PVC Water Absorption: fallacy?

From:  Malcolm Watts [SMTP:MALCOLM-at-directorate.wnp.ac.nz]
Sent:  Monday, June 01, 1998 5:58 PM
To:  Tesla List
Subject:  Re: PVC Water Absorption: fallacy?

Hi Richard, Bert, all,

> From:  Bert Hickman [SMTP:bert.hickman-at-aquila-dot-com]
> Sent:  Saturday, May 30, 1998 7:12 AM
> To:  Tesla List
> Subject:  Re: PVC Water Absorption: fallacy?
> Tesla List wrote:
> > 
> > ----------
> > From:  R M Craven [SMTP:craven-at-globalnet.co.uk]
> > Sent:  Thursday, May 28, 1998 2:49 PM
> > To:  Tesla List
> > Subject:  PVC Water Absorption: fallacy?
> > 
> > Many people recommend that PVC coilforms should be varnished inside and out,
> > in order to seal them against water ingress.
> > 
> > Having spent an hour or so in our library with various materials handbooks,
> > I see no mention  of any significant water absorption of any of the rigid
> > PVCs which are encountered. It is a very good plastic, comparable with HDPE
> > (rigid polythene, alkathene). The volume resistivity is not affected by
> > immersion in water, and the surface resistivity is only marginally worsened
> > (same is true for most amterials: even touching the test sample will cause
> > an OM change in ohms per square). Nylon and PTFE do suffer, but pretty well
> > all of the thermoplastics are not prone to absorption. They are impermeable.
> > 
> > So, why is it recommended that PVC in particular is treated inside and out
> > with varnish? Is it actually the case that, in doing so, we create a more
> > tacky surface on which to wind our secondary wire? Is it so that the wire
> > will slightly embed itself in the varnish, get a grip, and thus exclude
> > air-pockets which might harbour water vapour?
> > 
> > I think the reason for varnishing might well be moisture related, but it is
> > to exclude air pockets, nothing to do with water absorption. If so, then
> > surface prep. should be carried out on all coilforms (which I acknowledge is
> > recommended by the experienced builders on the list)
> > 
> > If someone can quote me a BS or ASTM or other document that states
> > hygroscopicity of PVC, then I guess i'm wrong!
> > 
> > Any comments from people who've built an untreated PVC secondary and been
> > able to do a comparison with a varnished one?
> > 
> > Richard Craven, Malvern, England
> Richard and all, 
> I agree that technical information on this effect is sparse, and bulk
> PVC certainly is an excellent insulator. However, there does seem to be
> some empirical evidence for poorer performance at high humidity, at
> least by some Tesla coilers. For example, I've got a tandem RQ/TCBOR
> Static gap with a total of 12 gaps housed in single 6" PVC pipe. The PVC
> was not specially pre-treated in any way. On humid days, there appears
> to be a mechanisn whereby the surface of the PVC becomes significantly
> more conductive, leading to sporadic flashovers exceeding 2" long along
> the surface of the PVC. This doesn't happen on days with lower humidity. 
> I also notice that the coil's output does not appear to be as great, and
> there's more difficulty with breakout when the coil is run on humid
> days. Best performance is in the winter when the relative humidity is
> quite low. The PVC coilform was sanded and pre-sealed on the outside,
> but not on the inside. 

The highest Q secondary I have ever wound was done on thickwall 
unsealed PVC drainpipe - done totally against the common "wisdoms" 
of which I knew nothing at the time. On measuring this coil with
great isolation I found that at 5% humidity, I hit an unloaded Q
of 320 odd and in a 50% humid atmosphere this dropped to 300. 
Performance of this in a system is noticeably worse on a humid day 
but I attribute that to the atmosphere and its ionization 
characteristics, not the coil itself. A difference in Q of a few 
percent cannot account for a 20% decline in maximum discharge length. 
Moreover, discharge length does noticeably improve after running for 
twenty seconds or so in the damper atmosphere but does not attain the
performance heights of working in a dry atmosphere.