[Amps] RF insulating materials - engineered plastics

[John Lyles]

David, you are correct, should be 200 mA for the starting plate current for that column. 

-----Original message-----
From: David Kirkby david.kirkby at onetel.net
Date: Tue, 28 Aug 2012 00:48:41 -0600
To: jtml at vla.com
Subject: Re: [Amps] RF insulating materials - engineered plastics

> An interesting set of results. Is there a zero missing on the row
> "PTFE+Glass        84     98        20      205  goodmechanical"
> since it simplies the current started at 20 mA and rose to 205 mA,
> when I expect you mean 200 and 205 mA. Or have I interpreeted things
> differently.
> 
>

I have used a 3D printed plastic part as an air shroud around the screen contact ring on the TH628, operating at 250 kW average power at 200 MHz. It is not directly across the insulator between screen and anode, or it would have never survived. It is shielded beside the screen ring itself, out of electric fields. However, even so, the part has changed dimensions slightly and had to have foam tape applied to block air leaks. This is likely due to the thermal instability of this material from the radiant heat. I haven't found any yet that is a good RF insulator, be warned!

> I was particularly interested in what you had to say about the
> material for 3D printing. I was looking to get a radome made by 3D
> printing for a helical antenna at 2.45 GHz. It looks like that might
> not be such a good idea, though perhaps there are different materials
> used. I believe people have done 3D printing with chocolate!
> 
> So that's that an electric field of 20 kV/inch or 787 V/mm, which is a
> little under the breakdown of air  I believe, which is generally said
> to be 1 kV/mm.

Yes, any higher and it sparks. 

> It would make an interesting comparison if the loss tangent could be
> measured at low RF voltages, such as using a vector network analyzer
> to do a measurement. I wonder if the loss tangent is different at high
> electric fields to low electric fields?

I don't think the loss tangent is nonlinear like that for these samples, but I would need to read up on Von Hipple's text book from MIT again to be certain. Loss tangent is a strong function of temperature on some of the materials as evidenced by the runaway plate current as it heated. This is correlated with the temperature of glass transition in the material. Rexolite, an otherwise excellent radome insulation, has a horrible tendency of running away if it is not cooled with air. We use it for insulators in our 14 inch diameter coaxial transmission lines carrying the 3 MW of peak power. 
 
> I believe one way that would probably work on material this size is to
> put them against an open-circuit transmission line connected to a VNA.
> I've recently been given some thought to if that is practical just by
> pushing the flat material against a sexless APC-7 connector. Agilent
> sell a probe for this sort of thing, and charge a fortune for the
> software to work out the results. I assume that means the mathematics
> of it are non-trivial, otherwise the software would not be so
> expensive.

I have one of the HP sets you mention, but not the software for it. Found it in someones storage cabinet. 

For lower loss materials such as these, the old standard way would be to use the HP Q meter, the wedge shaped one. It had capacitance parallel plate fixtures. But it wouldn't go up about about 100 MHz. 

73
John
K5PRO