Building A Transformer

[Tom McCarthy]

Hey all,

This is not a thread for designing Optimus Prime
I've recently been pondering the idea of building my own transformer as there seems to be a shortage of cheaply priced ones on eBay since I last looked - quite some time ago - and it would allow me to start building at least one part of my Fusor.

I've asked my parents to get me The Art Of Electronics and one other elec. eng. book for Christmas, so hopefully they'll give me some safety grounding before I try my hands at anything if I decide to. On the other hand I got an Arduino recently, it's not a tool that's going to kill you, but I've been getting some good use out of it in designing simple circuits.

So, would any of you recommend building a transformer? If not, why not? And if yes, where would you recommend starting and what tips could you give me?

Thanks,
Tom

[Ross Moffett]

That's my plan.. I'm building a high-frequency one because I found I can purchase kW rated ferrite cores at Mouser.

I own "The Art of Electronics" which contains minimal, not very useful, information pertaining to designing a high voltage switched mode power supply.

Several years ago I was looking to build an audio output transformer and couldn't find a single source for low quantities of steel laminate material, so my guess is that may still be pretty much the case and leave you with the only option being high frequency supply. The exception being those who are willing to de-construct other transformers and re-wind them. Usually it's shoddy looking, difficult work. Pole pigs already contain this kind of transformer, but that's very dangerous territory if you're not familiar with them.

A friend who has done what I'm doing before recommended that ordinary magnet wire can withstand up to 10kv in air, or 20-30kv submerged in oil and vacuum evacuated of air. I'll probably pot in epoxy instead. At higher frequencies voltage multipliers become more efficient, so you can get higher currents at high voltage than is possible with 50-60hz neon transformers.

[Jerry Biehler]

I would not recommend it. A transformer is more than just wrapping magnet wire around a core.

[Rich Feldman]

An ebay vendor called siliconsector sells bespoke toroidal cores wound from steel ribbon.
But mechanical winding of wire onto toroids is far from simple.

Tom:
Making custom transformers can be rewarding and educational, though the era when it was common among ham radio builders has passed.
The quantitative EM physics knowledge to do it properly would be a very narrow slice of an introductory college course.
Not too much for you w/ good book or Internet references or tutor.

But for fusor voltage levels, a mains frequency transformer would be an extremely advanced project. You need 10000's of turns of extremely fine wire, mechanically wound, and must be set up for vacuum potting. Better to learn using a scrounged NST or XRT.

"Switcher" designs reduce the turns counts by factors of 10's, maybe even by 100 if followed by voltage multipliers.
But you need _lots_ more serious electronics up front, and still need your own potting for HV.
The electronics knowledge is much more than you would otherwise need in a fusor project.

[Tom McCarthy]

So the general advice, based on my age - realistically speaking - would be not to build my own transformer from scrap?

And for acquiring a transformer, either to buy a built transformer, or to buy an NST or X-ray Transformer and manipulate them to suit my needs?

Thanks,
Tom

[Ross Moffett]

Don't kid yourself.. age isn't getting in your way, you need to understand the electronics in order to design something like this though.

Buying something pre-made is always easy.

[Tom McCarthy]

I'll rephrase that - based on my experience in the field of electronics, the general advice would be not to build my own transformer, or spend quite a while acquiring the knowledge to build a transformer suitable for a Fusor.

I'm beginning to get on the construction front so I'm trying to find out what I can build myself rather than buy. I'd rather purchase a transformer, or get one off a College, but if it is more economical and efficient to build my own then I'm willing to do that, alas, safety-wise and the time it would take to get what I need to know...'knowed' and the work there seems to be involved, I don't think it's wise.

Tom

[Andrew Robinson]

Don't get discouraged. I built a few transformers several years ago during my Tesla days and it can be done, but as many above have said, it's not as simple as winding wire around a core. There are a lot of variables to consider. I would recommend starting small and do some experimenting. Learn as much as you can about transformer design and operation, materials, etc. Design and test small, then scale up once you get the hang of things.

[Troy Peterson]

Hi,

This website has lots of information on building your own heavy iron transformers: http://ludens.cl/Electron/trafos/trafos.html. Might be worth a look.

Cheers,
Troy.

[Tom McCarthy]

I would recommend starting small and do some experimenting. Learn as much as you can about transformer design and operation, materials, etc. Design and test small, then scale up once you get the hang of things.

I'm going to follow this path. However, when you say start ''small'' how small would you reckon I should start, I don't think I want to go straight into the big transformer types that Ross is talking about, but something relatively easy for me to do in my free time.

I'll check out that site over the weekend Troy.
Edit: I've started reading on that site Troy and it's really useful. Everything is explained - so far - nice and simple with a good bit of detail, thanks for that!

Thanks for the advice guys,
Tom

[Garett Goodale]

Hey Tom,
It sounds like your on your way to building a transformer, but I figured I would throw in my two cents nonetheless. I have been learning about how to build transformers have so far had frugal results. That's not to say that it can't be done, and I most certainly think it could be done in a decent amount of time, it will just require more effort. I would suggest a full wave CW multiplier. There is an abundance of information on them online, they are easy to build, easy to insulate (compared to a transformer), and on the whole probably an easier supply. I have built an operational one that I calculated to put out about 50kv. I don't have a probe, but I was able to arc 2.75 cm so I would say that I was probably close. Whether it would be the best power source for a fusor, I am not sure, but it is an idea.

[Tom McCarthy]

Ok! Thanks for the help Garett, I'll check them out online and see what's involved.

Thanks,
Tom

[Ross Moffett]

There's not a whole lot out there about making this kind of trafo. I happen to be over at my buddy's place in Austin to snap it for you.

The blue stuff is Litz wire, a bundle of small gauge magnet wire designed to transfer high-amperage, high frequency AC current in a small package. The high frequency skin effect normally limits current flow to the outside surface of the wire, but since litz wire is a bundle of a bunch of tiny wires, it allows for better transmission efficiency.

The secondary is wound from 22 gauge magnet wire with each layer insulated with some relatively thick mylar. That protects the secondary windings from arcing over to one another.

The two cores will be joined to form the complete transformer.

I'm not sure exactly what the objectives are so far as voltage and current with this one, but the final result will be encased in a container filled with mineral oil and pulled to a vacuum for some amount of time. It may take days with some fluid agitation, due to oil viscosity, to clear all the air. If that's not done, he says that plasma bubbles form in the air gaps, which then create carbon trails and everything is ruined.

Steve tells me that magnetic core saturation calculators are available, and that using those in concert with LTSpice allows for a successful project simulation.

[Tom McCarthy]

Sorry for the late reply Ross, I read the reply before I went to bed and decided that I'd reply the next day.

So you're saying that to actually build this type of transformer is not necessarily as physically complicated as it has been described?

Regarding working from a mains frequency, I've been reading up lately and it seems possible to CW multiply from a NST. However, would I be able to go about feeding a NST's output into a 'normal' transformer?

Three more points:
If I were to feed a CW multiplier mains/NST voltage (through a Variac), put it through a small number of stages (2-5), invert the DC output to AC and repeat the voltage multiplication process with another few CW stages, would this work?

When I'm reading about the, CWs are commonly described as working on no load. As the Fusor does not 'pull' power, a CW would still work, is this correct?

I'm assuming a CW does not break tradition in the way that input wattage will equal output wattage with no losses, so would I be better off starting with a current around 5A (I'm on 240v mains) to allow the voltage to multiply while not reducing current to a few milli or micro amps?

Cheers,
Tom

[Ross Moffett]

It's the driver, not the transformer, that's complicated with switched mode supplies of this nature.

You can multiply the voltage from a NST, but the low frequency means that efficiency is not good. Your current output will be lower than that of a SMPS with caps of the same value. There are calculators you can look up. You'll need a very expensive capacitor bank to get voltage multiplication at good currents.

[Richard Hull]

I hope everyone is fully cognisant that in all the years of folks working on HV supplies here and in other venues, if it were in some fashion easier or cheaper to do it this way or that way, that it would not have already been done hundreds of times and openly recommended here!

There is no cheap, fast, instant way to obtain full fusor fusing voltage other than......

1. A lucky find of a rather inexpensive supply of the proper voltage, polarity, and current handling ability on e-bay, a hamfest, a surplus store or by gift.
2. A disused old supply donated or loaned from an outside source that you have interested in your project.

A proper supply that is new and ordered outright will run over $3,000, period!
You can build your own supply for as little as $100.00 or as much as $1,000.00 based on your skill sets, knowledge of electricity and electronics, your scrounging abilities and your mechanical skills.

Barring all the above, it can be stated that........

No where on earth will you find exactly what you want that is fully capable of doing fusion at an obvious and easily detectable rate of 2 million + fusions per second in a fusor that is a simple, fast and inexpensive solution! Nor can you readily devise one from a neon sign transformer, a flyback, an Information unlimited kit, a tesla coil, etc.

The two great hardware issues that hold back the "Git r Done", DIY , "I'm gonna do this", kind of hopeful newbie is in obtaining and proper operation of a suitable power supply and suitable neutron counter. Beyond this, an even greater limiting factor is the lack of knowledge to even understand certain minimal basics involved with these two critical components. Any guy in shop class can build a fusor, but making it "go" separates the non-hackers from those who pack th' gear needed to actually do fusion.

The hundreds of rehashings of all the power supply issues with hundreds of folks trying to find non-extant work-arounds on the cheap or on the fast has been a long and oft repeated process here. Fortunately, the gifted and talented will win out and it is why the number of folks in the neutron club is limited.

Yes, money can buy you into this exclusive club rather quickly if you outright purchase all you need at the professional level of function, (approx $25,000), but somehow doing it all from scratch for well under $1,000 by using acquired knowledge and skill sets is more internally rewarding and will serve one better in the continuance along life's path.

Richard Hull