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

Re: Fwd: [jlnlabs] TESLA COIL REVISED6



Original poster: Bart Anderson <classi6-at-classictesla-dot-com> 

Hi CK,

Tesla list wrote:

>Original poster: "C. Kollett" <ckollett-at-falconaerosystems-dot-com>
>Hi,
>Although I am a newcomer to the list, I respectfully disagree that several
>hundred turn secondaries reduce the gain. I built a 20" dia x 48" L coil
>with about 560 turns of 20G Teflon covered wire. Resistance was only a few
>ohms, compared to the inductive reactance which is quite high. One of the
>equations for Q is "Q = X(subL)/R".  If for example the inductive reactance
>is 1K ohms and R is only a few ohms you will have a very high Q like 200! In
>my opinion Q is what its all about along with a lot of drive power rich in
>harmonics.
>A Tesla secondary impedance could be complex in that it is a lumped
>distribution of inductance and capacitance. However at resonance the total
>reactance is highest and it is the point where capacitive inductance equals
>inductive reactance.

Your coil would have a DCR of 30 ohms and reactance of 34182 at resonance 
(roughly 125 kHz unloaded).
The Q of XL/R you describe is textbook and would be 1139. This would be 
true if there weren't other losses involved (lots of losses). The equation 
does not account for losses (it is the "ideal" case). If you measure Q at 
resonance, you would measure a much lower number, probably between 100 and 
300. However, when the sparks start, Q plummits like rock in a shallow creek.

There are methods for measuring Q in such a dynamic system as a Tesla coil. 
If you want to do the measurement sometime, let us know. The method 
involves an "indirect" sampling of Fres above and below the 3dB points and 
applying Q = Fr/(Fhi-Flo). The generator must have the lower output 
impedance you can muster (ideally 0). I built a low z-amp for this purpose 
whicih our Moderator Terry designed. The generator feeds the low-z amp, the 
low-z amp output feeds the base of the secondary. A scope probe is dangled 
as far away as possible to reduce loading. That's the basics. More details 
if interested.

>Also for what its worth (maybe nothing at all!) Has anyone tried an
>additional resonant coil (or coils) coupled loosly in the proximity of a
>running coil which "should" add to the overall Q due to mutual inductive
>coupling(i.e. multiple tuned circuits interacting with one another). Im not
>an expert coil builder but I do have some experience building a fairly large
>coil and have worked in electronic most of my life.

I posted a maggy design a couple days ago (All components are now built but 
I've modified the design which is quite a challenge). If you can imagine 
12.75" x 15.125" coil (18g, 350 turns) surrounded by a helical primary 
(0.375" copper tubing, 11 turns total to be tapped), and sitting directly 
above is a 3rd coil (4.5" x 20.165", 24g, 900 turns), topped off with a 
toroid. It's kind of a magnifier without the long transmission line.

This isn't easy. The capacitive loading of the upper coil and topload 
dramatically affects the lower coils (Fres). I've found and confirmed by 
measurement I can use Javatc to model Fres with this coil and topload. I 
simply treat the topcoil as a cylinder topload object at it's position 
(along with the rest of the system). Fres comes out correct. But it's still 
a challenge to get the all 3 coils at the same Fres (lot's of fun trying!).

When I finalize it, it should make a nice compact maggy. I turned it on for 
a quick test today. Arcing between bottom 2 coils immediately (insulation 
time, but I'm off to Neveda again tomorrow, so it's have to wait a couple 
days). I would almost bet there are a few others who have tried this.

Too bad the program can't model multiple coils (that would make this 
challenge go faster [number crunching]).

Take care,
Bart