Re: Sparks jumping from 4' to 9' (was Re: 20 joules)

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Steve Ward <steve.ward@xxxxxxxxx>

Hey Steve,

Some comments:

On 8/8/05, Tesla list <tesla@xxxxxxxxxx> wrote:
> Original poster: Steve Conner <steve@xxxxxxxxxxxx>
>
> Hi all,
>
> I was thinking about Steve Ward's observation that the sparks on his
> coil suddenly grow in length from about 4ft to about 9ft as the input
> voltage is increased.

Did some more tests with a bigger 9.5"x44" toroid:

http://www.stevehv.4hv.org/bigtoroid/123_2397.JPG

I still didnt exceed 10 feet, despite the bigger toroid and nearly 1kW
extra input power (4400W now) :-(.  The effect of the sharp change in
output seems to depend on tuning as well i found.  If the primary is
tuned higher (to where i produce notches) then the knee is softened,
but only because it produces longer sparks with smaller bangs at the
low input voltage.  As i increase the input voltage, the primary needs
to be tuned slightly lower to compensate for the big streamer,
otherwise the streamers dont grow as long.  So when the primary is
tuned to support the big streamers, then its just a bit out of tune at
low power, making the knee more noticable.

>
> Then I thought about an old post from Marco DeNicolai where he says
> that discharges longer than about 1 meter grow by a different
> mechanism that needs only 1kV/m electric field, rather than 5kV/m for
> smaller gaps. (are these figures really right? they seem very small)

Those E-fields do seem pretty weak, but maybe when you are 2 meters
out from the toroid, the E-field gets pretty low.

>
> 4ft is about 1 meter, and 9ft is a bit over twice 4ft, and 5kV
> divided by 1kV is (a bit over twice) squared. So I'm thinking maybe
> Steve is seeing this transition as he comes up on the voltage that
> produces a 1 meter discharge?

Possibly so.  But now i remember something from my tests last summer
that used secondary base feedback.  It turned out that i could graph
an almost linear line of input V vs spark length.  But, my operating
conditions were much different then.  I used 350uS driving periods,
allowing ample time for the driver to pump plenty of energy even into
the smaller leaders.  Now i only use about 150-180uS of driving time.
Maybe i should do a test using 350uS (if i dont blow the breakers) to
see if the knee dissappears...  I do have a suspicion that ON period
could play a big role here.

>
>
> If this hypothesis were true, it would imply:
>
> The effect would only be seen on coils that give all bangs the same
> size. It would not be seen on coils with async rotary gaps, for
> instance. If the bangs are assorted sizes, some will be big enough to
> push it beyond 1m and others won't so I expect the transition will be
> a lot more gradual.
>
> It would also only be seen on coils with a long enough energy
> transfer time to allow growth of a >>1 meter discharge in the first
> place. If the field is all depleted by the time the tip makes it past
> 1 meter, nothing much is going to happen. Hence why I never saw it on
> my OLTC2, which is rather too short at only 60us.

But i didnt notice it when using extra long bangs?  But i should do
the test mentioned above to verify.  The knee seems sharper with
shorter ON times... at least for my coil.

>
> DRSSTCs tend to have lower peak power and longer energy transfer
> times than spark gap coils so maybe this is why Steve sees the effect
> so clearly.
>
> Does anyone have any evidence that would confirm or deny this?
>

Just what i mention above.  But at the moment im not sure enough to
confirm or deny anything.

>
>
> On a broader note, all this kind of suggests (to me anyway) that the
> key to long sparks is filling a large enough volume of space with a
> big enough electric field to drive growth, for long enough to let the
> growth complete.
>
> The way we talk of TC leaders suggests that growth at the tip is fed
> by conduction from the toroid through the discharge channel. I am
> beginning to think what really happens is that the tip feeds on the
> electric field in its immediate neighbourhood. So you can pump all
> the power in you like, but if your coil isn't tall/big-toploaded
> enough to generate 5kV/m (or 1kV/m if over 4ft) in the furthest
> places you want the sparks to go, it ain't gonna happen. You'll get
> lots of action near the coil and nothing further out.

Very interesting hypothesis.  Now the question is, how to test it?
What assumptions can be made if we know, for example, the voltage on
the toroid during breakout conditions and a given spark length?  Maybe
then i can finally get 12'  ;-).

Steve Ward

>
> Steve Conner
>
>
>