Re: LTR Pig Project - Long Post!

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "by way of Terry Fritz <twftesla-at-uswest-dot-net>" <FutureT-at-aol-dot-com>

In a message dated 2/17/01 4:27:35 PM Eastern Standard Time, tesla-at-pupman-dot-com 
writes:

> Hi John,
>  
>  At 02:52 PM 2/16/2001 -0700, you wrote:
>  snip...
>  > I would 
>  >say that one does not want to run a large degree of LTR in a pig
>  >because you want a high voltage for better primary efficiency, and
>  >you want to keep the power factor good.
>  
>  LTR coils should still fire at the full rated transformer voltage.  You
>  basically just keep loading up capacitance until you can't charge anymore C
>  to the full firing voltage.

Terry, all,

I agree that the LTR coil will still fire at the rated xfrmer voltage, but
I think a smaller cap may fire at a higher than rated xfrmer voltage
because of resonant charging that can occur at certain ballast
settings.  This higher voltage should help to improve the tank
efficiency provided that too much xfrmer saturation does not occur
due to the over-volting.  This is what I meant when I spoke of higher
voltages above.  In my PT powered TC's I'm getting about a 32kV
firing voltage from a 14.4kV xfrmer at certain ballast settings at a
slight LTR setting.  Using a slight LTR setting is not a new idea.
It is spoken about in Morecroft's book from 1921.  He says
that a frequency about 20 % lower than resonant should be used
because it gives smoother operation than a resonant value.  I may
have been closer to reso than 20% though when I got the 32kV firing
voltage.
>  
>  >
>  >Consider Kevin's coil.  His pig combo rating is 14.4kV at 20kVA. His 
>  >cap is 0.138uF or something similar.  He draws 100 amps from a 
>  >240 volt line.  He runs at 120 bps.   I just did the calcs, and a
>  >matched sized cap for Kevin's coil would be 0.31uF or so, and
>  >a typical LTR sized cap would be about 0.6uF or so.  This is 
>  >huge cap, and I don't recommend anything that large.  His cap
>  >seems to be less than 1/2 the resonant size.  (If you calc the
>  >reso-size based on the pig's specs.)  You may want to check
>  >my calcs here in case I made an error.  But if his reso-size
>  >would be 0.31uF, I certainly would not suggest using anything
>  >larger than 0.375uF or so.  Since Kevin's cap may be about
>  >1/2 the reso-size, it's possible that his power factor might not
>  >be that good, but I have no real idea if it is or not.  It is also 
>  >possible that Kevin's coil may work much better with a 0.375uF 
>  >cap, if it greatly improves the power factor.  It's too bad we don't
>  >know what the power factor is on Kevin's coil. Then again, it is
>  >possible that a reso or LTR cap is NG for a big coil.  It may hurt
>  >the quenching too much or something, who knows.  Maybe 
>  >someday someone will build a big coil with LTR.
>  
>  Forgive as I think out loud about Kevin's coil...
>  
>  240 volts at 100 amps gives us 24000 watts.  We'll assume we can add power
>  factor caps until the PF is good and we can use all 24kVA. 
>  
>  24000 / 120BPS gives a 200 Joules per bang.  The firing voltage for a
>  14.4kV transformer is 14400 x SQRT(2) = 20365 volts.  200 Joules = 1/2 x C
>  x V^2 so the LTR cap size is 964.5nF.  Yeah!! :-))
>  
>  If his present coil is running 138nF then he would have to increase the BPS
>  to match the LTR case.  120 x 964.5 / 138 = 839BPS.  However that is async!
>   There for he instantly has to double that to match the sync case for 1677
>  BPS (async gaps deliver 1/2 the power of sync gaps of the same BPS do to
>  firing on less than full voltage).

You're assuming that his firing voltage will remain constant regardless
of the cap size.  If this were true, Kevin's coil (0.138uF) would be drawing 
only about 3430 watts at 120 bps, (not counting losses).  Yet his coil 
draws 24kVA or so!   To me, the only explanation is that his firing
voltage must be much much higher than 20365 volts, due to a ballast
setting that gives some degree of resonant charging.  Although the
power factor may not be perfect, I don't think the true watts drawn 
is only 3430 watts.  If he was using only 3430 watts now, I don't 
think he would be getting 24 foot sparks.  So to me, this is a reality
check.... his caps must be firing at a much higher voltage than
20365.  I believe that resonant charging must often be considered 
when analyzing low bps pig powered coils.  Otherwise things just
don't add up.

Kevin is using a combo of two ballasts, I believe, and it is possible
that the ballast is lossy.  I know that when I used a variac as a 
ballast on my small coil, it doubled the input VA, for the same
spark length.  I think Kevin said that his ballast gets quite hot,
so it may be burning some power, but I have no idea how much.
If we assume that his ballast burns up 3000 watts, and his other
wiring and xfmer losses are 5% more or another 1000 watts, then
he'd be supplying still about 20kVA to the system.  If the power
factor is only 70%, then his true wattage would be about 14kW.
This would mean that his bang size is 117 Joules at his present
120 bps.  This is much larger than the 28 Joules that is predicted 
if one uses 20365 volts as the firing voltage.  Now if one uses
37kV for the firing voltage, then a figure of 97 Joules per bang is
obtained, which seems to fit his situation better.  It's still not
117 Joules, but his losses may be even higher than I figured,
which would explain the difference.  Or I may have made a mistake,
and his total input VA may be 20kVA, not 24kVA.  If this is the
case, it would drop his bang size to 97 Joules, which would
match up with a 37kV firing voltage.  Another thing that makes
me think that my analysis is correct, is that it agrees with Greg
Leyh's results.  Greg has posted his firing voltages, etc., so we
kind of know the bang size vs. spark length relationship.  We know
for instance that 3430 watts cannot give 24 foot sparks at 120 bps.

>  So to match our theoretical 120BPS  964.5nF LTR coil he has to run his
>  138nF coil at 1677BPS.  Since the RMS current in the primary cap is pretty
>  much proportional to the BPS, his "little" 139nF cap has to take 14 times
>  the current as the big 964.5nF LTR cap!  If Kevin knows his breakrate, we
>  can pretty much figure out his power factor since 1/2 x C x V^2 x BPS is
>  the real power input.  Once the cap is charged, LTR coils are just like any
>  other coil,  I would think the low BPS would help power arcing in the gap
>  considerably.

I can't understand what you mean by, "the RMS current is
proportional to the BPS".  It seems to be that if a larger cap is used
it will have have a higher peak current but a low rep rate, and the
small cap will have a lower peak current, but a higher rep rate, so
I would think the RMS current would pretty much balance out.
The only difference I see is that high bps is less efficient, so more
power has to be processed for a given spark length.  So from this
point of view, the low bps large cap case will pass less RMS 
current, but I don't see it as being 14X less.  Rather I see it as
being maybe 50% less, something like that.  Maybe I'm missing
something.  If so, I'd need it explained in more detail to understand.
>  
>  One should also consider Jeff's Model 8J which is very close to a true big
>  LTR coil.  It works darn good too!!  I am not sure if he does much with
>  power factor or really optimized it for LTR (I think it was built before
>  the LTR days) but it turned out well.
>  
>    snip...
>  
>  >
>  >Well, a low bps coil will indeed use a more expensive cap, but if
>  >it outperforms, it may be worth the cost. 
>  
>  As in the example above, if the RMS current is only 1/14, then the cost of
>  the cap may be dramatically less.  
 
Again, I don't see how the RMS current could be only 1/14, using
a large cap at low bps.  Maybe my understanding of RMS currents
for these caps is not correct.  The way I see it, a small cap, high
bps system will fire at a lower voltage, so the cap can have a lower
voltage rating.  This lower voltage rating, and small C value, will
make the cheaper.  The large cap, low bps system, may fire at a
higher voltage due to some resonant charging.  This high C value,
and high voltage rating will make the cap a lot more expensive,
but well worth the cost, due to more efficient operation.

Cheers,
John Freau
--

>  Rating TC caps by RMS current is very
>  important!!  In the past, it has been very hard to do but now we can get
>  pretty close.  The fact that the big cap manufacturers eyes get big when
>  you mention 90 amps RMS to them suggests that some of those exploding caps
>  of the past could have been avoided if we only had known.  However, the
>  "right" cap's price may have made our eyes get big 8-) 
>  
>  snip...
>  
>  >
>  >Cheers,
>  >John Freau
>  >
>  
>  Cheers,
>  
>   Terry 
>