[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Pushing a TL494
Original poster: Shad Henderson <shenderson@xxxxxxxxxxxxxxxx>
Hi All,
For what it's worth, any time I get "funny" output on a chip that looks
"almost right, but not quite", I throw some capacitance across VCC and
ground and see if that cleans it up. When working with exposed
electronics (non-shielded) around high frequency, I sometimes find that
a parallel combination of a 10-20uF tantalum and a small ceramic disc
(1nF or so) between VCC and ground *at the chip* helps soak up some of
the crud.
Keep in mind that gate drive chips (TC4420, etc) benefit greatly from an
electrolytic and film cap in parallel across their VCC and ground pins.
The film provides the quick-burst of grunt for the output, and the
'lytic prevents the voltage at the chip from sagging too badly.
Things to watch for -
If you're bypassing the TL494 and driver chips with fair amounts of C,
then ensure that the inhibit line of the TL494 is referenced to the
incoming 12v line. That way when you turn the switch off, it stops the
TL494 from running.
I found out the hard way, that if you don't, the caps on the 494 and fet
driver chips (TC4420's in my case) continue to run for a moment. The
falling voltage on the TC4420's ends up presenting low gate drive
voltage to the H bridge (which was still powered... oops) and high-loss
switching ensued. Blew 2 fets, shorted the gate to the drain and the
spike fed back into the TC4420's, blowing them, the TL494, and the 556
interrupter chip.
Or you can just be sure to kill power to the mosfets' supply rails
first. :)
Hope it helps!
Shad H.
On Fri, 2007-04-06 at 00:52 -0600, Tesla list wrote:
> Original poster: "Joshua Johnson" <jjohnson98@xxxxxxxxx>
>
> Thanks for all the advice.
> Usually you get some 'engineer' margin but you just never know. Now I
> don't have to waste time testing ICs and I can get right to building SSTCs.
> Good suggestion about the caps; ICs should just start coming with a
> 10uF tantalum between Vcc and GND. I bet it would save people a lot
> of trouble.
>
> JJ
>
> On 4/5/07, Tesla list <<mailto:tesla@xxxxxxxxxx>tesla@xxxxxxxxxx> wrote:
> Original poster: Brett Miller
<<mailto:brmtesla2@xxxxxxxxx>brmtesla2@xxxxxxxxx>
>
> Steve,
>
> Awesome, that's good news for me too since I've wanted
> to try an audio modulation project with either a
> flyback driver or a small single MOSFET SSTC.
>
> Of course a flyback runs plenty slow enough for the
> 494.
>
> -Brett
>
>
> --- Tesla list <<mailto:tesla@xxxxxxxxxx>tesla@xxxxxxxxxx> wrote:
>
> > Original poster: "Steve Ward"
> <<mailto:steve.ward@xxxxxxxxx>steve.ward@xxxxxxxxx>
> >
> > Ive run the TL494 at 1MHz in the past. Im guessing
> > the upper
> > frequency limit is given either because the slow
> > output drivers (open
> > collector / open emitter, depending on
> > configuration) or because maybe
> > the deadtime control is out of spec or something
> > along those lines
> > (perhaps a slow comparator for the DTC is used
> > internally). Anyway,
> > it should work ok.
> >
> > Steve Ward
> >
> > On 4/5/07, Tesla list <<mailto:tesla@xxxxxxxxxx>tesla@xxxxxxxxxx> wrote:
> > >Original poster: "Joshua Johnson"
> > <<mailto:jjohnson98@xxxxxxxxx>jjohnson98@xxxxxxxxx>
> > >
> > >I am trying to throw together a audio modulated
> > SSTC. The
> > >secondary/topload I am going to use looks like 400
> > to 500 kHz for a
> > >resonant frequency. I was wondering if I could
> > stick with the TL494
> > >controller even though the spec sheet says it's
> > only good to 300 kHz.
> > >The Plasmasonic II setup,
> >
>
><<http://www.easternvoltageresearch.com/designfiles/plasma02_presentation.pdf>http://www.easternvoltageresearch.com/designfiles/plasma02_presentation.pdf>
> http://www.easternvoltageresearch.com/designfiles/plasma02_presentation.pdf
> > >is tested to 400 kHz with the TL494. Can I push it
> > a bit farther?
> > >Anyone run at higher frequencies with this
> > controller?
> > >
> > >I skimmed through Texas Instruments' SMPS
> > controllers that handle
> > >higher frequencies but they don't have the handy
> > Dead-Time Controller
> > >pin like the TL494 where you can just feed in the
> > audio for modulation.
> > >
> > >Any ideas much appreciated.
> > >
> > >Thanks,
> > >JJ
> > >
> > >
> >
> >
> >
> >
> >
>
>
>
>
>
____________________________________________________________________________________
> Now that's room service! Choose from over 150,000 hotels
> in 45,000 destinations on Yahoo! Travel to find your fit.
>
<http://farechase.yahoo.com/promo-generic-14795097>http://farechase.yahoo.com/promo-generic-14795097
>
>
>
>
>