TSSP: List Archives

From: Bert Hickman
Date: Wed, 01 May 2002 00:06:59 -0500
Subject: Re: [TSSP] Secondary voltage stress factor

Paul,

Responses interspersed...

Paul wrote:
> 
> Bert wrote:
> 
> > creeping leader surface discharges "hug" the surface of a
> > dielectric, and are enhanced when there's a conductive plane on the
> > opposite side at a different potential. As the leader propagates
> > along the dielectric surface, streamers fan out and collect higher
> > displacement current than if the leader was propagating in free air
> 
> One of my favourite HV games is playing a 'hot' electrode across the
> surface of a pond or pool, which seems to give impressive displays -
> longer 'streamers' with much more branching, spreading out across
> the water surface. Is this a related thing?

This is indeed a similar phenomenon. As you describe, you can make very
interesting and beautiful Lichtenberg-like surface discharges in this
manner. During the brief intervals that the leaders propagate, the water
behaves mostly as an insulating surface. You can readily demonstrate the
difference in length between the discharges coming from a positive HV
wire (much longer) versus those from a negative HV wire. Dr. Colin
Pounder's site (referenced in an earlier posting on the Tesla List) has
a couple of nice photos of these positive Lichtenberg figures over
water:
http://www.geocities.com/CapeCanaveral/Lab/8063/lichtenb.htm

My favorite shot of dielectric breakdown over water (from water
insulated transmission lines breaking down and "bleeding" off the tail
end of a huge pulsed power discharge) is from Sandia's awesome
Z-Machine:
http://www.sandia.gov/media/images/jpg/Z02.jpg

> 
> I suppose then the enhancement by a conductive surface on the
> opposite side occurs because of the resulting 'shunt' capacitance
> that the streamer current has to charge up, ie more current,
> therefore hotter.

Exactly. This phenomenon is also used in specialized low inductance,
ultra high current HV spark gap switches to stimulate multiple breakdown
paths between parallel electrodes along a dielectric surface.

> 
> Is this surface leader propagation enhanced because the
> mobile charges in the leader have less opportunity to escape,
> since half of their lateral escape route is blocked by the
> dielectric surface across which charges cannot move.  Is it this
> lateral confinement that makes the difference?

Possibly, but more likely, the positively charged leader really doesn't
"want" to escape, since it's electrostatically attracted to the opposite
plate of the capacitor that's formed during the discharge.
 
> 
> I'd like to know the peak base current of some coils that are
> exhibiting racing arcs, so that we can work out what the V/cm is
> along the coil.  I'm keen to get some examples which show how much
> lower than 26kV/cm the problem starts at.  They would be very
> useful numbers for coilers.
> --
> Paul Nicholson,
> --

Sounds like a pretty good way to get an estimate of Vmax. Once a surface
discharge leader is initiated (the literature indicates that this can
occur at about 10 kV for a creeping leader) the leader can then continue
to propagate with average fields of 5 kV/cm or less. This could be
further reduced if there were regions of stranded negative surface
charge on the outer dielectric coating of the winding. I would expect
that the initiating conditions for surface discharges could exist at
E-fields of 10 kV/cm, perhaps less...

Best regards,

-- Bert --
-- 
Bert Hickman
Stoneridge Engineering
Coins Shrunk Electromagnetically!
http://www.teslamania.com


Maintainer Paul Nicholson, paul@abelian.demon.co.uk.