Re: Continued Problems (fwd)

---------- Forwarded message ----------
Date: Sat, 23 May 1998 17:20:50 GMT
From: Jim Fosse <jim.fosse-at-bjt-dot-net>
To: Tesla List <tesla-at-pupman-dot-com>
Subject: Re: Continued Problems (fwd)

>Date: Sat, 23 May 1998 00:18:38 -0500
>From: Bert Hickman <bert.hickman-at-aquila-dot-com>
>To: Tesla List <tesla-at-pupman-dot-com>
>Subject: Re: Continued Problems (fwd)
>Bart and all,
>Assume that the main gap is across the HV transformer. In a resonant
>charging system there are two independent resonant conditions at work.
>The first condition, low frequency resonant charging, ONLY occurs when
>the main gap is open, AND when the "effective" secondary inductive
>reactance of the HV transformer is close [in magnitude] to the
>capacitive reactance of the tank cap at the mains frequency of 50 or 60
>Hz. As long as the main gap is not conducting, the tank capacitor
>voltage and charging current will climb on successive mains half-cycles
>until the main gap fires {or the transformer or tank cap gets overvolted
>and fails). If the main gap fires, this "spoils" the Q of this resonant
>system, and the resonant charging process begins anew once the gap is
>extinguished. Ideally, the gap fires _at least_ once every half-cycle,
>reducing the possibility of "letting the smoke out of your system".
>The second condition occurs ONLY while the main gap is firing. In this
>case the heavily-conducting gap effectively "shorts out" the transformer
>L, leaving only the primary/tank inductance and capacitance to resonate
>at the operating frequency of the coil. The HV source is effectively
>"out of the picture", and will not influence the high frequency tuning
>of your system. While the two resonant conditions are basically
>independent and mutually exclusive, non-linearities (mostly transformer
>or choke core saturation) may also introduce chaotic operation and
>"bumping" in real-world coils...
	Additionally, when the gap IS conducting, the ballast inductor
is storing energy: E = 1/2*L*I^2. (in my case. E=1/2*30mH*20A^2 = 6J)
When the gap opens, the  current from the mains transformer stops and
the field in the ballast inductor collapses. This causes a very high
induced voltage to be  applied to the primary of the mains transformer
which charges the primary cap to a higher voltage than it otherwise
would be charged to. Given E=1/2*C*V^2 and solving for V:
6J=1/2*16nF*V^2 => V=27.4kV on top of what ever voltage the mains
transformer is applying to my primary cap.

This topology is that of an open loop boost mode switching power
supply.  If the break rate matches the resonant frequency of the
ballast inductor, as transformed through the HV transformer, and the
primary cap, you then have the topology of a quasi-resonant switching
power supply.

During one of my test runs, I placed my ballast inductor to close to
the grounded case of my RFI filters. 1/4" arcs jumped from the last
turn of my ballast to the grounded case. Needless to say, 220V -at- 20A
across a 1/4" gap was quite impressive. Additionally, I found that I
had to place a 50uF PFC cap at the ballast RFI filter junction to the
other side of the 220V line to prevent this induced voltage from
blowing the filter caps inside the 600V rated RFI filter. I took out 2
of the 4 sections of this filter before it dawned on me what was going
on - apply power; bam! my 50A circuit breaker opened because of the
now shorted RFI filter.