> It seems that using a flyback transformer would be a good way to go, but this > is itself an example of exactly the sort of thing that has me all riled up! > Now it is more than possible that I might be under some gross > misunderstanding in what follows, but a transformer is a transformer. There > is no such thing as a flyback transformer.
That's not entirely true. There are many parameters that affect how useful a transformer is for a specific application, and in particular, transformers designed for flyback duty include a gap in the magnetic path, to store energy to be released in the "flyback" phase when the magnetic field collapses. Transformers for some other uses are not gapped. Note that the gap itself is not always a physical gap in the core, and some powdered cores (there are several designs) have a discontinuous magnetic path on a small scale, which can act as a "distributed gap". > You can use a transformer in a flyback design, but that doesn't make the > transformer special. It is still just a transformer, and the same transformer > could be used in a variety of other designs and applications. Yes and no. Gapped transformers have different behaviour when it comes to magnetic saturation, and some designs depend on those differences to work well. > However, the world seems to have decide that there are a wide variety of > different types of transformer. This is a small sample: PoE, SMPS, Pulse, > CFL, Power, Energy harvesting, Coupled inductors (this last one deserves its > own place in hell, isn't that just another term for transformer?) etc. Many of these are optimized for various parameters. All transformers are a compromise between a bewildering number of parameters (several loss mechanisms, leaking inductance, capacitance, high voltage standoff capability, etc.), and which choices are appropriate depend on how the transformer is going to be used. > Distributors and manufacturers split their offerings into these categories, > making it especially difficult to actually find a transformer when you are > trying to search on windings ratio! Quite. The design notes for the LT8304 point to some useful places to obtain transformers with a variety of ratios. For high-ratio transformers, I find that CCFL drivers are a reasonably common and affordable category. I sent a request to a manufacturer of mains transformers a while back, curious about their CRT power supply units, which are still listed in their catalogue, but their distributors didn't carry them. I was quoted nearly US$500 apiece for them. > Oh. And it turns out that a lot of these actually have a 1:1 windings ratio. > Sigh. With a gapped transformer, and sometimes even without, you can get a large step-up with a 1:1 unit. There are plenty of geiger counter supplies out there built around 1:1 audio coupling transformers, deriving better than 400V from a 9V supply. I eyeballed those, but decided that since I'm aiming for much higher voltages (I'm looking to power things like lasers and CRTs that operate on many kilovolts and consume several watts), I'd explore other concepts. > But wait. it gets worse! Even assuming that you can actually find the > windings ratio, the world has then decided that it is a good idea to talk > about primary and secondary coils. Now it could be that there is some useful > distinction between these two coils in a real world design, but as far as I > know transformers are symmetrical devices. I can feed either coil, and use > the other as the output, as long as I don't exceed the specs for each. Mostly true, but you can get some unexpected results, especially when operating at high frequencies (the regime where flyback circuits usually operate). I've seen some doozies in my time, one due to the fact that the outermost winding happened to be an efficient RF radiator. > It gets even worse! It seems like the world of transformers is especially > fluid. Parts come and go like leaves in a storm (o.k., there might be a > better metaphor, but I'm on a rant here!). Oh, absolutely. And most transformers are produced for a given buyer, and the specifications aren't available (although there are ways to derive them). A few builders here design (and even wind) their own transformers. They can probably offer much more concrete explanations than I can. I do understand your frustration: most writing on the subject treats transformers in a fairly simple idealized manner as just two coupled inductances with a turns ratio, leads with pages filled with imposing equations, or just considers magnetics as "black magic". It took me a while to find an accessible primer on the subject, but I can recommend a reasonably compact and fairly readable reference that covers a lot of the basic details without getting bogged down in the physics and mathematics involved: The "Trilogy of Magnetics", by Würth Elektronik, a very nice book (in English) by a German transformer company (an earlier version was entitled something like "The ABCs of Transformers and Coils", and retailed for $14). Unfortunately, the current version is a bit dear at €49, and used copies are tricky to find. > At this point, I'm figuring that I'll just have to wind my own, which is > possibly a step further than I had wanted to go! Anyway I've bought an LCR > meter to help me out, and I'm poking around in the old parts from the boiler > in my basement. 20kV anyone? I feel a digression coming on. Ah, you share my interest in high voltage! Digress away! At those voltages, I start eyeing the "LOPT" or "flyback" transformers used in CRT circuits and things like "neon sign" and "ignition" (both oil burner and automotive) transformers. That stuff is fun, the failures are spectacular, and safety is absolutely critical (and difficult: things that seem like insulators at lower voltages aren't really, most test equipment can't protect you, you can fill an entire room with a powerful RF field by mistake, and arcs can jump surprisingly far to get you). I've been working with high voltage for decades, and I've made every one of those mistakes. - John -- You received this message because you are subscribed to the Google Groups "neonixie-l" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send an email to [email protected]. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/67539DE9-1B76-4203-ADE3-0A6B5EAE731D%40mac.com. For more options, visit https://groups.google.com/d/optout.
