Re: OLTC Questions

["Tesla list" <tesla-at-pupman-dot-com>]

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

Hi,

A single IGBT from Digikey would do it.  I use the mighty IRG4PF50WD:

http://hot-streamer-dot-com/temp/irg4pf50wd.pdf

I have indeed run 750 amps through 'one' of these:

http://hot-streamer-dot-com/temp/OLTC08-30-01.jpg

http://hot-streamer-dot-com/temp/OLTC08-30-02.gif

If you look closely at the last picture, you will see that I didn't bother 
to turn it off at the zero crossing either :o))  You do need to run high 
gate drive voltages for that much current.  Like 25 volts...  I was only 
running about 16 at first:

http://hot-streamer-dot-com/temp/OLTC08-30-03.gif

But I put in more batteries to get to 750 amps.

You could always use two IGBTs too and be within their ratings.  They are 
pretty cheap ($8.89).  I still have yet to blow an IGBT, much to the 
chagrin of Marco and Ken H. ;o))

My cardboard form was very lossy which really hurt performance.  However, 
as it dried out over weeks it became far far better.  Don't use cardboard ;-)

You also want high coupling to transfer energy to the secondary as fast as 
possible so the losses can't steal much power.  I could get to 
0.25.  Higher than that and primary to secondary and racing arcs started 
(big ones!!)

Cheers,

         Terry



At 12:58 PM 1/20/2003 -0800, you wrote:

>The IGBTs that Terry used worked at currents up to 750 amps, but were 
>rated at (I think) 204 amps. you would probably be fine with 1 IGBT.
>
>140pf is a very large topload. With 400 amps through 3.5 uh, the toroid 
>voltage wont pass 65kv. If there are going to be any streamers, you better 
>have a sharp breakout point. I would go with a higher turns ratio, I think 
>65kv would be too low. If you have to go to really thin wire to get enough 
>inductance, you might want to put multiple layers of wire on the secondary 
>( in paralell, of course) to lower the wire loss. You dont want to have 
>the same problem as Terry's OLTC. Terry, do you know how much of the 
>secondary loss is the wire and how much is the form?
>
>Higher break rates tend to produce shorter sparks. I think I remember John 
>Freau saying it took about 20% more power at 240bps than 120bps for the 
>same spark length.
>
>  Tesla list <tesla-at-pupman-dot-com> wrote:
>Original poster: "Stephen Conner by way of Terry Fritz "
>
>At 21:57 19/01/03 -0700, you wrote:
> >Original poster: "Terry Fritz"
> >
> >Details on this are at:
>
>(snip)
>
>OK, thanks for the information. I put together a basic schematic of my
>proposed OLTC. Note: I left out EMI line filters etc and I may need to
>change the value of the charging reactor.
>
>http://homepages.strath.ac.uk/~cnbp111/oltcschem.gif
>
>It's basically a small DC Resonant charging tesla coil with the spark gap
>replaced by an IGBT. Rather than triggering the 'gap' as soon as the
>charging current hits zero, I used a de-Qing diode like in normal DC coils,
>so I can run at any break rate I want. Does anyone have any thoughts on the
>validity of this approach?
>
>The main issue is that the primary ! current is going to hit 400A peak but
>the biggest IGBT I can find just now (Infineon part# BUP314) has a pulse
>rating of 160A. So I have to either parallel three IGBTs (using a separate
>primary coil/tank cap/gate drive for each one to ensure sharing) or use a
>tank cap one-third the size and then go up to 1200bps to get the same power
>as before. Can anyone comment on whether increasing the BPS would impact
>the spark length even if the power level stays the same? I.e. do a few
>large bangs give a bigger spark than many little ones?
>
>Your thoughts much appreciated
>
>Steve C.
>
>
>
>
>
>Jimmy