[Amps] RF insulating materials - engineered plastics

[David Kirkby]

On 28 August 2012 20:26, John Lyles <jtml at losalamos.com> wrote:
>> 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.

OK. I was wondering if there was a simple way of measuring the
properties of a dielectric without access to high power equipment,
before incorporating it into high power equipment.

>> 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.

I've been looking for one on eBay. Not seen one come up, but I've only
been looking for a week or so, since I bought a 20 GHz VNA a couple of
weeks back. But as far as I can see, there's nothing very clever about
it, though they one they sell can be used for liquids too. I don't
think it's anything one could not constuct in a garage to be honest.

The big advantage for me at least in using an APC-7 connector would be
that one could easily attach a short circuit, open circuit then a load
to calibrate a VNA. You also have a ready made transmission line, with
a nice flat end.

There appear to be a number of methods of measuring the dielectric
properties of materials. I can't recall them all off hand, but some of
them are:

1) Push material against an open transmission line.
2) Make material part of a transmission line (coax or wave guide)
3) Make a resontator, put the material in, and check the change in
resonate frequency.
4) Sandwich between two plates and make a capacitor and measure with a Q meter.
5) Put material between a transmit and receive antenna.
6) Use a very high power source at 90 MHz, cutting the materal 3/4" thick.

http://cp.literature.agilent.com/litweb/pdf/5989-2589EN.pdf
covers 1-5 and more, but not #6.

> 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.

That's what I like in principle about a sample on the end of a
transmission line. You can cover from kHz to GHz. But like all
methods, it has regions where it works well, and others where it does
not.

Dave