[AMPS] dielectric losses

[John T. M. Lyles]

I've been meaning to get back to the dielectric materials topic with 
some suggested alternate materials:

At 11:50 PM -0500 1/15/00, owner-amps-digest@contesting.com (RF 
Amplifier Discussion Dige wrote:
>I have been advised NOT to use Teflon, Delryn, and a few other
>insulating materials for an RF choke. Instead only use ceramic. It was
>then stated the "D" factor was the reason.
>
>OK, but why as I could not find info on this subject in the ARRL
>Handbook. Any help to point me in the direction of knowledge on this
>subject would be appreciated.  May as well learn something while
>building my amp. And I want to do it RIGHT.

The following reply is correct, thanks Dave for inserting facts into 
that discussion.

>Teflon is a (fluorocarbon resin) or (flourine plastic) and has the same
>structure as polyethylene, in which the hydrogen is replaced by (flourine).
>OR  a part may be replaced by (chlorine), in which case it is called a
>(fluorohalocarbon resin).
>
>Polymerized (tetrafluoro ethylene), aka PTFE, which carries the E.I. du Pont
>trade name of  Teflon,  formula (CF2*CF2)n  is probably the most common
>solid form used for insulating materials, bearings, gaskets, insulating
>tapes, wire insulation, etc.  This is a white waxy solid with specific
>gravity of up to 2.3, typical tensil strengths of up to 3500 psi, elongation
>of 250-350%, dielectric strength of 1000 v/mil and melting point of 594*F.
>
>Another commonly encountered material in electronic use and sometimes
>referred to generally as "Teflon", is Teflon FEP, which is a fluorinated
>(ethylene-propylene).  In thin films (typically down to 0.0005"), it is most
>frequently found as insulation in capacitors.  Typical tensil strength is
>3000 psi, elongation 250%,  and dielectric strength 3200 V/mil.
>
>Neither PTFE or Teflon FEP emit phosgene upon decompositon at elevated
>temps.
>
>However, fluorothene plastic (which may be what Rich's info was referenced
>to) has the formula (CF2*CFCL)n  and differ from (Teflon) in having one
>chlorine atom on every unit of the polymer chain, replacing  the 4th
>fluorine atom.  This is a (transparent) material with a specific gravity of
>~2.1, has tensil strengths of up to 9400 psi, but has a lower withstanding
>temp (as related to Teflon) of around 300*F.  This material is sometimes
>called "Teflon" due to its similarity in chemical and polymer-structure
>make-up, and will emit phosgene upon decomposition at elevated temps.
>
>73, Dave, K1FK


I worked for E. I. DuPont in RF reseach in the late 1980s. We made a 
LOT of dielectric measurements up to 3 GHz. We used Teflon* as a base 
material (a substrate or boat) in a lot of heating experiments in 
high power RF. It is one of the best materials to use. Even when 
there is a direct arc, unless it burns deep into the material, it can 
be reused without repeat arcing. Teflon* is also rather inert, so it 
does not readily combine with other chemicals in experiments. Surface 
arc-resistance of pure Teflon* PTFE (* DuPont trademark) is 
excellent, and carbonized arc paths do not easily form on the 
surface. As others have already mentioned, the dielectric losses are 
among the best for polymeric materials (or even natural materials 
besides gases like air and vacuum). It is often used in radomes and 
in microwave windows. But it cold flows very easily, so if clamped or 
mechanically held, it will eventually deform. So it is rarely used 
where mechancal strength is needed. Coil forms should be no problem, 
for smaller coils wound with magnet wire, however.

There is a sickness called the Teflon* flu. Machinists can get it 
when working the material at higher temperatures near the smoking 
point. This is to be avoided. It was not phosgene gas, but a fluorine 
byproduct I think. I don't know how harmful it is, but sounds like 
Rich's friend succumbed to an extreme case, or something else.

The following Teflon* safety information came from E. I. du Pont de 
Nemours and Company website:
"Before using TeflonŽ, read the Material Safety Data Sheet and the 
detailed information in the "Guide to the Safe Handling of 
Fluoropolymer Resins, Latest Edition," published by the 
Fluoropolymers Division of The Society of the Plastics 
Industry-available from DuPont.
                       Open and use containers only in well-ventilated 
areas using local exhaust ventilation (LEV). Vapors and fumes 
liberated during hot processing, or from smoking tobacco or 
cigarettes contaminated with TeflonŽ or TefzelŽ fluoropolymer resins, 
may cause flu-like symptoms (chills, fever, sore throat) that may not 
occur until several hours after exposure and that typically pass 
within about 36 to 48 hours. Vapors and fumes liberated during hot 
processing should be exhausted completely from the work area; 
contamination of tobacco with polymers should be avoided. Mixtures 
with some finely divided metals, such as magnesium or aluminum, can 
be flammable or explosive under some conditions. "


A lot of flexible conveyor belts for RF heating systems in industry, 
use glass-reinforced Teflon*  FEP polymers with silicone rubber. They 
don't get very hot in the intense RF fields used to heat polymers.

To circumvent the weak mechanical properties, microwave industry uses 
crosslinked polystyrene, such as Rexolite, Polymer Q200.5, and maybe 
a few other tradenames. This stuff is the hard brittle plastic that 
is translucent, although it can be polished to be like glass. We use 
it to carry 3 Megawatts of RF, as the spider in our 14 inch coaxial 
lines in my present work. We buy it in 4 foot sheets, and cut and 
turn the discs. We have a special polishing wheel that looks like a 
phonograph. Rexolite does leave a carbon track when arced, and it can 
really melt into a goo if ignited. But the dielectric properties are 
only slightly inferior to Teflon*. K is 2.55 instead of 2.05-2.1. 
Loss tangent is still 0.000x range. Its good mechanically. By the 
way, 3 Megawatts is about 12 kV RMS, assuming perfect match in the 
coax line. It is worse than this, because of the high VSWR during off 
tune (Q of our load is 60,000).

Another material used a lot is ultra high molecular weight 
polyethylene or polypropylene -also called UHMWPE. This stuff is soft 
like Teflon*, a lot cheaper, and has good low loss, if you are 
careful in what you spec and buy. However, it is very easy to get the 
wrong material, so a dielectric test is really needed before 
designing with a piece. This stuff is used for waveguide windows 
sometimes.

Amateur applications of lower RF voltage and frequency can get by 
with materials such as PVC, G10, maybe even Delrin (very questionable 
to me though). In high voltage RF service, in the VHF region, none of 
these can be used, due to excessive losses. Believe me, I have tried 
them all. I have a standard test i perform, in a 90 MHz dielectric 
heater, putting about 15 kV peak RF across a small block. In a few 
minutes, most of those materials will either burst in flames, or will 
begin to smoke and stink, as they are internally cooked.

If you want to use fiberglass materials, G7 has lower loss than G10 
or G11, due to the use of silicone instead of the more lossy epoxy 
resins in the other. It is harder to machine and cannot be turned for 
coil forms, we tried it 2 years ago and it just delaminated and fell 
apart. For the RF choke in the 100 kW 2.8 MHz amplifier, where the 
coil is exposed to the E field alternating at the plate of the tube, 
I chose polysulfone rod. It is easier to machine than Rexolite 
(crosslinked Polystyrene), and has fairly good dielectric properties. 
It is sold under one tradename UDEL* or Thermalux*. We are now using 
this same material for HV insulator posts for a 90 kV klystron power 
supply (not AC dielectric problem, only DC standoff).

Another interesting new material is polyetherimide. This is 
trademarked as ULTEM* by GE and Westlake Polymers and Hydex* by 
Polypenco/DSM Polymers, I think. It can be had in glass reinforced 
sheets, and has good RF properties. It is expensive but very strong 
for mechanic support, such has holding an 80 pound tetrode so that it 
does not crash into the finger stock at the bottom of the filament 
connector in the socket.

There are numerous other materials that are good, mostly more 
expensive, like polyimides Torlon* and Vespel* with incredible 
strength due to glass loading, etc. These are really beyond the range 
of most of us. They should be considered in cases where the ionizing 
radiation (gamma, Xrays) is high, and temperatures are high.

The BEST reference book on dielectrics is still one of A. Von 
Hipple's books "Dielectric Properties and Materials", MIT. Long out 
of print, but I heard that they got reprinted not too long ago. Of 
course, these texts don't have all the modern miracle polymers that I 
refer to above.....

If it works for you, then use it. But be aware that it may smoke now, 
or later, when moisture is absorbed in your material. Pick something 
that is not very hydroscopic if you live in a more humid climate than 
I do in New Mexico.

73
John
K5PRO



















--
FAQ on WWW:               http://www.contesting.com/ampsfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm