NEW!! FETCOIL II, half-bridge design, no more kickback

problems...even longer sparks!  Click here!

Hi Folks...thanks for dropping by.  This is the business end

of my new SSTC (solid-state Tesla coil) putting out a brush

 discharge (about 6.5" max) at about 100 W in.  The power

 supply is just the mains voltage, half-wave rectified and

filtered with several small electrolytic caps.  For testing, I run

the thing from a variac.  At this setting I am putting in about

90 VAC.  The topload is a pan lid, the secondary is wound

with 30 AWG magnet wire and has 8.5 " of winding on a

~4.4" sewer / drain PVC pipe.

Here is the FET and accessories in the drive circuit.  The two

black devices bolted to the aluminum heatsink are both

FETs, although only the one at the top is connected in the

circuit.  This is an NTP12N50 (12 A, 500 V) from ON

Semiconductor.  The other one is an IRF740 (10 A, 400 V).

Both offer nearly identical performance, except I have lost

 a couple 740's in this circuit (hence my attempt to

 upgrade to the beefier part.)  The two diodes above and to

the left of the FETs are 3 A, 40 V schottkys in parallel

(I wasn't sure if their peak current ratings were good

enough by themselves).  Directly above the NTP12N50, and

partly obstructed by a reddish capacitor, is a fast diode

(U1560 from ON) which I believe is rated for 15 A / 500 V.  A

strip of Cu-clad board behind the FETs is a ground, and the

one behind that is the B+ (high voltage) rail.  As you can

see, I put a lot of various capacitors across B+ and ground.

All are cut out of old equipment.  Not seen are a few

electrolytic caps and the drive circuit itself, which is just a

little piece of perfboard with a 556 timer, a FET driver chip,

some pots and whatnot.  One of the 555s in the 556 acts as

a "staccato" interrupter, enabling the other timer which is

the RF oscillator.  The gate driver chip sharpens the pulse

edges up and directly drives the FET.    The small reddish

capacitor on the above circuit that goes across the fast

diode is a 0.001 microfarad that I put there to short out very

fast spikes.  Its worth in the circuit is dubious.  Let me

know what you think.

Here are the 'scope traces of gate drive (the

square wave) and the voltage level at the FET drain.

The gate voltage was measured directly with the probe of

channel 1 on the gate.  The drain voltage was divided by a

94 kOHM resistor and a 6800 OHM resistor, in series,

going from drain to ground, and the 'scope probe for

channel 2 was connected across the 6800 OHM.  DC

supply voltage at this time was 50 volts, so with the sharp

peaks of the drain at about 7 V on the 'scope, this means

that the drain is seeing about 100 V at this time, or twice

the DC input.  Let me know if you think this is BAD (I do!)

You'll also note that the duty cycle is not 50%.  I have been

told this is bad, but the coil resonates with this type of feed.

The "on" pulses are 2 microsec. long, the "off" pulses are

only ~750 nsec or so.  Comments?

My e-mail is cwillis@guilford.edu or just make a post

on the pupman list.

 Here is another 'scope shot looking at the drain

voltage.  I have always thought that the drain voltage should be a nice square wave ideally...although

the drain current may vary sinusoidally (in the ideal case anyway).  The FET is off when the

drain voltage is high in the above traces.