Re: OL-DRSSTC - 13

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Steve Ward <steve.ward@xxxxxxxxx>

Hey Terry,

13 already? i wonder where it ends? :-)

On 10/15/05, Tesla list <<mailto:tesla@xxxxxxxxxx>tesla@xxxxxxxxxx> wrote:
Original poster: Terry Fritz 
<<mailto:vardin@xxxxxxxxxxxxxxxxxxxxxxx>vardin@xxxxxxxxxxxxxxxxxxxxxxx>

Hi All,

Today I did some single shot testing at ~170v buss:

<http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-001.gif>http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-001.gif

Primary current yellow, buss voltage is blue.

As expected, the burst shortens up so most of the real power is in
the first 400uS.  This is an interesting tuning spot since you can
see the modes switching between poles as the streamer forms.


I think the beating is just because you are exciting both frequencies 
simultaneously, but usually one or the other is more dominant, so you 
dont get full cancellation when the 2 frequencies cancel.  Just do an 
FFT on your primary current and you should see the two peaks, one of 
them being smaller (usually the upper pole is smaller).  So i dont 
think its "switching" between poles, but rather, exciting them both 
at the same time.

I notice for small DRSSTC's that tuning the primary extra low, so as 
to just about eliminate the upper pole allows for more energy to be 
*quickly* delivered to the spark.  You no longer get a beating 
waveform (indicating energy transfer back and forth), but rather the 
primary current builds linearly over many cycles until finally, the 
secondary can zap it all out (at which point you do get a steep dip 
in primary current).  Im not sure if the OL-DRSSTC can take benefit 
of this tuning.  In my smaller DRSSTCs im limited to maybe a 24" 
spark with tunings that cause the beating effect, but when i detune 
it to get only one "beat" i can do 37".  This whole scheme doesnt 
seem to matter for big coils, since it seems you can deliver enough 
energy in a few cycles anyway, so one normal "beat" is usually enough.


The droop is not bad at all anymore but the turn-on timing is still
about 950nS out:

<http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-002.gif>http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-002.gif

It tries to turn on at ~600nS but it seems to have to build up steam
to drive the current out.


Perhaps this is due to some minor shoot-through with the reverse 
diode and the IGBT in each leg turning ON during the transition?

  It stutters for about 200nS while the gate
voltage builds up.  The turn on transition is about 350nS against
about 220amps.  Not sure how bad that is.  I will find out when
running it full blast ;-)  The total power is not bad, but the peak
power is high there...  I may get away with it...

I moved to the big variac now.  The little OL-DRSSTC get lost between
the variac and scope ;-)

<http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-003.jpg>http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-003.jpg 


Note the "blast shield" between the OL-DRSSTC and the scope now :o)))

Here are some little sparks at only 35V buss for DC Cox :o))

<http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-004.jpg>http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-004.jpg

I am a little worried about loop current.  With a primary inductor,
the buss voltage can hit 240VDC where the computer says:

<http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-005.gif>http://hot-streamer.com/temp/OL-DRSSTC-2005-10-15-005.gif

It can hit 600 amps!!  That is sort of "high"...


Makes for great sparks!

I can reduce system power buy lowering the buss capacitance, reducing
buss voltage, and decreasing the primary capacitance.  But I think I
will leave as is since I really don't want to use the buss inductor
at all.  That would put the nominal firing voltage at 170VDC which
might be just right....


Is there any way to run 340VDC?  Might be better to use more V and 
less I for the IGBTs.  You could most likely make better use of the 
IGBTs that way.  Im wondering if a half-wave voltage doubler could 
work (as seen in microwave ovens, or our VTTCs)?

Computer says...
435 amps peak
36 Arms primary current


Those MMCs are going to get a little warm ;-)

18A peak AC line current
7.9 Arms AC line current


948VA?

390 W buss power


Crummy power factor :-(


225 W streamer power
150kV peak output voltage

Seems just fine ;-))  I should note that streamer models are weak and
the above numbers do depend a lot on that (especially streamer
power).  But history says that those are good numbers in any
case.  Steamer power could probably be better estimated as "BussPower
- SystemLoss" in the DRSSTC case...


I would agree... where else could it go?

Steve Ward

Cheers,

        Terry