Air cooled tube chimneys seem to be a big topic these days, as much
as discussion of EPAY and that other service. I want to suggest that
people consider that these chimneys are typically situated at the
highest impedance of an RF amplifier, across the anode to ground
space. With 2-3 kV of peak RF swing there, at, say 28.25 MHz, one can
expect a minimal but measureable dielectric heating effect. With a
thin cross section, like transformer paper that Will suggests, this
may not cause appreciable heating in such a narrow piece of
dielectric. As long as moisture doesn't ingress into the paper, it
should be OK. Vinyl chloride-containing polymers (like PVC) are just
the opposite of what you want for a low loss dielectric. According to
Arthur Von Hipple's classic compendium of dielectric data from MIT in
the 1950s, (Dielectric Materials and Applications, available in
reprint from Artech house now), PVC formulations have typically 100 x
the loss tangent of polyethylene at 1 - 10 MHz range. RF sealing
process is used to stamp PVC articles together, like the patches on
auto carpets under the pedals, like inflatable rafts, like door
panels for autos, packs and raincoats. The harder polymers such as
used in pipe may be a bit more resistant than the flexible variety,
but they still have a lot of ionic content. We don't use PVC pipe
fittings in commercial Deionized water systems as they tend to
release ions into the solution.
As an RF engineer I try to stick by rules of things which are
acceptable and not, and apply them universally. I spent 6 years doing
just dielectric heating experiments for applications at E. I. DuPont
Co, so I got very familiar with what plastics I could trust or not.
My overly conservative stance (go ahead AG6K, fire away!) often gets
criticism from those who say it will work fine at lower voltage, or
for low duty operation like amateur transmissions. It true that cost
is no object in many high powered systems where you don't want to run
the risk of a burn-out. In low Q applications like pi-networks, the
voltage is well defined. In cavity and line circuits, beware, as the
Q can be higher, the voltage rises significant. I still suggest that
we don't tempt Mr. Murphy too far, as one may be sitting up on 10
meters (or worst yet, 6) key down and wonder why output power seems
to be dropping while smoke begins to rise from the PA compartment
(and that hideous smell of melted PVC). Think of it this way, would
you use the same material as a standoff insulator in an RF network
with high voltage? Do you mind the lowering of Q that lossy
dielectrics can introduce? Unfortunately, the old microwave oven test
doesn't work well enough to apply the results, as constant fields are
not produced, they purposely move them around to cook a chicken
uniformly. Applying a fixed electric field in one direction through a
thick dielectric, with minimal spaces (air gaps) between it and the
metal, is the exact method used to get temperature rise in dielectric
heating industry. Be careful; check your homeowner insurance policy
if you care to experiment and take materials close to their limits.
John
K5PRO
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|