Re: PLL-controlled SSTC?

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Jim Lux by way of Terry Fritz <twftesla-at-qwest-dot-net>" <jimlux-at-earthlink-dot-net>

Given the frequencies we are operating at (few hundred kHz), some sort of
DSP based loop might actually be the best approach.  Run the secondary base
current into a current transformer then a fast A/D, then implement the
tracking/filter/etc, and finally waveform generate the drive signals to the
FETs using the DSP (you could probably use the onchip PWM circuits).   Then,
putting in weird phase shift vs frequency curves and dynamic operating point
tracking, while non-trivial, is at least feasible, something you probably
can't do in straight analog circuitry.

You'd want some sort of "rock solid" FET driver that could take the CMOS/TTL
kinds of swings.  Maybe one of the PWM motor drive parts from IR?

Over the years I've built a variety of analog,digital, and hybrid motor
servo and stepper control loops.  At first, the analog PLL seems simplest,
but when you start adding a little circuit to deal with the noise from the
encoder, then another little circuit to handle the loop stability on
acceleration, then another little circuit to handle the switching transients
kicking back from the motor inductance, etc.etc.etc. all of a sudden you get
to a point where you say, I should have "bit the bullet" from the beginning
and just implemented something with enough "computational power" to do it
all.  At least, changing software doesn't require breadboarding, white wires
on PC boards, etc.

You could build the PC board ONCE. Given all the RFI/EMI/EMC issues with a
SSTC, you're going to need a PC board anyway.... random wired protoboards
just aren't going to hack it, and "deadbug" construction works ok(actually
great) for low level stuff, but when you handle serious power, you need
those big wide traces and attention to current paths..

So who's going to pioneer the "universal SSTC FET driver kit" business...


----- Original Message -----
From: "Tesla list" <tesla-at-pupman-dot-com>
To: <tesla-at-pupman-dot-com>
Sent: Saturday, July 06, 2002 5:59 AM
Subject: Re: PLL-controlled SSTC?


> Original poster: "Jolyon Vater Cox by way of Terry Fritz
<twftesla-at-qwest-dot-net>" <jolyon-at-vatercox.freeserve.co.uk>
>
> Would it not be possible to use a some means (eg. a current transformer or
> feedback winding) to detect the change in the resonant frequency, then
feed
> the signal into the input terminal of the PLL so that the PLL's internal
VCO
> could track the secondary frequency?
>
> ----- Original Message -----
> From: "Tesla list" <tesla-at-pupman-dot-com>
> To: <tesla-at-pupman-dot-com>
> Sent: Saturday, July 06, 2002 2:19 AM
> Subject: Re: PLL-controlled SSTC?
>
>
> > Original poster: "Daniel McCauley by way of Terry Fritz
> <twftesla-at-qwest-dot-net>" <dhmccauley-at-spacecatlighting-dot-com>
> >
> > After some thinking. . .
> >
> > Actually a PLL circuit really wouldn't work in this case.  The problem
is
> > that when a arc streamer propogates it actually changes the resonant
> > frequency of the secondary and this ultimately kills the propogation of
> the
> > streamer.  The output frequency of the oscillator doesn't really change
> much
> > during such a propogation.
> > A PLL on the output oscillator really wouldn't do anything since the
> > frequency is relatively constant.
> >
> > What would be needed is some kind of feedback to detect when the output
> > current begins to drop or something similar.
> > Richie Burnett uses a simple technique to simply feed back some of the
> > output voltage back into the feedback of the TL494 chip (See his
website).
> > Its not very elaborate, but pretty effective for what its worth.
> >
> > Dan
> >
> >
> >
> >
> >
> >
>
>
>