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

Re: Calculation of Q



-- BEGIN included message





Hi all,
           I would like to say hear hear to what Mark has written on 
this...

>   TESLAC calculates coil Q by first determining the RF (not DC!!)
> resistance
> of the wire used in the winding, and then dividing that value into the
> calculated
> reactance of the winding at the coils self resonant frequency. This is
> the 
> classical method of Q calculation for series resonance. Then the Q value
> optained is "derated" in a attemp to compensate for the coil form,
> coupling to
> the primary and such other things that reduce the effective Q of the
> coil.
>  TESLAC calculation of Q is NOT absolute, and no other programs
> calculation
> is either, they are best used in comparisons between one coil design and
> another.  The main thing I've found in high Q coils is LOTS of inductance
> and
> coil diameter,  it rather seems that any wire size that will allow 800 to
> 1000
> turns close wound at a H/D ration of 2.5 to 3.5 to 1 will result in a
> pretty decent
> secondary coil in a standard tesla coil.
> 
>             Mark Graalman

In fact, Rac is further modified (increased) by the proximity of the 
turns to one another so the derating is valid for that reason as much 
as anything else.
     Some time ago I measured a range of coils with H/D's ranging 
from 5 to about 0.5, both close wound and spacewound. I scored the 
highest Q (around 500 taking generator impedance into account) in a 
space-wound coil of H/D=1. However, its Q dropped dramatically with 
the addition of a terminal - clearly useless for TC work. Other 
measurements (and a theoretical analysis) agree with Mark's guideline 
above for coils with a top hat.

Aside:
Funnily enough, that is precisely the recipe for Tesla's final 
Colorado Springs extra coil although he did add rather a lot of 
capacitance with a topload (and a hell of a lot more with the wire 
leading up to the sphere). I was just mulling over this in Richard 
Hull's guide last night and realized that Tesla never really got to 
grips with this. I stand to be corrected here but I don't think he 
ever measured the effect of the wire alone before ading a terminal to 
it. It would not have given an absolute either because connecting the 
sphere would have modified the L/C distribution in the leadup wire.
End of Aside.

     A high Q does not of itself guarantee performance. It simply 
gives an idea of the *losses* in the coil. Q's in toploaded 
secondaries are usually in the 150 - 350 range depending on winding 
type so there are few losses there. Concentrate on the primary is my 
best advice. There is still debate about the best value for secondary 
Zo (L/C ratio). Of the 18 or so values of coils I have examined so 
far, the bare secondaries have Zo ranging from about 20kOhm to 100kOhm
but with the terminals the builders used, these ranged from about 
15kOhm to 50 kOhm. I haven't correlated relative performances yet. 
The calculations are rather tedious and I just want to get the chart 
finished at this stage.
     A further piece of info on this exercise: the use of a high 
Cterminal/Cself ratio in the secondary seems to be well justified in 
terms of spark output for power input from the figures I have. So far 
the highest ratio is about 5 but I have many more coils to work 
through and I have not looked at Nemesis yet.

Malcolm






-- END included message