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

Re: Notches and things

Subject: 
            Re: Notches and things
       Date: 
            Fri, 28 Mar 1997 07:04:11 -0800
       From: 
            Bert Hickman <bert.hickman-at-aquila-dot-com>
Organization: 
            Stoneridge Engineering
         To: 
            Tesla List <tesla-at-pupman-dot-com>
 References: 
            1


Tesla List wrote:
> 
> Subject:
>             Notches and things
>        Date:
>             Tue, 25 Mar 1997 17:35:41 -0800
>        From:
>             Skip Greiner <sgreiner-at-wwnet-dot-com>
> Organization:
>             Greiner, Ltd.
>          To:
>             tesla list <tesla-at-pupman-dot-com>
> 
> Hi Everyone
> 
> This post is partly an experiment. I made a very crude drawing in MSWORD
> which I have attempted to attach to this post. I do not know if you will
> be able to receive it. It depicts what I think are notches and I wanted
> confirmation.
> 
> The following applies to my new 17.5 x 24 "secondary which I described a
> couple of days ago.
> 
> With the toroid sparking the output frequency is about 260khz which is
> right on the non powered resonant frequency of the sec/toroid
> combination. The strange thing is that the ion cloud has not lowered the
> frequency. The frequency varies little if at all from low to high power
> inputs.
> 
> At very low power inputs where the toroid is just breaking out about 3"
> sparks there are four obvious notches before the quench. As I increase
> the power to maximum with the streamers out to 56", the quench seems to
> take place at the second notch. These two statements are on the basis
> that a notch is what I think it is.
> 
> I obtain the scope pattern with the scope placed about 10' from the TC.
> The probe is connected to a 2' vertical wire attached to a ceiling
> rafter. The horizontal scan is set slow to display the energy packets
> which are at the mains frequency.
> 
> The interesting part is that the rotary is only running at 1800rpm and
> the .25" contacts are on a 3.5" radius which gives a very long dwell
> time in the millisecond range. My question is how can I be getting a
> quench at the second notch with this long dwell time or don't I know
> what a notch is.
> 
> Comments appreciated
> 
> Skip

Skip,

Excellent question! The waveform you're seeing can be likenned to a 100%
amplitude-modulated signal with a superimposed exponential decay. The
"carrier" is the resonant frequency of your secondary/toroid. If you
didn't quench at all, you'd see the "envelope" of the carrier look
something like the crude ASCII waveform below, only perhaps with more
"humps". The gap "fires" at T0, and transfers maximum energy to the
secondary halfway at point TQ1. If you don't extinguis the primary gap
at this point, energy can now couple back from the secondary to the
primary until there's no more secondary energy left (T1). The process
then reverses, transfering energy back into the secondary, and so on. 
Note that the primary energy is at a minimum at points TQ1, TQ2, and
TQ3, corresponding to energy maxima in the secondary. For "magic" values
of the coupling coefficient "k", the primary energy will go to zero
Joules at these points, and the gap will self-extinguish. When the
secondary starts dumping energy back into the primary (TQ1 to T1), the
gap may reignite, and we will have to wait until the next pruimary enegy
minimum point (TQ2, or later) to quench.

However, during all this time, energy is also being lost. It's this
energy loss that accounts for the lower amplitude of successive peaks.
The "big" losers are the primary gap and any streamers you've got coming
off the top. Other significant losses come from skin effect resistance,
groundpath resistance from the base of the secondary, capacitor
dielectric heating losses, unwanted corona, and EM radiation. 


T0       .      T1              T2              T3 
      .     .             .
    .         .       .       .                 
  .             .  .            .     .    .  
 .               ..              . .           . 
-------------------------------------------------.--------------
 .               ..              . .           .               
  .             .  .            .     .    .                       
    .         .       .       .
      .     .             .  
         .
    
        TQ1             TQ2            TQ3 


As your streamers increase in length, the amount of energy being "lost"
to reinitiate and maintain them increases. This means that under fully
formed heavy streamer loading, the rate of decay of the enevelope above
will markedly increase. 

If enough energy has been lost from the secondary, there won't be enough
to reignite the primary gap, and we'll see the quenching take place
EARLIER, with NO change in mechanical dwelltime. On my system, the
static vacuum gaps will quench at TQ3 typically, but once I'm really
cookin' with heavy streamers, this goes to TQ2 and sometimes TQ1. This
means that the actual "dwelltime" now becomes a function of streamer
energy loss and becomes virually independent of simple electrode
overlap. BTW, this is one reason why the Milwaukee Museum tandem coil
still gets excellent performance even though it has massive, and
"theoretically" poorly quenching, rotary electrodes. This is one of the
many areas in coiling where theory must be tempered by experimental
wisdom...

Hope this helps!

Safe coilin' to you, Skip!

-- Bert --