Re: An Economical Magnecharger (inductive charger)

[Lee Hart <leeahart@earthlink.net>]

Arthur Matteson wrote:
> If I were doing this project, I'd make it a high switching frequency.
  
Yes; but others have already done it this way. We have working examples
to show the consequences; what it costs, and how it works. To match the
isolation, efficiency, and power factor of a constant-voltage type power
supply, a switcher is significantly more expensive and less reliable.

> [a switcher] allows sources besides utilities - such as solar panels -
> to be connected up, without any efficiency loss.

This is true. But if you have a DC source, wouldn't it be more efficient
to either use it directly for battery charging, or with a non-isolated
DC/DC converter?

> The transformer will definitely be more efficient at a higher
> frequency

A 1500 watt 60 Hz transformer is around 95% efficient; and a constant
voltage transformer around 90%. Rectifiers cost another 1-5% depending
on voltage.

A high frequency transformer can be slightly more efficient, but hardly
enough to matter. The switching transistors and rectifiers needed to
make the high frequency AC make it less efficient overall. 80-90%
efficiency is about all you can get out of a high frequency isolated
switcher.

> I don't think the cost would be too much more with a switcher.

Look at cost per watt of any of the thousands of switchers on the
market. A 1500 watt isolated PFC switcher is around $750 for a
made-in-China special; higher quality units are much more.

> I don't think you would be able to build an I and E out of these,
> since they would have to be perfectly aligned when put together,
> so they wouldn't short out. From the transformer that I took apart,
> the E's and I's were staggered.  The entire reason for having
> the laminated layers is to isolate them.

Laminations are given an oxide or varnish coating to electrically
isolate them from each other. That way, they don't short together when
you stack them up. There is less than a volt per turn at 60 Hz, so this
layer can be microscopically thin and still provide isolation.

If the ends of a core are machined, then they have to be re-insulated
from each other (can't have little smears or burrs short them together).
But you don't need to insulate them from the mating lamination's edges.

The E's and I's are interleaved on a normal transformer so the only gap
between them is the thickness of the insulating coating. Plus, the gap
is spread over the larger area of the overlap, rather than just the
mating edges.

>> In the PFC20s we are stuffing 5+ kw thru a 4 inch Torroidal core.
>> It's even smaller than the Avcon receptical. With HF we could get
>> this thing pretty darn small.

> Oh, definitely.  With ferrites you can choose between a few types of
> cores - you could cut a toroid in half, to acheive lower EMI, or you
> could just get two U or E cores and slide them together.

It's not going "thru" the core; the PFC20 is no isolated. The core needs
to be about twice the size for isolation.

With a switcher, you always need to keep leakage inductance in mind. If
it is more than a few percent, you have hideous voltage spikes, huge
snubbing losses, and unacceptable EMI. As switching frequency goes up
and your main inductance goes down, it gets harder and harder to get the
leakage inductance low enough. You start having to worry about
nanohenries. You get caught between needing the primary and secondary
within thousandths of an inch of each other, and the need to have
thousands of volts worth of insulation for safety and isolation.

You can't build a practical switching transformer with the primary and
secondary on different legs of a U or EI core. Not enough coupling. With
a toroid, you have to distribute the primary and secondary over the
entire core to get coupling, making them inseperable.

At 60 Hz, these problems vanish. You're driving it with a sinewave, so
there is no EMI. The windings have henries of inductance, so even
millihenries of leakage inductance is not important.

> That transformer sure would be heavy...even five pounds might be
> tricky to line up to the car for some people.

I'm guessing that 5 pounds is about the same weight as an ordinary gas
pump hose and nozzle. If these are acceptable, so should the 60 Hz
paddle.
-- 
Lee A. Hart                Ring the bells that still can ring
814 8th Ave. N.            Forget your perfect offering
Sartell, MN 56377 USA      There is a crack in everything
leeahart_at_earthlink.net  That's how the light gets in - Leonard Cohen