Re: The 1500t secondary myth (long)

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: FIFTYGUY@xxxxxxx 

In a message dated 12/5/04 3:20:20 PM Eastern Standard Time, 
tesla@xxxxxxxxxx writes:

> Below I'll discuss some things about coils and magnifiers.
    Thank you very much, John (and others) for once again clearing things 
up a bit for me.

> Obviously the toroid must not be so large that
> the spark cannot breakout.  Breakout points can be installed
> to permit a too-large toroid to breakout, but this tends not
> to increase the sparklength.

    Can't imagine a toroid of any size too big to not break out at all, 
but I have read here of cases where this happened. I also can't imagine why 
a lot of folks run without a breakout point. I want that darn streamer to 
go where I want it and NOT go where I DON'T want it!


> I did some tests comparing various input voltages and found only
> a minor improvement by using for example twice the voltage.  I tried
> using voltages up to about 50kV, but didn't see much difference.  It is
> true however that higher voltages are theoretically more efficient,
> and are likely to give slightly improved performance.  A lot depends
> on the coils original design.  For example if a coil has too few
> primary turn in use, it will tend to be lossy.  Then if a much higher
> voltage is used, it will permit more primary turns to be used
> (because a smaller capacitor must then be used).

    Interesting that you didn't get much performance increase from higher 
primary voltages. But why must a smaller primary capacitor be used with the 
higher primary voltage? Why not a cap of the same specs and performance but 
higher voltage rating (obviously there are cap construction tradeoffs, but 
if cost was not a factor?)?

> Richard eventually changed his mind about this and added more
> capacitance to Nemesis for a total 1.1uF value.
    Thank you for bringing me up to date... like most coils, theirs 
probably were in constant evolution.

>  It has also been found that basically what's good for a classic TC is
> also good for a magnifier;  low breakrate, large capacitors,

    Well, that kinda flies in the face of what Bert just posted... any 
wonder I get confused so easily?

> Hull used a series quenching rotary.  This design has a number
> of electrode pairs all in series.  This makes the gap both a rotary gap
> and a multiple gap in a sense.  It's true these types of gaps quench
> a little better.
    I think every rotary gap design I've seen has at least *two* actual 
gaps. Just not sure what Hull was calling the magic number. Didn't he 
advocate at least one static series gap as well? If so, why?

> In reality the actual overall coupling of a magnifier is very similar to
> that of a classic coil.  It's only the driver coupling that is tight.  This
> is because the secondary and extra coil of a magnifier "work
> together", and the coupling of the secondary and extra coil
> combination must be used to determine the actual coupling.

    I think I've got a handle on at least this aspect - you need some 
coupling, obviously, to get any energy into the secondary/resonator. But if 
it was perfectly coupled at k of 1, there wouldn't be a resonant rise, only 
what would then be the turns-ratio effects. So there has to be an 
un-coupled (from the primary) portion of the coil that can resonate. Is 
this correct?
    Interesting that the overall coupling ends up about the same between 
mags and "classic" coils. I suppose this makes sense - the "uncoupled" 
upper part of a non-mag coil's secondary is essentially the same as a mag's 
third coil, and of course the lower coupled portion is the "driver". Which 
brings us back to the post topic again :) - the extra secondary turns in 
debate here (1000 vs. 1500) are there to increase the inductance of the 
secondary, not to couple them to the primary. There can't be much coupling 
up there, and in fact the toroid doesn't have much effect as all as a 
shorted turn (in fact, base-driving my secondary with a low-voltage 
oscillator showed a split toroid made no difference at all in the 
secondary's characteristics). So I suppose a magnifier could have a 
geometry advantage, where the resonator is designed for lowest losses at 
the desired inductance. However, running all this through an inductor 
design program, taking into account skin effect, DC resistance, coil 
geometry, wire size, etc, generally gives me coils of the same general type 
as currently recommended (4-6 aspect ratio, typical size, wire turns and 
gauge). I've noticed that a 50% reduction in wire length (with 
corresponding re-arrangement of dimensions) can incur only a 10% efficiency 
penalty, and thus a big cost savings. But this also means there's a pretty 
broad range of what works well.

> In reality there is no space savings.  One must consider the
> entire Tesla coil.  When the size of the driver unit is considered along
> with the resonator, the whole thing is about the same size.

    Again, thanks for clarifying (you too, Mr. Terren!). I think I was 
confusing the old mag recommendations with the latest I got from Dr. 
Resonance - who said the greater the ratios of the inductance of the 
resonator to that of the driver, the better. But I've played with Antonio's 
MagSim, and a relatively small inductance driver (L2) requires a 
correspondingly large driver (or transmission line) capacitance. This does 
seem to jive with the last major magnifier discussion here, where it was 
mentioned that the more "transmission line" capacitance the better. 
However, MagSim says for a big coil design I should have around 600 pF of 
C2. That seems like a lot to try to obtain from a single-surface capacitor 
like a sheet of flashing, a pipe, or a coupla toroids.

> I haven't heard about those results.  The last I heard Gary got his
> best results by using a somewhat larger cap than most folks would
> think of using.  The resonant value for his system is 0.01uF, often
> LTR values of 2.5 or 3.3 x reso are recommended, but Gary got
> his best results using a 0.04uF cap.  He got about a 70" spark or
> so.  This is using a 15/60 NST.  Gary demonstrated this coil at
> Ed Wingate's last Teslathon.
    Whoops! My mistake - I misread that post! I had read it as ".004" uF, 
and was wondering if the high-breakrate small-cap idea was at work there.
    How is Gary (and you, John) getting such good performance? Gary has 
some interesting pressurized gap designs on his site, but are they what he 
is currently using?

> Hull never experimented with low breakrate 120 bps synchronous
> rotary spark gaps.

    Really? I would think that would be the obvious starting point, so I 
wonder what led them away from that.

-Phil LaBudde