Re: Solid State Gap

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

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

Hi Malcolm,

At 05:05 PM 3/19/2001 +1200, you wrote:
>Hi Terry,
>           Nice goin'.  That's close to what I was doing with the 
>MOSFET gap. I tried looking at the no quench trace but couldn't get 
>thru for some reason. 

The address to the site got kind of long.  Be sure you get the whole thing
in the address line of the web browser.  Maybe try again too.

>Did the beat envelope assume a linear of log 
>shape (guess = log)?  

Both the loaded and unloaded cases looked very log like.  The voltage was
only 50 volts so the turn on voltages were eating at the losses.

>BTW - have you tried cutting the gap off at Ip 
>max (hint - interesting to test one or two old theories out there - 

:-)))  I will have to try that...  I ran through the cutoff range and
didn't notice anything "odd" ;-))

>watch the spikes ;)? 

IGBT's have lower high-current losses and the fancy reverse diodes cut down
drastically on the usual spike problems.  The signals look perfectly clean
so far.  Amazingingly clean in fact.  I never could get MOSFETs to work
real well but these IGBTs seem to love this stuff.

>
>Regards,
>Malcolm
>
>On 18 Mar 01, at 20:26, Tesla list wrote:
>
>> Original poster: "Terry Fritz" <twftesla-at-uswest-dot-net>
>> 
>> Hi All,
>> 
>> I did more testing with my solid state gap tonight.  I actually have
>> it hooked up to my small coil but was only running with a primary
>> voltage of 50 volts so I could play with things :-))  Pictures of the
>> setup are below.
>> 
>> http://hot-streamer-dot-com/TeslaCoils/MyCoils/SolidStateCoil/2001-3-18/Se
>> tup.jpg
>> 
>> http://hot-streamer-dot-com/TeslaCoils/MyCoils/SolidStateCoil/2001-3-18/SS
>> gap.jpg
>> 
>> As this picture shows, it works (no load on top terminal, thus the
>> very long ringdown time) :-))
>> 
>> http://hot-streamer-dot-com/TeslaCoils/MyCoils/SolidStateCoil/2001-3-18/Te
>> k00007 .gif
>> 
>> The IGBT and the reverse diode seem to work perfectly well.  I looked
>> at no quench and first notch quenching:
>> 
>> http://hot-streamer-dot-com/TeslaCoils/MyCoils/SolidStateCoil/2001-3-18/No
>> Quench .gif
>> 
>> http://hot-streamer-dot-com/TeslaCoils/MyCoils/SolidStateCoil/2001-3-18/Qu
>> ench.gif
>> 
>> I was originally thinking of just not worrying with quenching and
>> shutting the IGBT of after a few hundred microseconds.  However,
>> unlike regular gaps where getting first notch quenching is really
>> hard, with this gap I just set the pulse width trim pot :-))  I was
>> using a 3pF probe (Tek 5100) in series with a 220k Ohm resistor hooked
>> to the coil's top terminal to simulate a streamer load.  The RMS
>> output was 78.72Vrms with no quenching vs. 87.03Vrms with first notch
>> quenching.  So that gives 10.6% more power. That 10% power is also
>> going into the streamer rather than heating the IGBT which is a
>> concern.  I was worried the cheap fiber optics may not run at this
>> high of speed but all seems well.  I was also worried that the soft
>> turn off of the IGBT would be a problem.  However, it seems that once
>> the IGBT begins to turn off, it looks like too high of a resistance
>> for the energy to swing back into the primary, so it's cutoff is
>> actually quite effective as shown:
>> 
>> http://hot-streamer-dot-com/TeslaCoils/MyCoils/SolidStateCoil/2001-3-18/Te
>> k00004 .gif
>> 
>> Even though the IGBT's cutoff is very slow, almost perfect quenching
>> is still achieved.  That is an unexpected bonus!  Although, it makes
>> perfect sense once one see what is going on.  The turn on time is
>> excellent so not problems there.  One can even easily slow that down
>> to help reduce that nasty initial pulse that causes RFI if it is still
>> there with this type of gap. 
>> 
>> So It looks like the IGBT gap has the ability to transfer the primary
>> energy into the secondary and quench very easily.  Of course, this was
>> only about 50 volts with a little over 1 amp peak in the primary. 
>> There is still a little way to go before I get to 13kV and 400A peak
>> the small coil normally runs at...  Right now that would run about
>> $650 but perhaps high voltage/current IGBTs would be cheaper once all
>> the bugs are worked out.
>> 
>> Cheers,
>> 
>>  Terry
>> 
>> 
>> 
>> 
>
>