Re: It works! Solid state driver

From: 	Alan Sharp[SMTP:100624.504-at-compuserve-dot-com]
Sent: 	Monday, December 08, 1997 4:12 PM
To: 	INTERNET:tesla-at-pupman-dot-com
Subject: 	Re: It works! Solid state driver


>This is my thinking from working with flourescent/HPS tubes. By pulsing,
>arc dies, and voltage can go up until the next one. I would like to hear
>from someone who has observed pulse vs CW for similar power levels, and
know which is more spectacular.

Pulse - you get higher peak voltages for the same power and its noisy!
Byland reckons that 50-60Hz is optimal.

Farnell components (UK and international)
Supplies a huge range of semiconductors including all the ones listed here,
also high frequency transformers, low inductance resistors - everything
the coil formers. 

I though I would start to put together a Solid state driver fact sheets -
additions and corrections welcome.


HIP5500IP   Combined PWM chip and Half bridge driver, 500v max
I'm putting together a design around this just now

HIP2100IB     Half bridge 2A 100v driver
HIP4082         Full bridge driver  80V
HIP2500IP    Half Bridge 500v driver
Haven't played with these - the snag is a slow response time -
400ns to respond to a short circuit.
Harris are on the net - great source of data sheets and application

Telcom Semiconductor
TC4421cpa  9A inverting high speed Mosfet driver.
TC4422cpa  9A non-inverting as above
TC4427cpa non inverting dual high speed Mosfet driver
all very fast.

PWM chips

SG3524   200kHz max - low capacity output. cheap and cheerful!
This will run as low as 25Hz with a 0-90% duty cycle -
use a timing capacitor of 0.47uF - ideal for the interupter.
(The circuit is given in the data sheet)

High speed, 1MHz, will drive FET's directly or through buffer chips.
A bit unstable - look at the data sheet.

Welcome to power transistor genocide country.
A source of smoke and a gate unhinged is a FET down the drain.

IRF540  100V   
IRF640  200V
IRF740  400V
TO 220 packages - fast and capable.
The bigger the FET the bigger the gate capacitance.

To avoid complaints from the FET presevation society.
Experiment in a sound proof basement so no one can hear them

These notes may help.

Things I learnt after blowing a couple of dozen FETs.

To disable the internal reverse diode in the FET (for long life
and happiness) put a beefy schottky diode in series with
the FET and then a reversed fast rectifier diode across them.

Have a fast current limiter - that can interupt the duty cycle.

Snubbers are essential.
As Hari pointed out to me V=L di/dt  (calculus :)
di/dt is say 40A switched off in 24ns, a very big number.
L is all the stray induction, a little number.
A little number times a very big number is a big number  (advanced calculus
So V could be big - could be goodbye FET.

Another implication of the above is to keep wire lengths short.
Make the design as compact as possible.

Resistors in the snubbers and especialy any low value resistor used
for current sensing must be low inductance types. Wire wound implies
a coil. (Dah)

Use a current inrush limiter across the mains to avoid the switch on
thump from big mains transformers. (Farnell Components lists these
with the thermistors)

Push pull is like a see-saw. The fulcrum is the centre tap to
+ive. Now pull one end down to the ground and the other end
goes up! So design snubbers and choose FETs that can
comfortably handle double the supply voltage.

Think about transformer saturation
More voltage - more primary turns, lower frequency more
primary turns

Good news - its heavenly once it works properly - I haven't
blown any fets for a year - due to withdrawal symptoms I'm
now messing with another design.

The alternative

Welcome to FET Hell

To really burn those FETs and your money - 
Instead of using an output transformer to bottom feed (there
are worse things than _swallowing_ a barium meal) a coil.
The magnifier arrangement. Drive a conventional primart
directly coil instead.
You can now blow FET's in two exciting ways:

1) The primary induces voltage in the secondary.
The voltage on the secondary rises through resonance.
The secondary induces a voltage rise in the primary.
My FETs didn't just smoke - they burned as well.

Very exciting:
2) The secondary has to be closely coupled to
the primary. So close that the sparks hit it.
This was the eventual (well fairly rapid) fate of Mark Bartons
"sheet of fire" mosfet driver. 

Have fun,

Alan Sharp.