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

Re: Fw: Arc Impedance Study - Computer Models



Hi Malcolm,
    Questions below:

-----Original Message-----
From: Tesla List +ADw-tesla+AEA-pupman-dot-com+AD4-
To: tesla+AEA-pupman-dot-com +ADw-tesla+AEA-pupman-dot-com+AD4-
Date: Tuesday, October 13, 1998 8:59 PM
Subject: Re: Fw: Arc Impedance Study - Computer Models


Original Poster: +ACI-Malcolm Watts+ACI-
+ADw-MALCOLM+AEA-directorate.wnp.ac.nz+AD4- 

Hi Barry,

+AD4- Original Poster: +ACI-B+ACoAKg-2+ACI- +ADw-bensonbd+AEA-erols-dot-com+AD4- 
+AD4- 
+AD4- Hi Malcolm,
+AD4-     Fr and k limit the speed agreed.  The surge impedance limits the
+AD4- amount of the delivery.  The higher Xp is the lower the primary losses
+AD4- agreed.  As the surge impedance goes up the current peak goes down,
+AD4- the energy lost in the spark gap goes down,  and the peak energy in
+AD4- the primary circuit decreases.  The peak energy in the tank scales as
+AD4- the square of the peak current.  The peak spark gap energy loss scales
+AD4- as an inverse 0.67 power of the peak current.  The net energy in the
+AD4- tank deliverable to the secondary, with loss to the spark gap,
+AD4- increases as the surge impedance decreases with the frequency kept
+AD4- constant.

Not sure I entirely agree. Tank energy is defined by what was 
originally in the primary cap isn't it? If one starts with the same
amount of energy and uses a high surge impedance in the tank there 
has to be more energy transferred to the secondary if k and Fr remain 
the same since tank losses are lower (Y-N)?

    I think I meant to say instantaneous instead of peak.  I was thinking
of how efficient it might be to couple the energy to the secondary system
with one pulse as opposed to many transfers in spite of spark gap losses.
Since this is not possible without overdamping the primary and wasting
energy it would be better to use base feeding from a controlled source.  

+AD4-     To transfer maximum energy to the streamer in minimul time the
+AD4- tank surge impedance should be equal to the impedance of the streamer
+AD4- reflected into the primary tank, should it not?

But the streamers don't really start until a transfer is near 
completion in most systems in which case the secondary and its 
topload are the source of streamer energy (Y-N)?

    Absolutely, for a conventional system+ACE-  A resonator would have to
have an L/R faster than the streamer rise time.  This would be in the
megahertz region.  Such a resonator would have too few turns to withstand
the voltage gradient imposed on it.  There is no way you could couple a
conventional primary system to such a resonator.  It would have to be base
fed like a magnifier.  The secondary that was base feeding this fast
magnifier would have to be equally as fast with the same L/R constraint
(together).  It would also have to be base fed.  The system could be
defined by the parameters derived by Bieniosek (Triple Resonance High
Voltage Pulse Transformer Circuit, Review of Scientific Instruments, 61,
1990, pages 700 to 702).  With this system the streamer could be fed in
real time.  The question I have is would the losses of such a system be so
great as to limit the producition of streamers up to 100 times the
resonator length??????

+AD4- Since the streamer impedance changes with time should not the primary
+AD4- tank capacitor also change with time?  How would one do such a thing?

But are you matching a tank capacitor with the streamer? Indeed, are 
you matching the conducting gap with a streamer? I used to wonder 
about both of those things until experiments showed that they don't 
really play a part once energy is in the secondary. Whether the gap 
quenches at that point or not then depends on (a) whether one can 
make it do that if output streamers are weak or nil and significant 
energy remains in the secondary, or (b) it going out anyway 
as a strong streamer sucks most of the energy out of the secondary.
The comparison between a TC and continuously loaded double tuned 
transformer in a radio set is weak for a number of reasons, not the 
least of which is the transient nature of the load in a TC.


    If you feed the streamer it will grow.  I do not know if their are any
limits other than a sustaining field.  I do not think that a simple
capacitor tank system will work.  I think that the secondary itself will
have to be base fed by a pulse forming network producing a waveform that
will result in a best match to the changing impedance of the streamer at
the toroid of the resonator.  The streamer has to absorb the energy as it
is transfered in real time.  If the streamer fails to do this the magnifier
will track catastrophically as can be seen in T. C. B. O. R. tape number 29
and anticipated by Rose and Greenspan (MEDEA II:  A two-pulse Electron Beam
Generator, IEEE pulsed power conference, 1985, pages 139 to 142, Loc. Sit.
Page 140, last paragraph).

  I have thought of doing this with capacitors but high voltage capacitors
of sufficient size are expensive.  Also, the risetime increase costs 1/2 of
your energy at each stage (Kurtz and Corcoran, Introduction to Electric
Transients, 1935, page 28, first paragraph contains the derivation).  The
process would appear to be horribly inefficient+ACE-

?
Malcolm 
+ADw-snip+AD4-

Barry