Re: Terry's DRSSTC - 6000 BPS >:o)

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Jimmy Hynes <jphynes@xxxxxxxxx>

Hey guys,

What is the point of this? Why don't you just have a longer on time,
and a reasonable bps?

At 6000 peak bps, the die doesn't have much time to cool between each
bang anyway, so you might as well throw all the energy in there at
once. Since the next bang is so soon that you have to worry about
phasing, you can start to think of it as a single bang, with some
interruptions in there for some odd reason.

It would be interesting to do a test comparing this new mode of
operation to one with the same bang energy in one pulse, and see how
they compare. You might find that it doesn't even make a difference.
I'd try that before spending more time with this. It's just some code,
so it shouldn't be tough (uCs are nice, aren't they?)

If you take it to an extreme, and split the bang in enough pieces,
it's basically PWM, which is essentially lower peak power. Steve's
tests have already shown that higher peak power is better for a high
streamer length/streamer power ratio. Your scope traces show pretty
well defined bangs, so you're not really approaching that yet.

I'd also vote for 'full power' testing. There can be some weird non
linearities when it barely breaks out, so that small differences can
show a 5x increase in length. At full power, you wont be fooled by
this. As long as you're careful with what you're doing, it shouldn't
be any more stressful than 'full power' in 'normal operation'. As long
as you dont try big bangs at elevendy billion bps or something, then
you should be fine.


On 5/2/05, Tesla list <tesla@xxxxxxxxxx> wrote:
> Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>
>
> Hi Gerry,
>
> At 11:55 PM 5/1/2005, you wrote:
> >Hi Terry,
> >
> >>Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>
> >>
> >>It seems that each bang just "adds" to the next.  I don't mean like a
> >>streamer follows the path from the last bang.  It is like the T1 time is
> >>just stretched out.  The T2 times seem to get lost and the streamer
> >>builds like it was just one long T1 time.
> >
> >What is the advantage in 25 bangs in a burst vs having one bang where T1
> >is 25 times longer.
>
> After about 15 bangs, the sparks just get thicker and hotter rather than
> longer.  In my case, I am thinking having all the interruptions in there is
> helping streamer length.  If it were true CW, the streamers would only be
> like a 5 inch brush...   The interruptions seem to be the key to long
> streamer length.
>
> >I would think the thermal time constant is longer than a burst
> >period.  Could it be the smoothing effects that would have on the main bus
> >ripple.
>
> You can tell if the buss is being loaded down from the primary
> current.  Right now, at 100 amps peak, that is not an issue.
>
> >Maybe there is a current issue, if so, how long can T1 be for a single
> >shot before overcurrent is an issue???
>
> The current will "top out" at about 200 amps with a streamer load when the
> input power and streamer power are equal.  But higher BPS rates make the
> total power higher.
>
> >I guess I'm thinking if the IGBTs and controller could take it, what would
> >happen if T1 were like infinite??
>
> "Just" a CW coil...
>
>
> >>I had to keep lowering the T1 time to keep the streamers at a sane 
length!!
> >
> >When you vary something else??
>
> As I raised BPS.  I fixed the program a little so that works smoother now.
>
>
> >>I worry about where the IGBT load is at.  The current should be low and
> >>the heating should be low...
> >
> >What is the concern if the IGBT's are over current and thermal protected???
>
> They could heat "very fast" blow before the temp sensor out on the heat
> sink sees it.
>
>
> >BTW, it looks like you have 26 bangs per burst by looking at the scope
> >pictures (might be an issue when you start doing real streamer experiments)
>
> Really????  Darn software...  Yep your right ;-))  I have one too many
> firing statements in the program...
>
> Cheers,
>
>         Terry
>
> >Gerry R
>
>