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

Re: High Tank Circuit Q

Hi Fellow Coilers,

Tank Q for discussing tank Q (sorry).  You know I just can't resist jumping in on a discussion of this 
topic.

It is my humble opinion that the spark output difference Scott is experiencing with .0125uF vs. .025uF 
is strictly due to the difference in the Lp/Cp ratio of the tank.  When you halve the capacitance, you 
must double the inductance which raises Lp/Cp.  This causes Qp to about double, resulting in a great 
increase in power transfer efficiency.  The problem is you can't take this to an extreme.  You can't 
have 10pF and 10H (or whatever) for example because you can't get any power into 10pF.

I believe that the best way to design is to wind a secondary at the lowest tolerable frequency, which 
also means very high self inductance.  This also means that the primary winding will need to have 
comparatively many turns for lower frequency operation given a certain capacitor size which matches 
well with the power transformer.  A many-turn primary does not result in a poor turns (inductance) 
ratio because the secondary inductance is huge.  Given Cp, ALL inductance ratios come out THE SAME for 
any size secondary.  It all washes.  Strange, but true.  So why not go for the configuration that will 
produce the best primary OSCILLATOR.

Following the above design philosophy results in the highest attainable primary Q for any power level. 

>It would still be true that the pulsing would be half as many with twice the capacitance.  But, 
>with twice the capacitance I should get twice as many Joules per pulse.  The problem is that I 
>didn't have enough current available for the greater capacitance. The 15,000 VAC -at- 90 mA is fine 
>for .0125 Uf, but it is about half of the current needed for .025 uF, at 15,000 VAC.  I had 
>made an oversight on the basics of matching the tranformer to the capacitor.  Don't I feel like 
>the fool! :-/

This is not entirely true.  True, a large cap will not charge all the way to the transformer rating, 
but the POWER goes in nonetheless.  You don't lose anything.  P still equals 1/2CV^2.  Essentially, 
the transformer just turns into a current source.  I believe the observed increase in performance is 
mainly due to the reasons given above.

About balanced tank circuits:  (The Tesla "Equidrive" circuit as some call it)
I have seen the same small increase in performance from balanced vs. unbalanced as Scott demonstrated. 
 In testing this, you cannot just remove one cap, re-tune and re-fire.  This changes the cap value.  
You must keep 2 caps in series and experiment by moving both to one side of the circuit.

I have measured circuit Q's of balanced vs. unbalanced circuits and have found about a 10% improvement 
in Q for the balanced.  I believe that this is strictly due to symmetry.  Kind of like having a tuning 
fork with unequal vs. equal length tines.  Intuitively, any oscillatory system is going to oscillate 
better (read longer) if it is balanced in every way.  It is up to the individual to decide if 10% more 
Q is worth the extra cap and the extra danger.

Zap,
Mark

P.S.
Judging by the posts and my own mail, your results have produced a benchmark that others are striving 
for.  Congrats again.