[Prev][Next][Index][Thread]
Re: Solid state design
>hours needed to make one didn't really seem all that exciting. When you
>actually calculate the impeadance of a 12µH inductor (typical helical pri) at
>400kHz it is actually quite high, about 30ohms. So if I use a 300V H-bridge
>drive (IRF3415 FETs) I can put 10A throught it, well within the capability of
So, you put 10A to it if coil draws no power. That is just the magnetizing
power the primary draws.
>the fets and a nice round 3kW power input. I am using the SG3524 pwm chip
3kW power drawn by coil would mean 10A power at 300Vrms voltage. Add
another 10A for magnetizing inductance and you get 20A current draw. Not
too efficient, I think.
Do also notice that if you have 300V bus H-Bridge the *peak* voltage is 300V.
However, as high-Q coil draws sinuoidal current at it's resonant frequency
only the RMS voltage is 300V/sqrt(2). So, approximately 200V RMS voltage.
This means more current.
>respectively. A question for all you smpsu designers is how safe is it to
>just rectify 240V mains and run this off that?
Safe enough.
>Should I use an iso
>transformer?
I do. It makes magnetizing inductance lot higher ie. lot less current draw.
> I also wondered if anyone had any ideas on my choice of pwm
>chip (does the 3524 run at 400kHz? - the data sheets are taking a long time
>to get here)? Is any of this really stupid?
3524 is definately out. It runs at up to around 250kHz. Choose another chip
like UC3825 for the job.
The idea is not stupid. Primary could be used if the frequency was high enough
and inductance high enough. Using a separate step-up transformer and
end-feeding
the coil directly from step-up transformer makes task a lot easier usually.
Do not try that as you're first SMPS project! It is not nice for that!
Harri