Need good link to output transformer design.

[Steve Conner]

For SMPS transformers, Furukawa TEX-E triple-insulated magnet wire is a brilliant invention. It meets the statutory insulation requirements of IEC 60950 all by itself.
Features of Triple Insulated Wire TEX-E : Insulated Winding Wires : Furukawa Electric

I guess you could make guitar amp OPTs out of it, but someone would no doubt claim that it doesn't have the mojo of vintage paper bobbins.

[Alan0354]

Guys, I am just learning about transformers, That's the reason I want a good model to start. Don't SMPS transformer use the equivalent model? I don't think I'll ever make an OT, but I have every intention to make the SMPS transformer. At that frequency, you can get like 5 to 7V per turn!!! You don't need the number of turns like the audio transform and you use ferrite cord that is pre made. They better be more accurate. I just want to learn the basic of the transformer equivalent circuit. I saw my engineer making transformers, in fact I was involved in the design by using 5KV 30 gauge teflon wire as secondary of a 7KV 200W transformer. BTW, we do not use heavy insulation in the transformer either, the whole thing is the teflon cable, by default, we have at least 10KV dielectric insulation. The wire might be rated higher, I just don't remember, I told my engineer that's what I want and asked him to look specifically for small cable for the secondary. We even wire the two half together that put more burden on the insulation. It turn out it is important to have symmetrical waveform in SMPS, if you wind one over the other, the waveform of the two sides is not the same and it is not good.

I am just thinking I kill two birds with one stone by getting into the modelling and dig into the output transformer first. The SMPS transformer should use the same model as OT, just different requirement......more relax requirement as it work at one frequency.

[J M Fahey]

I got in to answer and saw RG's and Steve's ... plus your own reaction.
FORGET RDH4 FOR SMPS UNLESS YOU WANT TO MAKE A 60 HZ, EI SILICON STEEL CORE, MECHANICAL VIBRATOR DRIVEN ONE.
Capisce?
And no, **IT'S NOT THE SAME**
Period.

[Steve Conner]

I'd say that "Switching Power Supply Design" by Abraham Pressman is the RDH4 of SMPS. Look for a recent edition, lots of new material has been added over the years.

I've designed DC-DC power supplies for the stuff we make at work, but I stick to off-the-shelf magnetics, our production volume is too low to justify custom transformers. For DIY projects I'll have a go at winding anything, and as soon as RG's back is turned I'll plug it into the wall.

[Alan0354]

Give me a source of SMPS that can do off line to 450V, I might not even care to learn transformers. You can easily find LV DC to DC transformer or even off line and up to 48V transformers. We had to design and wind all our transformers.

I never even interested in transformer, that's the reason I had been avoiding it and know nothing except in the basic EM theory. I tripled my knowledge just from the last two days of studying, there goes to show how little I know. The rest of the MOSFET switching and voltage regulation of the SMPS is much easier for me.

[R.G.]

The big issue is how much power you need to get out of it. Under about 20-50W, you can do a flyback switcher and it's not all that hard. We had to do discrete and custom designs back in the day, but today you can get the entire controls in an 8-pin chip. A MOSFET and a little cut-and-try on the trannie, and you're there.

Flybacks are especially easy as there is no direct primary to secondary power transformer. The transformer is just an inductor that gets loaded up with M-field, then dumped out through the secondary, selected by diodes turning on.

450VDC is not particularly a problem. You start running out of MOSFET voltage for single MOSFET switches at 500-600V, as a single switch on a primary needs to take at least twice the power supply in that setup. Then you have to do things like double-switch to get lower voltage on the primary side switches.

[Alan0354]

I need off line switcher, the primary is on wall AC. Auto tranny is out, I need isolation. MOSFET are in the primary which is 1.4X220V=311V max. It is not that hard to find. 50W is easy. This is for guitar amp, I won't want anything bigger. Who want a 100W amp now a days?

At 5V per turn, it wound be 45 turns primary and 200 turns secondary. Doing it with in pull with two MOSFET. This will accept 100V to 220VAC. I think I need a 650V MOSFET as it fly to double the voltage like the tube push pull amp. I don't think there is any way out of this as even using a single MOSFET, still need to have 650V.

To play it safe, 60 primary and 250 secondary. Any off the shelf switching tranny like this?

[Steve Conner]

Not that I know of. The easiest way forward would be to get an ETD core assembly and wind it yourself.

The Crate Blue Voodoo tube amps had switching power supplies. IIRC they were flyback converters driven by the all-in-one "SMPS on a chip" devices. I've seen the power supply schematic posted on the forum.

If you went for the flyback topology, you have to add an air gap to the transformer. A forward converter doesn't need the gap, but it needs a separate filter inductor.

[Alan0354]

Not that I know of. The easiest way forward would be to get an ETD core assembly and wind it yourself.

The Crate Blue Voodoo tube amps had switching power supplies. IIRC they were flyback converters driven by the all-in-one "SMPS on a chip" devices. I've seen the power supply schematic posted on the forum.

If you went for the flyback topology, you have to add an air gap to the transformer. A forward converter doesn't need the gap, but it needs a separate filter inductor.

I am really not familiar with SMPS. Is the flyback converter looking like a single end power amp with one MOSFET pulling one side of the primary down and the other side of the primary connects to the +V? if so, the transformer can serve as isolation transformer at the same time. I don't know forward converter and what gap!!! I am not there yet, still have a ways to go. I have no idea what is a forward converter. I am just starting to study the transformer, in fact, since I got stuck with the RDH4, I was reviewing magnetic circuits, reluctance and flux type of basic theory. I guess since I start studying, I am going to spend a few days on the OT anyway, I think the equivalent circuit is so basic that it must be the same as SMPS transformers. A little extra knowledge don't hurt.

Thanks for you and others' advice, going to have to study the SMPS. The idea of $70+ and 10lbs+ give me a lot of motivation to learn SMPS!!!!

[Steve Conner]

SMPS are trickier than you think, that $70 PT will look like a bargain before long!

I know a guy who built a 500W Class-D bass rig from scratch, including an off-line SMPS, but last I saw him he was gigging with an old Peavey. Maybe it exploded?

[Alan0354]

I know it's not as simple, that's the reason I am studying it first. I saw my engineer did a few, it is not that hard either, particularly we build all from scratch. MOSFET, pulse circuit and stability don't scare me at all, that's where my strong suit. I let SMPS go because I was designing all kinds of HV pulsing circuits include a pulser that generate up to 150V pulse with rise time of less than 1nS, a cascade pulser that can generate a 5KV pulse with rise time in like 10 to 20nS. That was the hard one, MOSFET was what I was using on all these. I even did the stability design of the SMPS for my engineer. It is just the transformer that I never got into.

In fact how we started to design SMPS was because we had to do a lot of control that had to be float on top of 15KV or higher. It was very hard to have complicated control at that high voltage level. I came up with the idea to design a 24V to 24V DC to DC converter that can have the secondary float to 20KV and provide 100W of power up the floating deck. I even design the feedback to control the voltage by "V to F" from the HV side, driving through a long fiber optic cable and "F to V" at ground level to closed the feedback loop for regulate the output voltage. With that, we float a whole micro controller with more HV power supplies and other circuits on top of the HV deck. We did it in the mid 90s and really beat out the competitions. We wound the transformer with 25KV 18 gauge cables as secondary!!!