Re: re: Bipolar question ?MOSFETS

To: "INTERNET:tesla-at-grendel.objinc-dot-com" <tesla-at-grendel.objinc-dot-com>
Subject: Re: re: Bipolar question ?MOSFETS
From: Alan Sharp <100624.504-at-CompuServe.COM>
Date: 15 Apr 96 21:29:37 EDT
>Received: from arl-img-5-dot-compuserve-dot-com (arl-img-5-dot-compuserve-dot-com [198.4.7.5]) by uucp-1.csn-dot-net (8.6.12/8.6.12) with ESMTP id TAA02757 for <tesla-at-grendel.objinc-dot-com>; Mon, 15 Apr 1996 19:31:36 -0600

Greetings,

I have just about decided to make one last circuit on strip board, aiming for
75v,
2A. And to start making PCB's . I'm tempted just to paint them, but if I go for
photo processing, how long an exposure should I give a board on a Sun Bed!

Have I got this right. I'm planning to use a double sided board, top side purely
for ground, leaving as much copper as possible, I take it that this is what is
meant by a ground plane. And then keep the power side compact with 1cm wide
tracks for the power bearing tracks.

I've been using a capicitor/ resistor in series for snubbing, I've seen an
alternative
arrangement. However I didnot consider the turn on current, I assumed that the
load
the primary coil was inductive and therefore it would limit the turn on current.
But of
course it is a transformer, so below resonance its an inductor, above resonance
its
a capacitor and at resonance its a resistor, that explains quite a lot, now I
have to consider
delibrately putting an inductor in series with the load (with a diode in series
with a
small resistor across it). 
I reckon that at 75v and 4A, I'll need 5uH to slow the current rise over 250ns,
(14ohm
resistor) does that sound right anyone? 

Paul wrote:
>PWM Circuits are'nt really what I had in mind.
>a much simpler circuit using feedback from the primary/secondary circuit and
>a schmitt trigger to drive a FET class B type circuit.

Go for it - I'm not sure that it will be simpler though, I did look at
cannibalising
a class B, mosfet audio amp, I wasn't sure that it would behave well at higher
frequencies, or how to avoid "shoot thru". Feedback is helpful, my coil (when
it works) goes off frequency as soon as you approach it and I have to keep
adjusting it. 

>I was thinking of trying to drive the primary as a LC series circuit (not
>parallel LC) The voltage on the primary would be determined only by the
>losses in the primary and secondary circuits.  The drive current would be
>high but it might be worth trying.  A Q of ten would turn 200V drive into
>a 2000 volt primary voltage. 

I was going to try a half bridge arrangement to drive an LC series circuit.
Lots of ideas but limited time - 

The 500w switching PSU's you use at work might do the job with only a
little bit of hacking, certainly get you started and to work out the frequency
of your coil (aim for low frequency - its easier on the FET's.

have fun,
Alan Sharp (UK)