[time-nuts] pin-wheel antenna

[David Kirkby]

On 20 April 2013 20:52, Tom Van Baak <tvb at leapsecond.com> wrote:

> For the rest of you:
>
> http://www.leapsecond.com/images/gps-pinwheel-1.jpg
> http://www.leapsecond.com/images/gps-pinwheel-2.jpg
>
> It's a thing of mysterious beauty. And the GPS World photo saves me from the temptation to break open my own pinwheel antenna just to see what's hidden inside.
>
> /tvb

Does the antenna work better than other types? As someone who used to
design antennas for a living, I'm well aware there are a lot of
antenna "designs" which are either badly understood by their
"designers" or are just put together to look impressive. One company I
used to work for based their specifications on that of their
competitors. This seems to be pretty common practise in the antennae
industry.

I once offered my boss at a company I worked for a bet. I would pay to
get one of our antennas tested at NPL, and if it met the specification
we claimed, then he owed me nothing. But if it failed, he had to pay
the cost of the testing. I could possibly not win any money with this
bet  - the best I could hope for was to break even, but I was
sufficiently convinced it did not meet the specification that I could
take that chance. Needless to say he would not accept the bet!!!

There are lies, damn lies, and antenna specifications.

This particular antenna design

http://www.rason.org/Projects/collant/collant.htm

has a claimed gain of 9 dB. We don't know if that is 9 dBi, dBd, or
dB_wet_string. Both myself and someone else have modelled this based
on the use of perfect conductors and we get a gain of about 7 dBi. I
used the dimensions given in that article, him a rescaled version for
2.4 GHz.

We used different software - me the exceedingly expensive HFSS 3D
electromagnetic simulator and him another very expensive EM simulator.
These both solve Maxwell's equations, although one uses  the integral
form and the other the differential form. I've contacted the author
and got no response.

One obviously way to convince yourself that antenna can't work as
described, is what would happen if the coaxial cable had a very high
permittivity lossless dielectric - say Er=10^6. All dimensions would
scale down by 10^3 from free space, and you end up with a very high
gain antenna 1.4 mm long.

Hence I tend to take antenna specifications with a pinch of salt.

Dave.