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

Re: Arc length vs pwr



Hi Mark,
         I'd like to address some of your points...

> The discussion had began in reference to a ringing primary with a
> very light or no load (under coupled or non coupled) like a
> secondary coil. Under conditions of load (coupling) the secondary
> coil looks like a series resistance to the primary oscillator loop.

Agreed. It may be modelled as such. It can also be modelled as an 
equivalent parallel resistance. Actually, it only looks resistive 
under one set of secondary loading conditions anyway.

> If resonant rise didn't occur a tesla coil couldn't operate, the
> principles of resonance apply if its a tesla coil or just a L-C
> circuit driven by a generator.

I would class "resonant rise" in the coupled system as the action 
whereby the primary imparts a nudge to the secondary with each half 
cycle thereby transferring its energy to the secondary over a number 
of cycles - i.e. secondary amplitude builds at the expense of the 
primary (see diagrams below for single tuned circuits).

> The source of voltage in the primary circuit under condition of
> oscillation is the capacitor AND the inductor, the inductors back
> EMF is what charges the capacitor, the only reason we need the
> drive transformer is to get things started and to keep them going
> due to the limited Q of the circuit which is why the output rings
> down in level.

As you originally said, the transformer is no longer part of the 
circuit once the gap has fired. In fact, the initial conditions are: 
cap charged, no primary current. A quarter cycle after gap fire you 
have cap empty, all energy in the mag field surrounding the primary 
(some of which will be coupled to the secondary if present) and 
primary current at maximum. A half cycle later, the cap is charged to 
opposite polarity but with a bit less voltage due to losses and the 
inductor current is zero, and so it goes until the energy is mostly 
gone and gap ionization has reduced to the point where the gap ceases 
firing. It is important to realize the phase relationships in the 
circuit. The energy the cap started with is apportioned between 
inductor and cap in proportions depending on which part of the 1/2 
cycle you take a snapshot of.
    The point is, the cap alone is the energy source initially. 
Whatever it transfers to the magnetic field is no longer part of cap 
energy.
     I have measured this, and I have _never_ observed the cap rising 
beyond its initial value after the gap has fired. This was one of the 
things I checked thoroughly in my MOSFET gap modelling exercise. I 
have posted a number of photographs taken off the scope of both 
primary and secondary waveforms simultaneously to a number of people. 
I also measured the primary alone with attendant photos.

   ______/ __________
   |                 |  (A) prior to gap fire
   |                 o
 _____               o
 _____  Vc           o
   |                 o
   |                 |
    ------------------

   _______ generator in series_________
   |                                   |  (B) with AC voltage source 
   |                                   |      in series.
 -----                                 o
 -----                                 o
   |                                   o
   |                                   |
   -------------------------------------   
 
Circuit (A) is definitely not equivalent to circuit (B). The 
phenomenon of which you speak only occurs with circuit (B) because
there is an external energy source in the circuit.

> Actually, I make take a big flame for this, but
> there is really no such thing as "parallel resonance" all resonant
> circuits are series in nature and operation series and parallel
> only really pertain on how the drive signal is applied to it. In
> order to be able to measure this resonant rise with any degree of
> accuracy you'd have to be very sure the measuring device offered
> VERY LITTLE capacitive or resistive loading.

True. For example, let Lp = 100uH, Cp = 10nF and the closed-circuit
Q = say 50.  Then the ESR of the circuit is: SQRT(L/C)/Q = 2 ohms.
The equivalent parallel resistance is close to ESR*Q^2 = 5 kohms.
Measuring with a 1 Mohm probe in parallel with 20pF across the 
capacitor gives negligible loading.

Malcolm

PS - If you would like, I'd be happy to send you some photographs of 
all this.