[Prev][Next][Index][Thread]

Re: Ferrite materials (was Re: A CW TC without a tube + Re:








Tesla List <tesla-at-pupman-dot-com> on 12.05.99 15:36:13

To:   tesla-at-pupman-dot-com
cc:    (bcc: Marco Denicolai/MARTIS)
Subject:  Re: Ferrite materials (was Re: A CW TC without a tube + Re:




Original Poster: "Malcolm Watts" <MALCOLM-at-directorate.wnp.ac.nz>


>I know the figures look pretty frightening but realise that the
>volume of your core is << 1 cu. m.  N27 is happily used in SMPS (with
>flat-topped pulse drive) to at least 100kHz.  If you really want to
>go for the doctor, try N67. That wouldn't happen to be the 3C90
>equivalent would it?

Correct: but N67 is marked as "not recommended for new designs" and not
available for such sizes of U-cores. N87 is available and recommended and has
got lower losses than N27 but, and this is the real issue, I do have samples of
the N27. There is no idea in again ordering another sample and again start
waiting several weeks. This is just what I have done with the Philips parts and
they should be on their way by now.

>If so, that material is used in resonant
>supplies up to 1MHz. At any rate, the losses for N27 are vastly lower
>than for powdered iron in your particular application.

Of course, of course. That was just a quick test from the series "Boy, I'm
going
to do something stupid but I just can't wait to think it over". I was just to
make a 160 uH inductor and see the resonance current shapes.

>    Correct me if I'm wrong but based on your core volume, Bpk of
>100mT and a frequency of 100kHz you'd be looking at a core loss of
>about 60W with N27. That translates to about 24W with 3C90. Not bad
>for a throughput of 5kW.

I have calculated the same figures: but isn't 24W really better than 60W? The
termal resistance of my N27 U-core is 3C/W: that makes a 180C temperature
increase at 60W. Not good...:(

>Please keep us posted.
>Malcolm

Here you go.
>From the same series "I can't way no more", yesterday night I quickly
wound a 50
turn primary and a 200 turn secondary on the N27 U-cores. As I had no bobbin
formers, I used two round pieces of PVC pipe 5 cm diameter. Then I
short-circuited the secondary and checked the primary current: almost no
change!
It must have been so that the coupling was too loose. (This is my first
transformer, after all :) )

So I left my primary and wound a primary ON the same bobbin of the secondary. I
shorted again the secondary and now the current did increase: I measured 14 uH
of leakage inductance. Then I started increasing the air-gap up to removing one
of the U-cores and then both U-cores: there was a minimal variation in leakage
inductance, about 0.05 uH.

I guess this is due to the fact that the most of the magnetic field is coupled
through air and only a little part through the core: remember, I have a huge
clearance between core and bobbin. It seems to me I really need a kind of
bobbin
former, so that the copper is very near to the core surface, like in commercial
transformers. Next step will be to build a kind of bobbin former, tonight.

What do you think about the above experiments? Suggestions?

Cheers