A CW TC without a tube + Re: Solid state TC progress ?

To: tesla-at-pupman-dot-com
Subject: A CW TC without a tube + Re: Solid state TC progress ?
From: "Marco Denicolai" <Marco.Denicolai-at-tellabs.fi> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)
Date: Mon, 03 May 1999 07:21:09 -0600
Approved: twftesla-at-uswest-dot-net
Delivered-To: fixup-tesla-at-pupman-dot-com-at-fixme





Tesla List <tesla-at-pupman-dot-com> on 02.05.99 07:46:34

To:   tesla-at-pupman-dot-com
cc:    (bcc: Marco Denicolai/MARTIS)
Subject:  Re: Solid state TC progress ?




>Original Poster: Alan Sharp <AlanSharp-at-compuserve-dot-com>

>Message text written by INTERNET:tesla-at-pupman-dot-com
>>Change topology: use full-bridge and a series resonant design.<

>Thanks Marco - I've got a book!

>I've looked at this but I've got a few questions.
>How do I calculate the leakage induction?

It is really not so easy. What I am going to do is to measure it and to add a
suitable series inductance to reach the 160 uH I need (there are equation for
leakage inductance of typical winding shapes but mine is not one of them).

>Is there a voltage rise accross
>the primary?

No. But in each cycle your switch (IGBT or MOSFET) will see twice the supply
voltage (300 V -> 600 V).

>What controller chip are you using?

UC3860 from Unitrode. It is specifically for resonant mode power supplies. But
this part is not critical.

>What size of transformer
>are you using to switch 5kW at 50kHz?

Two ferrite U-cores from Philips U100/57/25-3C90. Total transformer size about
10 x 10 x 5 cm. The U-core are big but not huge. The transformer will weight
about 500 g (not kg!).

>Why does toast land butter side down?

Sorry, I don't get this question...

>I also wondered about dispensing with the output xformer all together
>and feeding the coil of a resonant LC.

Then you would get into the kingdom of CW TCs. On the other hand, it is a good
idea: that could be tried easily. Basically you (or I) don't need anything else
more than a primary, coupled secondary and top load. And a resonance frequency
of about 50 kHz...

I could keep the same capacitor I have got (20 nF) put enough extra inductance
is series with the primary to reach 160 uH, and then I could be able to
feed the
primary with pulses at 50 kHz of 560 VDC and 12 A. Secondary and top load
should
be set also for about 50 kHz (not difficult) and here you are with a 5 kW CW TC
without a tube. Did I forget something?

Should we call this the "CW TC a-la-carte" ?

>Your design sounds like a great project - do keep us posted.

I have got all the electronics ready and tested: I am just waiting for the
U-cores to assemble the transformer. You bet I'll keep the list posted!

P.S. I wrote that at 300VDC 3A my IGBTs stood cool: my 160 uH test inductor
core
instead started boiling (its coating) after about 20 seconds. It was a 33 mm
O.D. iron-powder ferrite toroid with just 40 turns of wire. Core losses
were too
big at 50 kHz. It also did change color (from green to gray!). It is still
working but now its surface is full of bubbles :)


>Alan Sharp (UK)

>>A couple of days ago I have been feeding a test inductor at 50 kHz with
>>300V DC
>>and sinusoidal current peaking at 3 A. As I switch at zero current and zero
>>voltage, my IGBTs stay cool. More, as the current shapes are all
>>sinusoidal,
>>there are no switching spikes and ringing losses as long as you switch at
>>the
>>right moment. I calculated that switching 5 kW each IGBT will dissipate
>>about 4
>>W (!). At full power I am going to use 560 VDC and 12.5 A max.

>>A resonant series design basically features a capacitor in series with the
>>transformer primary. The capacitor and the transformer leakage inductance
>>define
>>the resonance frequency and the maximum current the primary will sink.

>>You'll find more about it on swithcing power supplies textbooks.