[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Coupling vs secondary voltage chart



Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>

Hi Steve,

At 08:33 PM 6/16/2005, you wrote:
Terry!

On 6/16/05, Tesla list <tesla@xxxxxxxxxx> wrote:
> Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>
>
> Hi Steve,
>
> The program says that your coil should work better at k=0.18 !!

Program appears to be right!!

:-)))))))))))))))))))))))


Just lowered the K to ~.18 according to JavaTC9.1.  I also arranged to
run the coil on my 240V line so i dont have to worry about the 15A
120V breaker ;-).  Solid 6 feet at about 80% input (out of 280VAC) and
only requiring 150uS burst lengths.  At full 85% input i got some 80"
strikes!  I didnt dare push to 100% yet.

Hmmmmmmmm, seems to be working "better" than expected. Might be due to the better AC line. But lowing the K does not seem to have hurt it ;-)))



The MMC is running very hot, and i saw a small flash on one of the
caps again.  So i gave it a break and tested the capacitance on my
meter, still 451nF dead on!  Fire it back up and it just spits out
those 6 footers plus some as i increase the BPS to 150 or so.

The program gave "ICprimary RMS/sqrt(BPS) = 3.976".... multiply that number buy the square root of BPS:


3.976 x SQRT(120) = 43.55 primary RMS amps.

The "sqrt(BPS)" factor seems strange.... But that is what it is.... Trust me ;-)))

Over six strings, that is 7.26 amps per cap. They are rated at 13... Are you caps "old or new"? The newer caps have a higher current rating.

My tests at:

http://hot-streamer.com/TeslaCoils/MMCInfo/GeekCaps/GeekCaps.htm

put them at 8 amps each (old caps).... There might be things that need to be looked at here... The caps "should" be fine. They might get to body temperature, but if they get hotter, something is wrong in our assumptions or the caps.

We may have to do an "autopsy" on your caps someday in the future. The problems you are having do not have an obvious explanation... I will have to review your website to see if any physical mounting things may be the cause... DRSSTCs seems to push caps ever harder than "every other" application on Earth :D



>
> There is more power delivered to the streamer and it is delivered faster!!
>
> I wrote a file explaining it (sort of quickly thrown together...) at:
>
> http://drsstc.com/~terrell/modeling/WardsCoil-18vs25.pdf

Cool, thanks.

I need to throw my current transformer on there to see if my waveforms
match the waveform shown there.  I notice im not tuning the primary
super low anymore but still at 13.3uH according to javaTC.  Thats a
little lower than what TeslaScan says should work (says 12.2uH).
Maybe i should tune it a little higher next time ;-).  But the
difference is like a 1/4 turn... so i'd say thats fairly good!

Just tune for best performance. There are a lot of "room effects" and such that might mess with the fine tuning. It is hard to even measure those things below 5% without really fancy toys...



>
> The chart on page three shows the streamer energy (joules) as a function of
> time.  Blue is the K=0.18 case and read is the K=0.25 case.  The higher K
> is actually hurting the speed of the power delivery and apparently wasting
> more of the coil's energy.

Wow, that is very cool to see!  Looks like i get another .5-.6J per
shot in there.  It would be nice to know the actual power draw on the
240V line, but those darn filter caps mess up the current waveform so
much you cant get a good number.  7J per shot just seems small for 6
foot sparks, does it not?

The very high K was killing the fast streamer "power" time. It had a high "first peak" time, but the "real energy" was delivered far later. Better off with a lower K that provides more power over a longer time even if it does initially get off to a slow start. Jimmy (an others ;-)) seem to have been right about that, but blindly raising the K is not the completely right solution...


John's formula (120 BPS) is 18.6 x sqrt(bang power) = 49 inch sparks. :-))) But you don't have a spark gap eating up all the power. We probably need a new number to replace the "18.6".... If you were getting 80 inch arcs at 7 joules, the new number is "30.2 x sqrt(bang power)" ;-))))))))) That actually IS "about right" for DRSSTCs... I have never had a thermal problem with my IGBTs and I don't think anyone else has either... We have all that old spark gap's waisted power going into "our" sparks now ;-)) We still need to watch primary losses....

Has anyone ever had their IGBTs run warm or hot on a DRSSTC????? We can certainly blow them to bits due to cross conduction, but from just "getting too hot" over time???

As far as low K goes, remember that folks have run coils with the secondary on the other side of the room ;-))


>
> The delivered power to the load goes from 6.253J to 6.822J and the power is
> delivered about 20 to 50uS faster.

Yes, i noted that i was getting 6' at 2 cylces less than before (which
is about 20uS).

  The primary currents go down a tiny bit
> and the capacitor voltage goes up about 100V so not big deal
> there.  Primary frequency goes up a tiny bit.  Secondary peak voltage is a
> tiny bit lower.

I'm quite surprised to see the program says my cap voltage stays under
4kV.  Maybe that other capacitor got so hot that the plastic just lost
its dielectric strength?  I cant imagine how hot they are on the
inside when they feel uncomfortable to touch on the outside.

Only 1700Vcap... But heat is all due to current, not voltage. One can barely hold a 50C object. Body temperature is 37C. The temperature "might" be "ok"... A lot depends on how fast they heat up. If it takes an hour to get to 50C, or 10 seconds to get to 50C ;-)) But in general, obvious dramatic heating is a "bad thing"... Keep the runs short...



>
> If you get 72 inches now, changing the k to 0.18 might increase that to 72
> x SQRT( 6.822 / 6.253 ) = 75.2 inches.

Or better ;-).

If you really are getting further, that may be an indication the models needs some improvements.... But a lot of old models seems to be on shaky ground right now... Getting better models is really what this is all about ;-))



>
>
> This is using the programs new "streamer power vs. time" function.  It
> shows the energy delivered to the streamer as a function of time.
>
> The latest program is here:
>
> http://drsstc.com/~terrell/modeling/ScanTesla620.ZIP

I should really download it and mess with it some more, try entering in
parameters for my other coils and see what it says will work better
;-).

I keep changing it all the time too ;-)) But the basic program seems real solid right now. Just adding stuff and very minor bug fixes... I need to write instructions to help those totally unfamiliar to see what is going on. It is super powerful and super easy to use but it does need instructions.... "I" obviously know waaaay toooooo much about it ;-)))


>
> Sorry to keep updating the program all the time.  But I just keep adding
> new features as I want ;-))  I need to write general instructions for it
> too....

Yes, instructions please!  Especially for me, im not good at just
figuring out programs like this, but i shouldnt comment until i
actually look at it.

Unlike MicroSim, we now have full control of the model and can just change things just the way we want them ;-)))



I surely hope this whole event wasn't some fluke and that the program
really is giving out good information.  It really does seem to right
so far!

We have been playing the "model game" for many years with great success. There my be problems in the little details, but the big picture is very solid and well understood. Lowering K for better performance in your case is based on very solid evidence beyond this program.... The program just make is look "obvious" now ;-)))


Very happy to here your coils is working better from all this!!!!!

Those of us that like to run coils on a computer screens, can kick "real" a** now and then ;-)))

Cheers,

        Terry


Steve

>
> Cheers,
>
> Terry
>
>
>
>
>
> At 12:41 PM 6/16/2005, you wrote:
> >Heh, thats pretty interesting that K=.25 is the peak output.
> >
> >Looks like i could even back off the K to .18 or so and still probably be ok.
> >
> >Its interesting to see that the graph looks smoother at the very low
> >K's. If these predictions are in fact meaningful, i would be curious
> >to see you operate your big coil from a K of say .03 to .12 or so and
> >see if peak performance lies at the peak in your graph there.
> >
> >So i wonder what causes this difference in coupling preference, is it
> >the tank circuit or the secondary characteristics?
> >
> >Maybe you could try a different tank impedance on your big TC, say a
> >.5uF cap and whatever primary it would need to tune with then and see
> >what K it prefers? Or maybe try my secondary specs with a .05uF tank
> >cap or something and the needed L to retune. That seems like it might
> >cover most of what DRSSTC builders might come up with... the next
> >thing would be to put these results to the test.
> >
> >And, good work Terry!
> >
> >Steve
> >
> >On 6/16/05, Tesla list <tesla@xxxxxxxxxx> wrote:
> > > Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>
> > >
> > > Hi Steve,
> > >
> > > The graph is at:
> > >
> > > http://hot-streamer.com/temp/KvsVtop-SteveWard-DRSSTC.gif
> > >
> > > It would appear your K of 0.25 is perfect! The first low K hump is there
> > > but far lower down for your coil.
> > >
> > > The graph of my other big coil at 140uS is at:
> > >
> > > http://hot-streamer.com/temp/KvsVtop-BigCoil-DRSSTC-140uS.gif
> > >
> > > Cheers,
> > >
> > > Terry
> > >
> > >
> > > At 11:00 PM 6/15/2005, you wrote:
> > > >Hi Terry,
> > > >
> > > >Let me just enter my values right into the text below:
> > > >
> > > >On 6/15/05, Tesla list <tesla@xxxxxxxxxx> wrote:
> > > > > Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>
> > > > >
> > > > > Hi Steve,
> > > > >
> > > > > The model is running now at 140uS...
> > > >
> > > >Did it change the coupling vs Vsec graph? OR when you say its
> > > >"running" do you mean its currently crunching numbers?
> > > >
> > > > >
> > > > > If you send me the following data on you coil I can run it. My
> > coil might
> > > > > be substantially different than yours. The coil I modeled was
> > never meant
> > > > > to be a DRSSTC.
> > > > >
> > > > > Lprimary =14uH (though it might be 2-3uH lower)
> > > > > CPrimary =450nF
> > > > > K =.25 (this is a changing parameter, right?)
> > > > > Lsecondary =130mH
> > > > > Csecondary =38pF
> > > > > Streamer length =72"
> > > > > Rail to rail buss voltage =400V
> > > > > T1 on time =175uS
> > > > > Current limit =850A
> > > > > BPS =120
> > > > > Max coil power =1800W (might be less... the killed the kill-a-watt with
> > > > my SGTC! arggg!!)
> > > > > If the coil is on your site, which one is it?
> > > >
> > > >The DRSSTC-1
> > > >
> > > >http://www.stevehv.4hv.org/DRSSTC1.htm
> > > >
> > > >I look forward to what your program spits out ;-).
> > > >
> > > >Steve
> > >
> > >
> > >
>
>
>