Re: SSTC, xfmr gate drive oddity .....GDIC protection, cap values, etc. SSTC

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

Original poster: "Justin Hays by way of Terry Fritz <twftesla-at-qwest-dot-net>" <pyrotrons2000-at-yahoo-dot-com>

Hi All,

(snip from Jan)
> O-oh. Mosfet datasheets say min 2V and max 4V as V_t. Yup. Now, 
> _that_really explains things!! Plus the C_gs => output coupling. 
> I hadn't checked the datasheets (surplus mosfets, so why bother 
> checking? ;) before and "assumed" it was just 1V. 

Yeah, that'd do it ; )

For sure, you want those mosfets SATURATED. Threshold voltages are
just what they are: thresholds. Barely on the edge. I suggest using
about 70 to 80% of the voltage that the mosfet gate fries Vgs(max),
to insure that those suckers are turned on. This will give you the
datasheet specified Rds(on) channel resistance, and will also give
you lots of noise immunity. 

Audio and RF linear amplifiers bias the MOSFET gate somewhere in the
middle of the upper and lower threshold voltages. (the linear
region). For us, the thing is a switch, and we need to stay as far
away from the linear region as possible. Otherwise there is a huge
amount of power dissipated in the MOSFET Rds(on) resistance as heat. 

About the gate drive IC's: I agree wholeheartedly, it would be
beneficial to make the P-ch/N-ch totem pole pair work. My best bud
used a MAX4420 to drive two IRFP460's via a gate drive
transformer....and at 4100pF Ciss a piece (ignoring Miller effect and
Coss), the gate waveform was a triangle as expected. They just can't
push that kind of current...I mean, an 8 pin DIP can only dissipate
so much power anyway.

*** Note, on MAX4420 and 4429, the input waveform must be a clean
square wave or it will fry. The input voltage must not stay between
.8V and 2.4V for more than 500nS ***

Also, we tried paralleling these and ran into problems. Still trying
that.

Make sure to use two diodes across the rail, with their middle leads
connected to the output of the gate drive IC. This clamps all voltage
spikes (leakage L in xfmr primary) to the rails. So like this:

GND ---> anode ---> cathode + output + anode ---> cathode ---> + rail

The diodes are reverse-biased unless spikes of opposite polarity come
in from the transformer, and in that case they conduct and clamp. 

Many people will suggest using shottky diodes for this
(1N5818,19,20), but I am using cheap silicon 1N4148's. Haven't fried
an IC yet, and I've blown up transformers, fried cap's, and many more
mean things!

One more quick thing, I am using 1uF coupling cap's on the
transformer primary. Their reactance is such that they are still a
dead short at my 300kHz Fo.

And for that matter, you know the coupling cap's for the primary of
the Tesla Coil?? Just for kicks, I tried used a .047uF tiny thing, it
worked just as well as the 1uF cap I had in there. 1 / 2*pi*F*C. I
suggest polypropylene or polystyrene dialectrics only.

Have a good one,

Justin Hays
KC5PNP
Email: justin-at-hvguy-dot-com
Website: www.hvguy-dot-com

__________________________________________________