RE: Quenching Theory Question (fwd)

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: List moderator <mod1@xxxxxxxxxx>



---------- Forwarded message ----------
Date: Fri, 18 May 2007 21:04:36 -0400
From: "Breneman, Chris" <brenemanc@xxxxxxxxxxxxxx>
To: Tesla list <tesla@xxxxxxxxxx>
Subject: RE: Quenching Theory Question (fwd)

This information was really helpful also, but really what I'm getting at
is somewhat more specific, relating to capacitor charging through a gap.  
Here's the overall picture: I was thinking of building a DC coil, and was
looking up various ways to do so.  The two ways that I found were to use a
charging reactor (which would have to be large and have to support a high
voltage) or to use a charging resistor (dissipating significant power).  
From what I can see, these are necessary to prevent shorting out the power
supply (presumably containing parallel capacitors) when the gap fires.  I
thought that there might be a better way to do this than these methods, by
physically breaking the connection with the power supply when the gap
fired.  The only way I could think of doing this was to use a modified
ARSG, with three contacts.  One contact would be connected to one lead of
the tank capacitor, and it would always be within gap firing distance of
the rotor, which would alternate sparks between two different sets of
contacts, one set attached to the power supply and one set attached to the
primary.  The other power supply lead, tank cap lead, and primary lead
would be connected together. In addition to solving the shorted power
supply problem in DC coils, this design would also have an advantage over
traditional gap AC coils, because there would be no loss due to the power
supply being shorted at peak voltage when the gap fired.  This would be
particularly advantageous for power supplies employing voltage doublers
because the capacitors in the doubler wouldn't completely discharge.  It
would also be an improvement over DC coils with a charging reactor because
the break rate could be varied continuously without changing the charging
reactor.  I have actually ordered parts to build such a gap, and have been
working on several equations to describe gap operation. I can't think of
any problems with this design, except for the issue with the capacitor
charging through a spark gap.  This is where my question was going.  
Because the DC power supply would have a capacitor in parallel with the
power supply of a significantly higher capacitance than the tank
capacitor, when the gap fired to charge the tank capacitor, the initial
current would be high.  My concern is that the tank capacitor might not
charge to the same voltage as it normally would with a charging resistor
or reactor.  So, what are the issues with charging the tank capacitor
through a gap? Also, as soon as the parts for this gap arrive, I'll
construct it, test it, and let you all know how it works.  I'm just very
curious about the aforementioned issues with charging a tank capacitor
through a gap.  I've derived several equations regarding the functionality
of the gap, bang size, power throughput, etc., but none of them take into
account the capacitor charging through the gap.

Thanks a lot, Chris



-----Original Message----- From: Tesla list [mailto:tesla@xxxxxxxxxx]
Sent: Fri 5/18/2007 5:06 PM To: tesla@xxxxxxxxxx Subject: Re: Quenching
Theory Question (fwd)
 
Original poster: List moderator <mod1@xxxxxxxxxx>



---------- Forwarded message ----------
Date: Fri, 18 May 2007 14:58:47 -0500
From: Bert Hickman <bert.hickman@xxxxxxxxxx>
To: Tesla list <tesla@xxxxxxxxxx>
Subject: Re: Quenching Theory Question (fwd)

Tesla list wrote:
> Original poster: List moderator <mod1@xxxxxxxxxx>
> 
> 
> 
> ---------- Forwarded message ---------- Date: Fri, 18 May 2007
> 11:19:35 -0400 From: "Breneman, Chris" <brenemanc@xxxxxxxxxxxxxx> To:
> tesla@xxxxxxxxxx Subject: Quenching Theory Question
> 
> Hello,
> 
> I'm trying to figure out some simulation parameters for a spark gap,
> and had a few questions about arc formation and quenching.  Is it
> correct that the point at which an arc forms is entirely dependent on
> the potential difference between the contacts?  

No. It's also dependent upon the recent history of the gaps, since 
residual ions, hot (i.e., rarefied) air, and thermionically emitted 
electrons (from incandescent spots on the gap electrodes) will all 
conspire to reduce the effective breakdown, or "restrike", voltage of 
the gap. This is why an AC arc, once formed (such as during welding) can 
reignite on every half cycle of the AC waveform even though the 
available open circuit voltage across the gap is much lower than would 
be required to initially bridge the same gap. Studies have shown that 
the gap restrike voltage increases in an exponential fashion, rising 
asymptotically toward the full ("cold") initial breakdown value. For 
spark gaps operating in air under standard conditions, full recovery may 
take as little as a few milliseconds, or it can be orders of magnitude 
longer between previously incandescent electrodes, and forcing cool air 
into/around the gap or using will help to reduce recovery time.

> I've seen a lot of
> tables that relate arc length to voltage, so if this is correct, I
> could find out all of the arc formation parameters myself.  

You've likely seen tables showing the spark breakdown voltage of an air 
gap (usually measured between spherical electrodes). The voltage drop 
across a sustained arc will be considerably lower - typically 10's to 
hundreds of volts depending on the arc current, electrode separation, 
and electrode materials. For low to moderate currents, the voltage drop 
is an inverse function of current - the higher the current, the lower 
the resistance, sometimes called a "negative resistance characteristic". 
 From a modeling perspective, an AC arc behaves similar to pair of 
back-to-back Zener diodes in series. The arc voltage drop is mostly a 
function of electrode materials and gap length. The Zener-like behavior 
is due to fairly constant a voltage drops near the anode and cathode 
electrodes (anode and cathode drops) and a voltage drop across the 
remaining arc channel that also tends to be relatively constant for a 
given current. Although the arc has a negative resistance 
characteristic, it always inserts an overall positive resistance into 
the circuit path. As such, the conducting arc always "steals" energy 
from the circuit it's imbedded within.

Following are some spark gap models and more information that you may 
find useful:
http://www.spectrum-soft.com/news/winter99/sparkgap.shtm
http://www.intusoft.com/nl50.htm#SPARK%20GAP%20MODELING
http://www.hmi.de/people/boenisch/articles/esd_protection_device.pdf

> Also, I
> was wondering what determined when an arc quenched.  I know that once
> a gap fires, the voltage can drop and/or oscillate significantly
> without the arc extinguishing, but how low can the voltage/current
> drop before a typical arc extinguishes for a given distance?  

Sustaining current is typically around a few amps. Once the arc is 
extinguished, the air begins to rapidly recover its original dielectric 
properties. In an AC circuit, whether the arc is reignited depends on 
whether the voltage applied across the gap can rise more quickly than 
the dielectric strength of the gap recovers. In a TC primary, most gaps 
cannot recover quickly enough to quench at the first notch unless there 
is very heavy streamer loading (such as a strike to ground). With open 
air streamers, a well designed gap may recover at the 2nd or 3rd current 
notch. If a gap is heavily overloaded, it may not recover sufficiently 
to even allow normal tank cap recharging from the mains between "bangs". 
When this occurs, you get fiery "power arcing" in the gap, typically 
accompanied by incandescent electrodes (and very dismal coil 
performance).

> The
> distance involved here is going to be only a few millimeters, so
> that's what I'm interested in. Like, how low could the
> voltage/current across a gap of a few millimeters get before
> quenching, with moderate air flow? 

It's hard to predict beforehand since it depends on many factors. These 
include the thermal mass of your gap electrodes, efficiency of air 
cooling, peak current in the tank circuit, the degree of coupling to, 
and the degree of streamer loading from, your TC secondary. Techniques 
such as using more inductance in your primary (i.e., higher surge 
impedance), using gap electrodes with significant thermal mass, making 
them from a good thermal conductor such as copper, and providing ample 
cooling for both the gap and electrodes will all help to improve 
performance. With a sufficiently high velocity air flow, a single gap 
can be significantly more efficient than a series ("quenching") gap that 
uses multiple gaps. You can estimate an average voltage drop of perhaps 
200-300 volts across each conducting spark gap.

I also had a few questions about
> resistance encountered in a gap.  Is the resistance generally
> constant with a given gap width?  

Yes, for a given current level...

And what kind of resistances are
> usually encountered?

The dynamic resistance seen during primary current peaks can range from 
several ohms (for a small coil), to a fraction of an ohm for large coils 
operating at kA-level primary currents. YMMV...

Hope this helped and good luck on your system.

> 
> Thanks a lot, Chris
> 
> 
> 
> 

Bert
-- 
***************************************************
We specialize in UNIQUE items! Coins shrunk by huge
magnetic fields, Lichtenberg Figures (our "Captured
Lightning") and out of print technical Books. Visit
Stoneridge Engineering at http://www.teslamania.com
***************************************************