[Prev][Next][Index][Thread]

Re: Beating Solved



Tesla List wrote:
(*snip*)
> 
> Brent,
> 
> I have to yet disagree on one point you mentioned.  The dielectric loss
> at Tesla Coil frequencies with kraft paper and foil in oil is full order
> of magnitude BETTER than capacitor grade mylar end foil construction.  I
> would go for the paper jobs over mylar any day.  These paper caps were
> the original pulse power caps.  They are still very good for that purpose
> and beat mylar by a factor of 10!  I have purchased Capacitor grade mylar
> some years ago and have personally constructed mica, glass, paper, mylar,
> teflon, styrene, polyethylene and polypropylene caps! The lowest heating
> (disappation factor according to the data sheets was indeed reflected in
> my experiments.  I chose 200 KHZ as the test frequency and 800 watts as
> the input power. In order of heat (dissapation losses), we found the
> follow results in order most lossey to least lossey.  Most lossey-mylar
> then mica, glass, paper, polyethylene, polypropylene, styrene, Teflon.
> 
> Radio guys love mica, but it is terribly lossey stuff!!!  Why do they use
> it??!!  It will run comfortably and survive well for years at continuous
> use temperatures of 700 degrees farenheit!!  Mica only fails when the
> structure of crystalization is lost around 900 degrees. (turns to dust)
> Transmitter systems need this kind of durability.  Also, Mica has one of
> the highest K factors 5-8 of any dielectric and so a smaller, higher
> value cap can be made with mica.  Also, their value is incredibly stable
> over enormous temp extremes. (important in final tanks)  Mica  has a very
> high standoff voltage.  Everything recommends mica for high voltage hot
> duty continous service, except its terrible lossey nature at RF
> frequencies.  This makes them terrible for Tesla coil work, if losses are
> a concern.  In transmitters they will pay for the losses (power companies
> will sell them all the power that want) in a trade for robust service.
> 
>  Mylar is really bad too its has a disappation factor of .001% at DC but
> rises above 1-5KHZ to about 5% in the low RF spectrum! (similar to mica.)
>  It is not normally ever used above 60HZ in circuits where energy
> transfer is critical. Main use DC filters and low power coupling in AF
> circuits.
> 
> Glass (potash based) is actually pretty good!  Amateurs make theirs out
> of lime glass (window glass) and the caps are no good.  Tesla used potash
> glass bottles only!  (specified by him in a letter to Scherff May 1899.
> Also mentioned in the Colorado Springs Notes.)
> 
> Paper-foil is superb for low end pulse work.  It was used extensively in
> the early atomic pulsed energy systems prior to world war II where the
> values of the caps exceeded micas cost/availability factor.  Paper will
> take a lot of heat, like mica, and under oil will hold off a good amount
> of voltage.  Good choice from DC to 100Khz.  They get real lossey above 1
> MHZ
> 
> The four best plastics are teflon, styrene, polypropylene, and
> polyethylene. in that decending order.  There is so little dissapation in
> all of these that they are virtually identical in dissapation factor!
> (.001% into the GHZ range) The deciding factor is usually one of voltage
> stand off with Teflon and Polypropylene being the best.  Heat is also a
> factor and Teflon is best with polypropylene second.  Teflon has micro
> holes in it like swiss cheese and is just not found in the cap industry.
> So polyprop makes the best pulse caps.  The most stable of these
> materials value wise is modified-cross linked styrene.
> 
> ANY CAPACITOR ON THE PLANET WILL PERFORM IN TESLA SERVICE.  How long, at
> what cost, and with what form of output is quite another matter.
> 
> The color of the spark always tells me a lot about a persons capacitor.
>  purple or purple-violet tells me the fellow is a novice and his caps are
> at the very bottom of their performance curve.  Brilliant pretty blue
> sparks are better but indicate high loss in the tank capacitor is still
> taking place.  Rich blue-white sparks indicate only satisfactory
> operation.  Snow white arc channels only fringed in brilliant blue
> indicate top performance operation and snyergistic tank operation and
> energy transfer.  With enough power input, to over come losses,  Mylar
> and mica can give quite impressive sparks.  Both will heat rapidly and
> only the Micas stand much chance of survival if long runs are the norm.
> 
> properly constructed poly caps, especially "volumetrically inefficient"
> ones just do not heat at all, under any conditions!  They will perform
> efficiently, flawlessly and forever in Tesla coil service when operated
> within their voltage specification range.
> 
> Richad Hull, TCBOR


Thanks Richard...meanwhile, back to the 'lab'...drawing board???

I guess in defense, all I can say is that the caps worked, they were
inexpensive, and I did get the darn thing to work pretty good. 14" from
a 9,000 volt, 30ma neon isn't shabby. Maybe I got lucky - that
'synergistic' thing, you know.

BTW - my large coil system runs with (3) 0.02uF, Poly pulse caps in
parallel. (The Hipotronics Bill Wysock specials) Apparently, I have more
tinkering yet to do - I do get blue-white arcs from it which taper into
lavender fingers at the ends - sometimes even pale blue 'haze' if the
color film is fast enough. Almost there - so far, I can push a 6' spark
out of the 22" tall secondary. One interesting tidbit I did discover -
The coil wasn't running happy, as it rasped and barked a bit until I
added a whole lot more series inductance to the primary side - whoooey!
I think I managed to witness firsthand your 'synergistic' description.
By throttling back the main transformer, (higher Z reflected back to
the primary side of the trans.) the primary tank circuit is free to
do it's thing, and due to resonance rise, had to re-engineer the rotary
gap a bit, as I was getting popping arcs from the rotary wheel to the
grounded frame on the gap! The coil sure came alive too - sings like
a banshee now!

Managed to get a few lovely pic's of the 'banjo' effect too. If you
are interested, I can compress and UU-encode 'em to you or anyone here
who would be interested.

- Brent