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

Re: LTR Pig Project - Long Post!



Original poster: "Kurt Schraner by way of Terry Fritz <twftesla-at-uswest-dot-net>" <k.schraner-at-datacomm.ch>

Hi Shad, John Freau, ...

[Remark: this posting was waiting in my drafts box, to be sent,
some time before, but was almost forgotten. I decided to send it
now, because the thread on the LTR Pig project continues.]

...your LTR pig project is just "after my heart": I like it! I
follow John Freau's efficency design rules with strong interest,
accepting (and even admiring :-) ) the good mix of theoretical
argument with experimental proof. Part of it was leading me in
the design of my bigger coil B&W. John's approach proves to look
very successful for (mainly-) NST powered coils of moderate
power, up to a few kVA. There seems to be a discrepancy of
practice vs. prediction in the higher power range: if, i.e., I
look at Bart Anderson's list of coils 97-99, from the first 20
TC's with the biggest sparks and power, there cannot be seen ONE
with a LTR, even not with a resonant cap, relative to the power
(and ballasting) used, but only smaller than resonants (STR's)!
The question of

          SCALEUP and SCALEDOWN of TC's,

for biggest spark at smallest input power, looks to me not yet
having a definite answer, neither theoretical nor just "crude
engineering design". (Don't mean, you cannot just design a
successful high power coil a rational way...). One of my main
questions is, if you extrapolate, say, a successful LTR-desingn
for 1kVA to 10..20kVA, with corresonding huge bang-size, and
shift of the whole frequency domain: how will the secondary
behave? Is linear extrapolation feasible? Won't different
mechanisms and constraints come to bear, like racing sparks or
power arcs between prim and sec? - I recognize, i.e., Robert W.
Stephen, in his BC-1 design procedure, obtained a reso-cap of
0.128uF, but reduced it 'wisely' to only half, 0.064uF, for a
10kVA design proposal. Even, if set to 0.12uF in the MTC-1 coil,
which looks fairly similar to the BTC-1 design, it was yet just
below the (dangerous) resonant value. I also see Richard Hull's
very successful designs of higher powered TC's and maggies
following an alternative way to LTR's, comprising high BPS and
small (shurely stressed!)caps.

On the other hand, scaling down to tiny coils, there might be a
problem with high inductance secondaries of > 1500 turns, as
mentioned by John F. himself, lately, 29th jan 2001, "2400 turn
sec report":

"I am not surprised by the results, because I've noticed in past
work that large secondaries do not work well with small toroids
for low powered work.  Low power and small toroids need to have
small secondaries."

In chemical-engineering, for example, I see scaling dominated by
dimensionless numbers (i.e. Reynolds for fluid flow). Do we need
dimensionless numbers for TC's? - My first guess is: no, - but
might be, we need to consider the change of limiting mechanisms
(-constraints-, like HV-breakdown in High power or mechanical
distances in low(-and high) power) for the different regimes of
power.

Further comments interspersed.

Tesla list wrote:
> 
> Original poster: "sundog by way of Terry Fritz <twftesla-at-uswest-dot-net>"
<sundog-at-timeship-dot-net>
> 
> Hi All,
> 
>  Firstly, Yep, I'm serious.  The LTR pig project is sitting in my
> garage.
> 
> Now, before you ask, "Why??!?", think about the "why not?", and
> let's put it in perspective.
> 
> The general consensus I've recieved on polepig powered coils is to
> run a high breakrate, small capacitor, and crank the power to it.

I dont't know of a big coil with a reso-cap, let alone an LTR, as
mentioned above.

> The problems I see in that...the high RMS current is murder on the
> caps.  The high power levels promote power-arcing on the RSG due to
> poor quenching.  Most common comment I hear, "Resonant value isn't
> important with a pig-based system."
> 
>  Now, let's say everybody "normally" ran on 10kv20mA OBITs.  A
> 15/60 would be the equivelant of a pole pig!  Literally GOBS of
> power!!!  Would you run a tiny cap, high breakrate and simply watch
> your RSG fry?  Nope.  Especially if it was proven you get longer,
> stronger sparks from a 120bps LTR setup on your OBIT.  A pig is the
> same thing.  It's just a big rabid NST.  Yep, the cap size is
> massive, and expensive.  I wouldn't recommend this project to an
> amature, or the average joe coiler.  The pig's output is the same
> (lethal), but the cap bank is literally 2 to 3 times as large.(very
> lethal).
> 
> Something to keep in mind here...this tank circuit is capable of
> *easily* powering 10-12' arcs.  Think big.  Very big.
> 
> I test-fired my LTR pig setup this evening, on an already-damaged
> 8" coil.  Output was dismal, only ~3' tops, with massive flashovers
> and the windings being blown off of the coil.

Flashovers.., seem a problem of higher power. I've not been able
to raise the coupling factor of B&W above 15% --> 18% without
introducing a polycarbonate sheet between sec and prim (distance
of the windings in air was 2"). Without just a critical value of
voltage and power, removing the breakout point on the toroid,
provoced power arcs between prim and sec, even 14" around the
polycarbonate shield, with ability to damage the secondary.

>  I knew it would
> fail, and I expected it to fail.  It was the tank circuit I wanted
> to test.  Though I wish it'd put up a better fight.  :\  Casualties
> included my primary tap (de-soldered itself from the lead, kept
> spot-welding itself to the primary.)  Also lost was an MMC string.
> I have yet to figure the RMS current the caps were seeing, but it
> should be nowhere near their 3A rating I use.  No, this was
> overvoltage.  While on a NST system, you can get away with an EMMC
> or even an EEMMC, because the current is low.  On my polepig
> system, my 12kv designed cap (16.7kv rating) is overvolted, but
> acceptably by NST standards.  But at 8kva, when one of the caps
> blasted through, the high current took it's toll.  No self-healing.
> Every single cap in the string literally exploded.  Ouch.  If using
> an MMC on a pig at high power levels, *DO NOT* skimp on it.  11
> caps popped like firecrackers.  Subsequent runs will have the pig
> throttled down to 12kv out.    Other than that, the project is a
> resounding success!! The setup behaved just like a bit NST system.
> Definate potential!!!!

My cap was about 40...50 J Bangsize. 
(109nF/300kV; 16kVrms 4//PT's -at- ~11kVA input power)
> 
> I'm in the process of designing a magnifier to run on this setup.
> Here are some of the things I've taken into consideration and why.
> 
>  The gap.  It's done.  Tungsten-carbide stationary and flying
> electrodes.  The huge cap bank and low breakrate means the gap will
> see tons of current when it fires.  But being LTR, after the
> initial fire/quench, the tranny will be doing all it can to charge
> the cap for the next bang.  Powerarcing is eliminated, as is
> multiple-firings.  My 1/8" tungsten rods used in the gap have
> almost no wear on them from the useage they've seen so far.

My SRSG's 10mm-electrodes were just handwarm. No serial
static-gap this time. 200 BPS.

>   The primary.  At least 1/4" tubing.  My primary tap is a 1/4"
> fuse holder.  It keeps desoldering itself from the lead, and
> spot-welding itself to the primary every time I fire the coil off.
> Problematic to say the least.  Plan on lots of turns.  Lots and
> lots of turns.  Primary inductance is the only thing keeping your
> gap current down, and high gap current can and will eat even
> tungsten-carbide like candy and destroy your caps from too much
> pounding.  Be nice to your caps, run high inductance primaries.

Agreed! - Running 7.2 turns primary tap (51uH) was not quite
there.

>   The secondary...keep in mind the huge bang size.  capable of
> making super-high voltage spikes.  Figure your target voltage,
> figure the toroid, then figure how many turns you need to keep your
> turn/turn voltage sane.  Use a good gauge wire to handle the ground
> currents.  (they will be heavy) Don't be afraid of very high
> secondary inductance, you'll need it to raise the primary
> inductance, hence keeping your gap current under control.

With B&W I didn't like the resonant frequency going too low, this
giving manageable sizes of the primary, relative to my available
caps. John's 1600 turns could not be maintained that way: I have
821 space wound turns only. I don't know, if this was a mistake.
However the toroid is quite big: 63"/16".

>   The caps...Given the high number of strings you'll be running in
> relation to your input current, RMS current ratings should not be a
> huge concern.  Pay very close attention to the voltage ratings.
> There's an obscene amount of power in the tank cap bank, and any
> small failure will instantly become really nasty.  Remember the
> exploded MMC string?


Current becomes more important in a big system. My 109nF caps are
300kV but only 25A rms. Stressed 50A rms! (however pulse current
of these Maxwells is up to 25kA).
> 
>   Ballasting and current draw...
>    this could take up a whole thread on it's own.  Right now I'm
> using pure inductive ballast.  The pig is throttled back to ~35A
> (direct short on the HV side).  I'll be using resistive ballast
> also to throttle the HV output to 12kv and help smooth the current
> draw.  As it is, the ballast lags the demmand, and the pig is
> getting current in huge chunks that the ballast then chokes down.

> I do not know if inductive kick is going on (probably is), but I
> want to limit the inductive kick into the pig.  It is *not* a good
> thing.  The caps already see a high enough voltage from the pig,
> the last thing you need is them seeing an extra 5-6kv spike because
> of the inductive kick.

As long as the caps are not stressed too much, I would see it as
a good thing: Voltage must be high for low spark-gap loss!

>  Hopefully rectifying the pig's output to DC
> will help with the charging current situation.  I'll be buying PFC
> caps for the pig as my budget allows.  They are beneficial to a
> NST, and the pig is behaving exactly like an NST.   A tappable
> inductor is being considered for this thing, and my target for it
> is 10 minutes of runtime at 30A.  10 minutes total runtime at that
> power level is a lot of time. Of course, not all at once.

>    And forget the variac.  By the time you get the voltage high
> enough to fire the gap reliably, you'll about be at full power.

True, but it's nice to fine adjust operation. Didn't have any
problems with my variac.

> Start with the system decoupled a *whole* lot, and use a breakout.
> Tune from the outside of the primary inwards for best spark, then
> increase coupling.  It's far to easy to set the coupling too close
> at half voltage, and then watch the secondary blow windings off
> when you crank up the voltage.  At multi-kva levels, the
> joules/bang is *very* significant, and because J=½C*V², cranking up
> that variac will drive your bangsize through the roof.  Run it at
> full power from the get-go, and use coupling to control how much
> power the secondary is seeing.  Plan on how long a spark you want,
> and when you get it, stop there!

Very good! - I planned 4m (157";13') and got 4.15m (163";13.6').
But I "think" may get 5m (197";16.4')...;o).

>  If you designed the system to
> handle that big of a spark, be happy! :) If you push it too far, it
> will certainly self-destruct, and when a high-joule/bang coil lets
> go, it's pretty spectacular! (not to mention expensive!)
> 
Had bad arcovers so..., see above!

>   This project is *far* from over. :)  Put on the back burner till
> I can properly assemble a coil capable of handling that much power,
> yep.  ANd even that's in progress (tracking down materials,
> planning, etc.)  I'll be working up more numbers for the LTR setup
> to post as time goes on.

>  I invite comments, questions, etc. I'll
> be posting photos and scans of the setup as I get time.  The
> projected magnifier uses a 55gal HDPE drum for the driver, and the
> tertiary is as of yet undecided.  My target sparks are 12', and
> more is better.
> 
>                                             Shad
Cheers,
        Kurt Schraner