[Prev][Next][Index][Thread]
Re: Solid-state tesla circuit
On Tue, 24 Sep 1996, Tesla List wrote:
> I looked at the circuits in Bylunds book - if you can get the PSU that he used
> and modify it then you might want to go for it.
That is a good alternate. PSU manufacturers have usually made things
right :)
> Otherwise use MOSFET's - IRF740 400v, 20A 100W fast switching are ideal.
u-hum.. IRF740 is rated for 10A rms..
> Much faster switching, and less heat than bipolars.
Depends on frequency. Ohmic losses of FETs are pretty large. Bipolars
certainly beat them at low freq range (say 20kHz or so). Beware of
voltage transients (spikes) with fast-switching devices!
> Rob and I both ended up with the uc3825 pwm chip - get the data sheet.
What a coincidence. I am just planning to make my first self-tuning
resonant driver for coiling. In principle: I'm intending to use
3825 in a wierd setup to get a nice zero-voltage transition PSU.
Detecting the current will enable self -tuning operation (like ZCS
or quasi zero current switching PSU's do). Why not use some better
chip for the job? Well.. I would but no nice chips are available
over here :(
> uc3825 driving a pair of 4422 buffers to 3 * irf740 on each side.
> Current limiting monitored across a 0.05 ohm resistor -
> this creates weird feedbacks - I work around it but its something I must
> sort out in the Mark 2.
Define wierd. Perhaps transformer is getting into some nasty resonances
becouse of parasitics? I have had a lot problems with that! Especially
at high turns ratio (say 15:1 :) when trying to also keep the power
factor good..
--
"If the troubles are a bit less sever (say, someone merely hit your keyboard
with an axe .." - Linux System Administrators guide V1.0
Harri.Suomalainen-at-hut.fi - PGP key available by fingering haba-at-alpha.hut.fi