transformers are definitely pushing it for continuous duty. I have no doubt you could obtain the voltages instantaneously, the currents instantaneously, or perhaps both instantaneously if the driver is suitably beefy, but I don't think you can bank on continuous operation at ~30 kV / 20 mA from such an item. My opinion is that you would do better to wind your own lower-voltage (~10-kV) transformer and follow it with a Cockroft-Walton multiplier.

-Carl