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

Re: 1-turn coil inductance nomogram



Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>

Hi Ken,

At 03:19 PM 8/26/2002 -0700, you wrote:
>I don't know if this will be new to the List or not but I'll prevail on
>Terry to put up http://hot-streamer-dot-com/temp/Loop-L.gif.  It returns
>inductance of a 1-turn loop using the loop- and conductor-diameters.  I
>found it in one of my ancient textbooks.
>
>It prompts me to notice that the inductance of a 12"-dia. x 1"-tube
>1-turn primary is just about 0.5 uH (and no doubt Terry reported
>something like that for his OLTC).  That yields ~0.44 ohms reactance at
>140 KHz and if I were to put 640 V across that, I'd need to switch 1450 A
>in a reconfigured s.s. primary, thru a bunch of MOSFETs (assuming much
>less resistance in the circuit).  So there's why I've gone to 3 turns at
>present, suffering the consequences.

I have three 13.75 inch loops in parallel of 1/2 inch tubing.  The chart
gives 0.7uH for one.  Having three in parallel seems to have cut that
inductance almost in half!  I was starting to think have three loops was
silly but that is what got the frequency so high which is very nice. I
wonder if some simple way could be found to add tuning to such a primary.
Tuning these OLTCs is a bit tricky.  I imagine a lot of proximity affects
are significant too.

>
>1/4" dia. tubing only increases the inductance to ~0.75 uH, resulting in
>967 A.  
>
>IRFP460A MOSFETs are rated at 85 A Idm -at- 25 C.  IXYS VMO 60-05F's, on the
>other hand, are rated at 240 A Idm -at- 25 C.  The 460As are $2.80 each/25
>(from Fortune) or 3.3¢ per Idm Amp; and I recall the IXYS being quoted
>recently at about $40 a pop (unfortunate word but there it is) or 17¢ per
>Amp.

My TO-247 package IGBTs do 204 amps peak ($7) and there is another 600V
version that does 280 amps peak ($13).  I am just not sure you can switch
them fast enough for your application.  IGBTs "can" be forced to switch
fast but the "shoot through" or Ccg capacitance can catch up to you and
blow the gate.  This is what Marco reported so much trouble with.  A sudden
step in voltage to the collector jumps the gate voltage and boom!  I just
slowed it way down in my case since I have some much time to turn on at
120Hz ;-) 

>
>Idm is the maximum for a 10 us pulse width, being limited by the
>junction-temperature rise during that interval, so it's probably a safe
>rating to use in choosing the quantity of transistors in this
>application, perhaps holding the spark duty-cycle to a modest value.  

My next project is to make a simple LC circuit to test the pulse current to
destruction.  It may be that "cold" IGBTs can take some pretty high pulse
currents.  Some folks I have talked to thought that the time and duty cycle
I was using "may" allow higher pulse currents.  That would be great if
true! ;-))

>
>Hmmm...using the 1" tubing, 1450/85 x 2 x 4 = 136, 460As are needed vs.
>1450/240 x 2 x 4 = 48, 60-05Fs.  
>
>Hmmm...did I hear someone mention IGBTs?  Which ones are going to turn on
>& off nicely in no more than 1 us, I wonder?  

There are driver chips that could force that and maybe transorbs on the
gate could stop shoot through.  I am not sure if the IGBT would start to
get really lossy though...

Cheers,

	Terry
 

>
>Ken Herrick
>