If correct, this will change everything!! Discussion
Last edited/updated: October 23, 2004
First Post
Date : Sun, 03 Oct 2004 13:02:27 -0600. Subject : If correct, this will change everything!!
Original poster: Terry Fritz
Hi All, I was working on DRSSTC theory last night. So I ran some models. Slept
on it... Wrote it down this morning on the laptop...
Just like the OLTC was optimized to giant low voltage currents, we can optimize
the DRSSTC to use small primary currents but higher voltages. Then we can take
advantage of the fact they have vast power reserves unlike a cap discharge coil
that only has a fixed amount of bang energy. Then we can optimize it all to run
off "little cheap"
IGBTs... This all went pretty quick, but unless I messed up, the whole world of
coiling will change now...
http://hot-streamer.com/temp/DRSSTC-Optimization.pdf
http://hot-streamer.com/temp/DRSSTC.sch
Of course, if I made some giant boo boo, forget I ever brought it up ;-)).
Cheers, Terry.
Date : Sun, 03 Oct 2004 15:39:43 -0600. Subject : Re: If correct, this will change everything!!
Original poster: Terry Fritz
Hi Matthew, It is a MicroSim model file for PSPICE style simulations. I just
wanted to save folks the time of typing the model in if they had the free
student version MicroSim 9.1 which most DRSSTC people use these days.
Cheers, Terry.
Date : Sun, 03 Oct 2004 15:39:52 -0600. Subject : Re: If correct, this will change everything!!
Original poster: "Matthew Smith"
<matt@kbc.net.au>
Terry Fritz wrote:
<blockquote>
I was working on DRSSTC theory last night. So I ran some models. Slept on
it... Wrote it down this morning on the laptop...
<snip/>
http://hot-streamer.com/temp/DRSSTC.sch
</blockquote>
What's the .sch file Terry? I thought that it was an Eagle schematic, but Eagle
just glared at me and sulked when I went to open it...
Cheers, M.
Matthew Smith
Kadina Business Consultancy
South Australia
Date : Sun, 03 Oct 2004 17:48:15 -0600. Subject : Re: If correct, this will change everything!!
Original poster: "Steve Ward"
Your gonna get a lot response on this one Terry ;-),
My thoughts. It seems that you have basically designed a DRSSTC with a rather
high surge impedance compared with the secondaries resonant frequency. 65nF
tank cap almost sounds normal compared to my first DRSSTC that used 75nf, but my
system works at 140khz while yours is designed for merely 40khz!!
Now here is sort of an oddity i have noticed with DRSSTCs to date. My systems
always prefer smaller tank capacitors and more cycles to drive the resonator,
while Jimmy's original DRSSTC used a HUGE tank capacitor (.6uF!) and many less
cycles by comparison. When i tried larger tank caps on my coil, the efficiency
dropped off and i could not achieve the great spark lengths i had seen with the
smaller C. *Same thing happened on my big DRSSTC!* I used a 360nF cap on the
big coil for most of its life time, but there was a period where i tried
450nF. It resulted in shorter sparks (maybe 9.5' compared to 11.some'). Also,
i lost quite a few IGBTs that way as well.
Anyway, i plan to take a look at your simulation file some more and try
modifying the driving method (you have notches in your waveforms... we typically
go for no notch now). Jimmy's first DRSSTC setup had notches too because he
drove the coil between the 2 split frequencies, but eventually he started
driving the whole thing at the lower split and got even better performance. I
want to see how your sim reacts to this.
>
> Of course, if I made some giant boo boo, forget I ever brought it up ;-))
Hehe, well not sure what to make of it yet. We still dont know what the best
way to drive these things is yet. Some methods seem to work very well while
others just flop. I might have to make an absurdly fat secondary with small
wire to test out your idea ;-). In fact, I'm open to suggestions on what types
of combinations i should try as far
as operating frequency, L to C ratios of both primary AND secondary circuits. Ive
been looking for something new to do for awhile now ;-). Steve.
> Cheers, Terry
Date : Sun, 03 Oct 2004 17:48:39 -0600. Subject : Re: If correct, this will change everything!!
Original poster: "Finn Hammer"
Terry, all
I'm sure I speak for more than myself, when I ask how it is possible to:
1) build up kilovolts from hundreds of volts, without a transformer, and
2) contain kilovolts inside the tank, with IGBT as gatekeepers. IGBT's are known
to be less than tolerant to overvoltage. Why don't they blow due to overvoltage?
Cheers, Finn Hammer.
Date : Sun, 03 Oct 2004 17:56:48 -0600. Subject : Re: If correct, this will change everything!!
Hi again,
I hope im wrong, but it seems like ive found something about your simulation
that makes it look exceptionally *good*. You gave the tank capacitor an initial
charge, and that seems to have made all the difference. Im not sure why because
it was only set to 10, which i assume may mean 10v, but maybe 10kv? Anyway,
resetting the IC
(initial charge) to 0 made things look just normal. 60kv on the topload, a bit
over 1kv on the tank cap, and not even 20A in the primary (... huh??). Anyway,
just looks like a mismatched setup :-(.
Maybe *I* missed something??? But now you have me thinking of something new...Steve.
Date : Sun, 03 Oct 2004 18:00:08 -0600. Subject : Re: If correct, this will change everything!!
Original poster: Terry Fritz
Hi Steve,
I had the 10V in there to use as an initial condition so I could ring down the
shorted primary system to determine its stand alone resonant frequency. I had
since forgotten it.
You are correct, this initial condition seems to be having some odd affect on
the model!! If I set it to 100V then I get 5 millions volts on the output!!!
I will try and sort this out but it does appear there may be an error in the
analysis!!
Cheers, Terry.
Date : Sun, 03 Oct 2004 18:16:07 -0600. Subject : INCORRECT!! - Re: If correct, this will change everything!!
Original poster: Terry Fritz
Hi again Steve,
As far as I can tell. The initial condition of 10V on the primary cap must be
forcing 10 / 0.03 = 333 amps through the primary system also as a phantom
initial condition. That false initial current is messing up the system's
energy!!
So Indeed I think the model is now completely wrong!!!!
Sorry bout that!! Darn computers!! :o)
I thought it was just a little too good to be true :-(( I still have another
idea so I will try that next. I still think there is much promise to this area
of study even if it does have a few traps lurking ;-))
Cheers, Terry
=====================
Hi Steve,
I had the 10V in there to use as an initial condition so I could ring down the
shorted primary system to determine its stand alone resonant frequency. I had
since forgotten it.
You are correct, this initial condition seems to be having some odd affect on
the model!! If I set it to 100V then I get 5 millions volts on the output!!!
I will try and sort this out but it does appear there may be an error in the
analysis!! Cheers, Terry.
At 04:51 PM 10/3/2004, you wrote:
>Hi again, I hope im wrong, but it seems like ive found something about your
>simulation that makes it look exceptionally *good*. You gave the tank
>capacitor an initial charge, and that seems to have made all the
>difference. Im not sure why because it was only set to 10, which i
>assume may mean 10v, but maybe 10kv? Anyway, resetting the IC
>(initial charge) to 0 made things look just normal. 60kv on the
>topload, a bit over 1kv on the tank cap, and not even 20A in the
>primary (... huh??). Anyway, just looks like a mismatched setup :-(.
>
>Maybe *I* missed something??? But now you have me thinking of something new...
Steve.
Date : Sun, 03 Oct 2004 18:50:52 -0600. Subject : Now incorrect!! Re: If correct, this will change everything!!
Original poster: Terry Fritz
Hi Finn, See these links for the explanations:
http://hot-streamer.com/srward16/DRSSTC-1.htm ,
http://hot-streamer.com/srward16/SolidStateTeslaCoils.htm
http://hot-streamer.com/chunkyboy86/,
http://www.easternvoltageresearch.com/sstc_coils.htm
Look for "DRSSTC" stuff. Basically it is a low voltage current source added in
series with the tank circuit that drives current into the primary. It rings the
L and C up to high voltages and currents while the voltage on the current source
stays low.
Since my fist idea went up in modeling flames (POOF!!), I will now try the long
ring up time approach with low coupling to pack as much energy as possible into
the primary... It's cool stuff! Probably the future, even
if I can't make the drivers cost less than $2 ;-)). Cheers, Terry.
At 04:50 PM 10/3/2004, you wrote:
>Terry, all
>
>I`m sure I speak for more than myself, when I ask how it is possible to:
>
>1) build up kilovolts from hundreds of volts, without a transformer, and
>
>2) contain kilovolts inside the tank, with IGBT as gatekeepers. IGBT`s are
>known to be less than tolerant to overvoltage. Why don`t they blow due to overvoltage?
>
>Cheers, Finn Hammer
>
>
>
>Tesla list wrote:
>
>>Original poster: Terry Fritz
<teslalist@twfpowerelectronics.com>
>>
>>Hi All,
>>
>>I was working on DRSSTC theory last night. So I ran some models.
>>Slept on it... Wrote it down this morning on the laptop...
>>
>>Just like the OLTC was optimized to giant low voltage currents, we can
>>optimize the DRSSTC to use small primary currents but higher
>>voltages. Then we can take advantage of the fact they have vast power
>>reserves unlike a cap discharge coil that only has a fixed amount of bang
>>energy. Then we can optimize it all to run off "little cheap"
>>IGBTs... This all went pretty quick, but unless I messed up, the whole world of coiling will change now...
>>
>>http://hot-streamer.com/temp/DRSSTC-Optimization.pdf
,
http://hot-streamer.com/temp/DRSSTC.sch
>>
>>Of course, if I made some giant boo boo, forget I ever brought it up ;-)).
Cheers, Terry.
Date : Sun, 03 Oct 2004 18:51:56 -0600. Subject : Re: If correct, this will change everything!!
Original poster: "Steve Ward"
Finn,
Comments:
On Sun, 03 Oct 2004 17:48:39 -0600, Tesla list
<tesla@pupman.com> wrote:
> Original poster: "Finn Hammer"
> Terry, all
>
> I`m sure I speak for more than myself, when I ask how it is possible to:
>
> 1) build up kilovolts from hundreds of volts, without a transformer, and
Many people find this hard to believe. Its all resonant stuff. DRSSTCs
build up many kV in the tank circuit all from just a few hundred volts an NO
transformer ;).
>
> 2) contain kilovolts inside the tank, with IGBT as gatekeepers. IGBT`s are
> known to be less than tolerant to overvoltage. Why don`t they blow due to overvoltage?
This high voltage exists basically at the node where the L and C connect. On
either end of the series LC the voltage is clamped to the power supply from
which it all originates. We need those reverse diodes on the IGBTs to clamp
this voltage safely to the filter caps.
Steve.
> Cheers, Finn Hammer
Date : Sun, 03 Oct 2004 18:51:14 -0600. Subject : Re: If correct, this will change everything!!
Original poster: "Antonio Carlos M. de Queiroz"
Tesla list wrote:
>
> Original poster: Terry Fritz
> Hi All,
>
> I was working on DRSSTC theory last night. So I ran some models. Slept on
> it... Wrote it down this morning on the laptop...
>
> Just like the OLTC was optimized to giant low voltage currents, we can
> optimize the DRSSTC to use small primary currents but higher
> voltages. Then we can take advantage of the fact they have vast power
> reserves unlike a cap discharge coil that only has a fixed amount of bang
> energy. Then we can optimize it all to run off "little cheap"
> IGBTs... This all went pretty quick, but unless I messed up, the whole
> world of coiling will change now...
For fixed capacitances, the input current is always inversely proportional to
the square root of the inductances, and the excitation frequency too. This
just impedance and frequency scaling. So, if you want to to halve the
input current, use inductors four times larger. This also divides the input
frequency by two. Antonio Carlos M. de Queiroz.
Date : Sun, 03 Oct 2004 19:27:45 -0600. Subject : Incorrect... But... "current" OLTC?
Original poster: Terry Fritz
Interesting.... The initial condition error was basically sending 333 amps into
the large primary inductor and storing 14.9 joules of energy in it as 1/2 x L x
I^2 !! Sort of a "current charged" version of the OLTC... Charge a big
inductor with current instead of a big cap with voltage... That theoretically
could charge 40pF to 863kV. 500kV is only 5
joules... 66% of the energy is still being lost but most into the streamer at
say 9 Joules "output"
If one could charge the "inductor" with current, pumping from each side of the
primary cap, and then quickly removed the current source with 20kV of
isolation... You would get a very large output voltage as the model originally
showed... No IGBT's at all would be needed!! Maybe another use for that old
arc welder!!
You would need practically nil voltage (just to overcome DC resistance) but very
high currents. With such a low R, the charging time constant should be
nil. t/RC, t/LR, ... Can't remember it... 5LR = 38uS...
The model then gives us 500kV peak!!..... Hmmmm.... We seem to have more ideas
than we know what to do with ;-))
I think this idea was mentioned before in the old OLTC days. It's going to
be a very long winter ;-)). Cheers.
Date : Sun, 03 Oct 2004 19:34:27 -0600. Subject : Re: If correct, this will change everything!!
Original poster: "Malcolm Watts"
Hello Terry,
A few comments about your paper:
- the notion of using high primary inductances to get primary currents to a low
level has been recognized as being beneficial for a some time now. I did a
quantitative analysis of this wrt spark gap coils pre-1998.
- there is typo in the text; I'm sure 250mH using 1/4" Cu tubing is a bit
unrealistic for a coil that might be made transportable ;) I note that mH
reverts to uH further along in the text.
- a 71 Ohm surge impedance is not terribly exceptional.
- the secondary losses are going to be rather high unless the coil is physically
huge in which case the associated capacitance is going to be correspondingly
high.
- You appear to be starting out with a charged primary capacitor which rings
down in conventional disruptive fashion. Have I misinterpreted your scope shots?
I suspect you saw the same nirvana I did a few months ago at which point I was
reminded by a couple of list members that in the process of ringing the system
up, the double coupling artifacts do not allow one to just ring up and up
without intermediate energy trades.
(Aside - I realized why after considering that a conventional disruptive coil
has the same topology save that its "supply" is effectively a zero Ohm piece of
wire; both conditions are modelled as being driven from voltage sources but in
the old disruptive case, the voltage source has zero output) - end aside.
The kind of analogue I had in mind was Dr Gary Johnson's coil which actually
took in the mS region to ring up to a useful level in that fashion with a modest
drive circuit. The results look far too good to be true - you have to pour
a
considerable aomunt of energy into the secondary to get its voltage to the point
where you can generate some stretching (i.e. you are not dealing with a drawn
arc which to my mind is implied by the secondary "load". Ideally, you want the
secondary to ring up with as little load as possible to get the output voltage
high before letting go. In short, I think the model is flawed.
Malcolm.
Date : Sun, 03 Oct 2004 19:52:23 -0600. Subject : Re: If correct, this will change everything!!
Original poster: Terry Fritz
Hi Malcolm,
At 06:53 PM 10/3/2004, you wrote:
>Hello Terry,
> A few comments about your paper:
>
>- the notion of using high primary inductances to get primary
>currents to a low level has been recognized as being beneficial for a
>some time now. I did a quantitative analysis of this wrt spark gap
>coils pre-1998.
Yes!! But it has new uses with IGBT drive coils were primary current needs to
be kept lower. So on old idea with state of the art electronics ;-))
>- there is typo in the text; I'm sure 250mH using 1/4" Cu tubing is a
>bit unrealistic for a coil that might be made transportable ;) I note that mH reverts to uH further along in the text.
Yup ;o))
>- a 71 Ohm surge impedance is not terribly exceptional.
>
>- the secondary losses are going to be rather high unless the coil is
>physically huge in which case the associated capacitance is going to be correspondingly high.
The higher capacitance is helpful in keeping the resonant frequency low so no
problem there! We need to beware secondary losses, but these coils are pushing
13 feet ;-))
http://hot-streamer.com/srward16/DRSSTC-2.htm
Even the old OLTC technology has gotten pretty good!!
http://www.scopeboy.com/tesla/t4spec.html
Secondary losses seem to be very manageable now days.
>- You appear to be starting out with a charged primary capacitor
>which rings down in conventional disruptive fashion. Have I misinterpreted your scope shots?
Excellent observation!! It was the inductor that was accidentally charged with
current. The primary cap had an initial condition of 10 volts left one
it. Trivial capacitor charge, but a 333 amp inductor initial charge of current
due to the low primary circuit resistance!!
>I suspect you saw the same nirvana I did a few months ago at which
>point I was reminded by a couple of list members that in the process
>of ringing the system up, the double coupling artifacts do not allow
>one to just ring up and up without intermediate energy trades.
DRSSTCs can add power throughout the firing cycle unlike our fixed break energy
capacitor discharge coils. In fact, that is the only time they can add energy
to the system, as odd as that sounds. That may be very important here very
quickly!! Energy trades and the losses (arc) perhaps can be controlled to an
optimal level. In fact, we know they can...
>(Aside - I realized why after considering that a conventional disruptive coil has the same topology save that its
>"supply" is effectively a zero Ohm piece of wire; both conditions are modelled as being driven from voltage sources
>but in the old disruptive case, the voltage source has zero output) - end aside.
>
>The kind of analogue I had in mind was Dr Gary Johnson's coil which
>actually took in the mS region to ring up to a useful level in that fashion with a modest drive circuit.
Yes, but now that they are putting out 13 foot arcs... Things have gotten "hot"
;-))
>The results look far too good to be true - you have to pour a considerable aomunt of energy into the secondary to get
>its voltage to the point where you can generate some stretching (i.e. you are not dealing with a drawn arc which to
>my mind is implied by the secondary "load". Ideally, you want the secondary to ring up with as little load as possible
>to get the output voltage high before letting go. In short, I think the model is flawed.
And so it is ;-)) I'll try not to mess up the next time. And, the time after
that :o)).
Cheers, Terry.