Maker Pro
Maker Pro

800+ Watt DC-DC converter ferrite transformer design

R

RobertMacy

Jan 1, 1970
0
You can of course apply the Litz trick to foil, but you don't have any
free lunch; the geometry reduction is still required whether putting
together a bunch of strips or strands. Ten strips woven together will
have less resistance than a single strip of the same width and
thickness, but higher than 1/10th of an isolated strip that size. I
know of at least one company that claims to have some sort of foil
technology that reduces Rac like Litz, presumably doing some kind of
weave. Tempting to buy a bigass custom part from them just to take it
apart and look, see how they put the stuff together. I can't imagine
it's all that easy to make, considering there are only two US companies
making the round stuff as is.

Tim

Somewhere there is an IEEE paper [I think] showing how multistranded cable
is a 'poor' man's Litz wire. The better performance of multistrand versus
solid is attributed to the poor 'cross-conductivity' between strands. It
seems the stranding vectorizes the conductivity. From memory they
presented data comparing losses between solid and stranded with the same
copper cross section.

Very interesting article, because I had always thought that solid versus
stranded didn't matter much. But, the paper is good news since it's easier
to wrap 10 strands of 18 Awg than a single strand of 8Awg. The only
'overhead' appears to be similar to stacking factor - in that you can't
completely fill the winding area.
 
B

Bill Sloman

Jan 1, 1970
0
Strange...

Note that anything with 7 bundles is 14% useless: you get six around one central core, which never moves out from the center and therefore exhibitshugely greater resistance than the others. I've only ever seen 3 and 5xbundles from NEWT, but I've seen Chinese stuff that's 7-way before. Doesn't make sense why anyone would make it that way, unless they simply didn't know how to do it properly.

They don't have to simply twist the seven bundles in one hit. If you twisted a three strand bundle and a separate four strand bundle and then twisted those two bundles together you'd get a rather better effect, if less good packing.
 
N

Neon John

Jan 1, 1970
0
Somewhere there is an IEEE paper [I think] showing how multistranded cable
is a 'poor' man's Litz wire. The better performance of multistrand versus
solid is attributed to the poor 'cross-conductivity' between strands. It
seems the stranding vectorizes the conductivity. From memory they
presented data comparing losses between solid and stranded with the same
copper cross section.

I have that paper or one very similar. In it the author was testing
the difference between Litz and finely stranded welding cable. His
claim was that the oxide coating on the individual strands was enough
to eliminate adjacent eddy current flow.

Sounded reasonable so I tested the theory at 80kHz and about 60 amps
using a Roy induction heater. 60 amps should not significantly heat
$4 wire and in fact, the Litz wire the transformer is wound with
stayed quite cool, as did the 4 ft long extension cables that I made
for the test. The welding cable, OTOH, got too hot to touch. I don't
recall the exact figures but it was in the 160 deg range when
equilibrium was reached.

So I set up to enhance the oxide theory. I made a fixture so that I
could flow oxygen into one end of the welding cable and let it escape
the other end. Then the pieces were placed in an oven and baked at
400 deg F for several days.

At the end of the run, cutting into the insulation showed the strands
to have a nice uniform dull oxide coating. Unfortunately, it only
made about a 10% improvement over the standard welding cable.

We use welding cable as output leads on the Roy heater for cost and
durability reasons and it does get hot. Our Litz wire manufacturer
will extrude a neoprene jacket over #4 equiv Litz wire if we'll buy a
certain minimum quantity. We're about to do that.

John

John DeArmond
http://www.neon-john.com
http://www.fluxeon.com
Tellico Plains, Occupied TN
See website for email address
 
R

RobertMacy

Jan 1, 1970
0
John

John DeArmond
http://www.neon-john.com
http://www.fluxeon.com
Tellico Plains, Occupied TN
See website for email address

was that F, or C hope it was F 71C is hot enough, I can't even hold onto
50C items if more than a few pounds.

Yes, think that was the paper. The premise sounded so good always wondered
if made much difference. Glad you went to the effort. perhaps the cable
was too tight and the O2 couldn't get down in there. Like a gas seal from
the pressure. Wonder what would have happened *if* you could have used
pre-dulled copper strands THEN spun into a cable.
 
T

Tim Williams

Jan 1, 1970
0
P E Schoen said:
I think I see what you mean - sort of an idealized pot core where the
magnetic material totally encloses the coil. ...

Yes! That. It can be so hard to describe things sometimes...
I can't quite grasp the concept, other than it seems like a rod core
coupled inductor.

Well, if it were a rod, it wouldn't be folded over. Consider the
"idealized wound pot core", but run a cutting wheel around the periphery
so you get strands splayed out, instead of a continuous winding. Then
clamp the strands back into place. That's roughly what Swinburne had, at
least in my mind. Just with longer strips so the ends overlap, which will
help significantly with airgap.
I designed a rather interesting high current toroidal transformer, using
four 1.4 kVA cores and four bus bar turns at 90 degrees, with bent bus to
connect to stab plates for high current circuit breaker testing:
http://enginuitysystems.com/pix/PI1000X-1.JPG
http://enginuitysystems.com/pix/PI1000X-02.JPG

Tim Allen grunt "Ohh Ohh Ohh!" :)

Only thing better would be using solid copper pipe for the innie and
outie. And brazed to end plates. Coaxial winding. That much meat would
probably be good for continuous duty, heh.

Coaxial windings on toroids work great at high frequencies, too; easy to
make a ferrite transformer of moderate size (as ferrite transformers go,
not as breaker testers go :) ) with bandwidth 50MHz and up, with high
turns ratio.

Tim
 
T

Tim Williams

Jan 1, 1970
0
Bill Sloman said:
They don't have to simply twist the seven bundles in one hit. If you
twisted
a three strand bundle and a separate four strand bundle and then twisted
those two bundles together you'd get a rather better effect, if less
good
packing.

Could, but you'd have to insist that they do it that way, and you'd be
paying extra for it. Litz is built in batches, so they'd have to go out
of their way to make the 3x, and seperately make the 4x. Then as the
final step, braid those to make the 2x. Which might not even be possible
because of the asymmetry.

Most of NEWT's catalog stuff shows multipliers of 3 and 5, on all bundles
except the very first bundle, which might be 3 to 50+ strands, twisted.
The second bundle (first braid) might be 7x, but I don't think they
recommend any subsequent braids be 7x or more. Really big cable, in the
10k strand range, looks like 5x5x5x5x16. They just use more steps to get
bigger cable.

Tim
 
N

Neon John

Jan 1, 1970
0
was that F, or C hope it was F 71C is hot enough, I can't even hold onto
50C items if more than a few pounds.

yes, deg F.
Yes, think that was the paper. The premise sounded so good always wondered
if made much difference. Glad you went to the effort. perhaps the cable
was too tight and the O2 couldn't get down in there. Like a gas seal from
the pressure. Wonder what would have happened *if* you could have used
pre-dulled copper strands THEN spun into a cable.

Dunno. Copper oxide is a semiconductor (re: copper oxide rectifiers)
so I suspect that it just wouldn't work very well.

I never had an appreciation of just how good Litz wire is at these
moderate frequencies until I got involved in induction heater design.

John

John DeArmond
http://www.neon-john.com
http://www.fluxeon.com
Tellico Plains, Occupied TN
See website for email address
 
Top