Re: Spark Gap Losses And Thoughts...

Hi Terry, all,

> Original Poster: Terry Fritz <twftesla-at-uswest-dot-net>
> Hi All,
> Taking many observations, measurements, models... into consideration, I
> think there are two general types of gap that one should try to shoot for.
> The first is the quench gap.  These gaps try to transfer the energy from
> the primary into the secondary and trap it there by opening the gap while
> the energy is in the secondary system.  They have the advantage of reducing
> stress on the primary cap and components, and being easy to make.  However,
> they tend to have much higher resistance and will loose a lot of system
> power if they fail to quench.  Since higher currents tend to be difficult
> to quench without making a really high loss gap, quench gaps tend to be
> better on low power (~1kW) systems.  The multi-gap types made from many
> pipe sections seem to work best.

Most if not all gaps designed to quench that I have seen have a 
number of drawbacks. They invariably increase primary losses if they 
are to even come close to stopping pri-sec transfers in a primary 
null. Multi-section gaps lose it because there is a voltage drop 
associated with each gap. Blown gaps lose it because the air is 
working against efficient conduction by pushing ions aside. I have 
seen a jet of air applied to effect an ideal quench actually blow the 
gap discharge into a 3 - 4" arc (in the curved sense). This suggests 
that one really wants as few gaps in series as possible (ideal = 1) 
and wouldn't it be nice to have timed pulsed jets of air to blast the 
gap at the most desirable time?   

> The other type of gap does not rely on quenching.  They depend on providing
> a purely low resistance path for current to flow while loosing as little
> energy as possible.  They are terrible at quenching but since they loose
> less power, they can accommodate many primary to secondary energy transfer
> cycles without becoming a major loss to the system.  Rotaries and simple
> gaps fall into this category.  They have the advantage of being able to
> handle very high currents and are thus suited to high power systems.

Anything that keeps the gap electrodes cool is desirable. The best 
way to quench a gap is to have a strong attached output discharge. 

> I also suspect that if one doubles the current through a spark gap, the
> resistance is half of what it was.  In other words if you have twice the
> current you dissipate twice the power and not four times as the I^2 rule
> would suggest.  If the I^2 rule were true, as you doubled the primary cap
> size, the system arc power would not change much.  But if you assume the
> resistance of the gap drops in half, the system arc power about doubles.
> This is gleaned by observation, measurement, and computer modeling.  It
> appears such basic rules must hold true for our coils to exhibit the
> behavior they do.

Very true. The losses are VI, not I^2R. This was modelled a long time 
ago. Giving a mean resistance figure for the gap belies its true 

> The gap losses account for around half of the power that goes into a TC.
> Reducing this loss can have a dramatic effect on system performance.  So
> far, the best gaps I have used are sync rotaries with only two gaps with
> close spacing and large cool running electrodes.  Such a gap does not even
> try to quench but rather provide as low of resistance as possible.  I
> suspect that a single gap rotary would be better but that would require a
> 500+ amp AC slip ring assembly...

A single pri-sec transfer can do far better than 50%. I have measured 

> In my coil's primary, my best guess as to the primary resistances are as
> follows:
> Primary cap AC resistance 0.060 ohms (10 x 5 EMMC)
> Primary conductors including primary inductor 0.488 ohms  
> These were measured with my HP meter and from poly cap data sheets and seem
> very accurate.
> The gap resistance should be about 2.5 ohms.
> So, the primary inductor (0.25" tubing) and cap are not loosing too much
> power.  A heavier primary may help a little bit.  But the big loss is the
> gap accounting for about 80% of the primary loss or 325 watts out of 860
> into the coil and 273 watts into the output arcs (also, 130 watts in the
> neon and filter system and about another 130 into other things).

Again, gap loss might peak at 325W but the figure is not constant 
throughout the transfer period. 

> So, the gap is where to extra power is.  At full power, a TC does not have
> that many energy transfers so I do not think quenching is that big of deal
> for a system that puts out pretty good arcs.  I think one needs to work on
> making the gap as low resistance as possible...

Agree completely.