[Top] [All Lists]

[Amps] RF Insulation

Subject: [Amps] RF Insulation
Date: Fri, 26 Mar 2004 10:43:28 EST
List-post: <>
The problem with all of this discussion is the casual use of the term 'RF 
fields". In fact, the electric field inside of a solenoidal inductor (a/k/a 
choke) is quite small, almost negligible. The large RF solenoidal field is 
predominantly magnetic. The dissipation factor of plastics relates to their 
absorption of the electric field (E) and is not relevant here, and there are no 
materials present which present losses to magnetic field (H). The latter, of 
course, is exactly the problem you run into when you try to use a ferrite core. 

I have built all of my plate chokes out of PVC for the past 20 years, and PVC 
is a fairly "lossy" dielectric. I like it because I live walking distance 
from a good hardware store! But the heat generated by the chokes is always 
dominated by the loss of the copper itself, not the plastic form.

In other components, there can develop large RF electric fields if the 
plastic is used to any extent as an insulator between high-potential RF and 
If the spacing decreases significantly below an inch, then the field 
(gradient) grows and dielectric losses can and  will occur. That is when teflon 
or poly 
becomes necessary. Nylon and all of its variations (including Delrin) are 
moderately lossy.

These effects can all be readily calculated using simple capacitance 
formulas. I have done it many times and never melted a piece of plastic or 
my favorites for RF construction. 

For those interested in the relative losses of different kinds of materials, 
there is a good summary of the RF loss properties of the various common 
plastics materials in an article I wrote several years ago that focused on the 
construction of capacitors in gamma matches, where it is a highly significant 
issue. Just send me a request for a copy of that article in MS Word format.

Eric von Valtier K8LV
Amps mailing list

<Prev in Thread] Current Thread [Next in Thread>
  • [Amps] RF Insulation, RFlabnotes <=