# [TCML] poor coil performance again- help!

bartb bartb at classictesla.com
Thu Apr 24 20:05:58 MDT 2008

```Hi Thomas,

Regarding LC values, there are maybe preferences by some, but ratios,
not so much in the secondary. It seems larger L in the primary helps to
keep gap losses lower than they would be for a small L primary due to a
higher impedance (affects current peaks). But this is still speculation
and it may simply be the gap itself for currents involved.

In the secondary, energy is stored in the inductance and capacitance of
the coil. Obviously, more L or C, the more energy that can be stored,
but "how" you increase those values makes a difference and in my little
head, geometry (shape and size) plays a "very" significant role. For
example, say I have a 5:1 h/d coil using 24 awg at 1111 turns (5" x
25"), and thus L, C, dcR, acR, Q, etc.. all have some value.

Case 1 = I double the "length" of the coil (5" x 50", 2222 turns):
From the original coil, L increases 200%, C increases 150%, dcR
increases 200%, acR increases 135%, Q decreases 110%.
Even though I doubled L, I introduced large losses both at DC and at RF.
Although L and C increased, the coil is lossy.

Case 2 = I double the "wire length" to match Case 1 but return h/d to
5:1 (10" x 50", 2222 turns).
From the original coil, L increases 200%, C increases 200%, dcR
increases 200%, acR decreases 115%, Q increases 115%.
Better than Case 1 and the reason is the geometric shape as far as h/d
is concerned. The coil will work, but still lossy.

Case 3 = I double the coil size as in Case 2, but now I increase the
wire size to have the same turns as the original (10"x50", 1111 turns).
From the original coil, L increases 150%, C increases 150%, dcR
decreases 150%, acR decreases 130%, Q increases 130%.
Here is better coil than Case 1, Case 2, or the original. It's h/d is
identical to the original, but it's geometric size has changed. We get a
good LC increase, some major reduction in DC and RF impedances, and a
nice jump in Q.

The LC ratio is not that important to me as is the geometry. Both L and
C store and release energy. The larger the coil is, the more energy that
can be stored and released, but note that it's "shape" or h/d is also
important for maximizing not only LC but also minimizing losses. The
basic guides like the range of turns, h/d, etc. are mostly empirical in
that a range of h/d has been found to work well via experience, but it
also can be supported with numbers. Someone only looking at say the
above data might think that larger wire is always better, but that is
"not" the case. Turns, wire size, and coil geometry are important
aspects for an efficient coil. These aspects will dictate LC ratios.

Your 22 awg is perfect for a nice range of coils. Suppose you select a
1000 turn coil and you want to use 22 awg. 22 awg is .0279" including
insulation. So, .0279 x 1000 = 27.9" length coil. Say you want a 4.5 h/d
ratio. 27.9" / 4.5 = 6.2" diameter. This is a decent medium sized coil.
So look for a form similar in diameter. It might end up right at 6".
That's ok. It only means the h/d ratio will slightly increase to 4.65
which is fine. You end up with 6" x 27.9" winding length at 1000 turns
using 22 awg. Run some numbers and you'll find that a 12/60 NST running
a static gap will be LTR at .02uF. This coil would need about a 10 turn
primary. Throw a decent airflow gap in and you'll have a nice coil. This
coil would have a Q a little above 300 which is pretty good.

If you go with higher power on "this" coil (which it could handle 200mA
nicely) and if you try to stay LTR with the transformer, then the
primary turns decrease, but that's the trade-off and it's not
necessarily a bad one if kept within reason.

BTW, above when I say "energy is stored", don't misunderstand. It's not
stored over several energy transfers or cycles, it's stored during a
single energy transfer event. The remaining stored energy simply
transfers back to the primary tank circuit minus losses encountered then
back to the secondary until there's no more energy to keep the gap
conducting. For C, energy is stored in the turn to turn capacitance,
turn capacitance to ground, top load, external objects to turns, etc..
For L, energy is stored in the magnetic field.

For frequency, the lower the better as RF losses are reduced. But this
isn't usually a big worry unless the wire size is small and coil is
small. Frequency affects AC resistance of the coil and Q. Some will say
Q isn't important because when the coil sparks, Q drops like a rock.
Well, yes their right, but I don't care about Q during sparking, I care
about Q only after the gap quenches and before a spark occurs. So for
me, it's important to keep secondary Q as high as possible and to
totally ignore spark time Q because it's not relevant to the coil in any
way. The off-time is when Q has a clean slate and if the secondary has
an "off-time" Q of 300, then that is a nice number to have. 300 Q in a
medium sized coil is what I personally shoot for (or better).

Take care,
Bart

Ryckmans, Thomas wrote:
> Thanks, I think I will just design a new one. By the way, are there
> specific rules about ratio of L and C in the primary and the secondary?
> In other words, for both resonant cicuits, what should be the relative
> contribution of L and C to resonance? And what resonance frequency
> should I aim for? As low as possible?
>
> Many thanks
>
> Thomas
>
```