Ferrite rod, side by side or in line

Klaus Kragelund wrote in news: snipped-for-privacy@googlegroups.com:

By this you mean turns ratio over coupled voltage efficiency?

I am sure also you are reading it no load. Efficiency drops severely when you actually try to drive a load.

The key is the complete magnetic circuit.

Those core half faces are what need to be as close together as possible. So those need to be at your 3mm spacing to take your readings and do your characterizations from.

There are long, flat cores that are pre split and faces polished that your windings could go on without much profile height rise.

Like little sticks of gum with little blocks on the ends.

I have it working, just need to reduce cost if possible

Efficiency about 60% is easy, I think for wireless power at limited supply load that is quite good

I need < 5W

I don't care, spend your time asking the Chinese or Americans to reduce their carbon footprint, not my small power supply

I have a load of about 1W, good efficiency

Cheers

Klaus

Makes it difficult to design their stuff in, if you can't get hold of a han dful to build a prototype or two.

Perhaps the business I was in was lower volume than you seem to be used to. We did know - and sometimes used - some specialised suppliers for ferrite parts, but everybody was much happier if you designed in parts that could b e bought off the shelf from broad-line distributors

ck

Maybe you could go back to telling your grandmother how to suck eggs. She h as to put up with you - I don't.

Klaus Kragelund wrote in news: snipped-for-privacy@googlegroups.com:

Then your solution is a pancake flat pot core pair and use hand wound "(scatter wound is fine) flat form windings that spiral in and back out. epoxy stripes lock it in place and the pot core form factor encapsulates the stray flux.

All you need are physical impressions and steps on the mating parts to insure that correct alignment ocuurs.

You can do it core free with the pancake winds, just a lot less efficient. Then misalignments are less a problem and cost goes way down.

Your dollar figure for what you want this to add to the cost of your product is very low. ANY configuration will likely have a higher cost than that which you are projecting in this thread.

snipped-for-privacy@ieee.org wrote in news: snipped-for-privacy@googlegroups.com:

There are almost always sample qtys that do not have a great effect on end year audits on hand. Grow up, even your nit picks are lame.

snipped-for-privacy@ieee.org wrote in news: snipped-for-privacy@googlegroups.com:

There is typically and historically no such animal in the High Voltage Power Supply segment. Every customer and every application is typically uniquie and there are next to zero off-the-shelf supplies around. The company I worked for started some small form factor, PCB mountable modules for 4kV and less, but real supplies like a triple output anode supply for the huge three gun projectors that used to be used on the 747s are quite custom. As are those that go up on weather balloons bought by the NOAA. and those in use on Geiger counters by the Navy nukes fleet wide.

Essentially each design must be made at the time of the order as the customer design, load conditions, and form factor requisites are always different.

I take it the barrier is already in place? The barrier is usually between the coil and the core, no?

Rick C.

The barrier is between the two coils

Like this:

Huh, that was going to be my suggestion. Two e-cores or something.

Maybe someone makes them bonded to some metal that can be soldered down?

George H.

Or some clip?

(random search for planar e-cores) GH

Bill,

If you would so kindly, stop feeding the trolls.

Thanks, Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Design Website:

Makes it difficult to design their stuff in, if you can't get hold of a handful to build a prototype or two.

Perhaps the business I was in was lower volume than you seem to be used to. We did know - and sometimes used - some specialised suppliers for ferrite parts, but everybody was much happier if you designed in parts that could be bought off the shelf from broad-line distributors

Maybe you could go back to telling your grandmother how to suck eggs. She has to put up with you - I don't.

-- Bill Sloman, Sydney

The question here is, what is the application ?. If you just need isolation (plastic barrier), then there may be better ways of doing that other than a transformer. If you must use coils, then a flat pancake coil at each side of the plastic barrier, both tuned to resonance, sounds like a possible way to go. Also the best way if you want to transfer power, as in a wireless charger...

Chris

In terms of regulatory compliance, a design with half-pot-core on each side would be somewhat self-shielding, and would have higher inductance (so you can use lower frequency and presumably get better Q). It'd be hard to stack a high set of flat pancake windings to match a bobbin in half a pot core. Pot cores, though, are likely not a low-cost winner.

It's possible to use a pot core in the power-source side, with pancake receiver windings, if the economics (one source, many receivers) work.

Pot core halves are brittle; potting or molding would be a good precaution.

I was just thinking that a pancake format would maximise transfer area across the effective 3mm airgap. There's also other considerations, such as what frequency to use, rfi shielding, though running both coils at resonance could minimise harmonics. What frequency ?. Guess anything from 100KHz upwards of 1 or 2 Mhz as a first stab, but it would need some labwork to see what's feasible in terms of power transfer, efficiency and any heating effects. Ferrites may help, say, flat sheet ferrite, would be cheaper that pot cores, where only the larger sizes could provide a decent transfer area across the gap...

Chris

Right, I was thinking two (planar) e-cores would be sorta similar. (Klaus is a clip cheaper than glue? I have no knowledge of high volume stuff, but it's interesting.... maybe you don't have enough area for a planar thing.)

George h.

End-to-end, but broad and flat - you need to increase the diameter from rods up to flat slabs, like pancake coils. Leakage will come mostly around the periphery, so increasing the diameter means a higher percentage is far from the periphery.

Never came across any myself.

Small, thin markets tend to take a while to sort out standardised modules.

Geiger counters verge on a mass market item. Military specs tend to be written to exclude using mass market components - it's pork barrel politics in action

Until the economic advantage of exploiting a common form factor trumps customer wilfulness.

Electron microscopes include a high voltage power supply that typically gets cranked up to 30kV to drive the electron beam. As far as I can remember, we bought them in, along with the 1kV supplies for the photomultiplier tube in the electron detector.

There was clearly an off-the-shelf market for the high-voltage connectors that hooked them up.

Grow up. That kind of sample quantity has to be negotiated out of the sales representatives.

Buying stuff from a broad line distributor is a lot quicker, and pretty much guarantees that the stuff will be there for purchasing to just-in-time-buy when manufacturing needs it. You seem to have lead a sheltered life.