[time-nuts] Why no cheap timing grade GPS antennas?

Attila Kinali attila at kinali.ch
Sat May 9 00:24:13 UTC 2015


On Thu, 7 May 2015 12:18:39 -0400
<lstoskopf at cox.net> wrote:

> Note in the last digest indicated 

If you are using digest mode, please disable it in the user settings
at https://www.febo.com/cgi-bin/mailman/options/time-nuts
It only makes sense if you never intend to reply to any mail and
messes up things when you do.

>that some of the patents run out in 4 years.

Only the first of the patents that is the basis of the Novatel Pinwheel
antenna. It defines the antenna as one with multiple radial (straight
or curved) slots that are coupled to a circular (but not closed) feed
on the backside [1]. The key point here is the combination of both
radial slots and a circular leaky waveguide feed in the back that
gives the whole thing its circular polarization.

I can think of a couple of ways of getting around this restriction,
either by using a leaky waveguid on the front side, or not using
slots, but radial "wires"[2], etc. I think that the Trimble Zephyr
antenna does something like that, as what i've read about it seems
very similar to the pinwheel, but with a different feed network.

> Looks to me that some plastic stamped choke ring with suitable
> metallic coating should be cheap in bulk.  Finding the phase
> center might be interesting, but.......

If it would be, someone would have done it ;-)

There are two problems with that. You need good conductivity for the
fins, and an even better conductivity for the back side (bottom of
the choke ring). Currently the material of choice is aluminium,
because it's (kind of) cheap, has a good conductivity,
and you can machine it very easily.

Coating molded plastic is not as easy as it sounds. Especially
not if you want 1) good conductivity, 2) have it perform
well at high frequencies, 3) being of reproducible quality
and 4) stable that it doesn't quickly degrade when used outside. 
My experience with molded interconnected devices (MID), which are
again a bit more difficult, is rather bad. Yes you can do high
frequency stable stuff, but you need a manufacturer who knows what
he is doing, at each step of the proces, from choosing the right 
plastic base material, over how to form the mold and how to inject
the plastic, to getting the metal deposit passivated.

Building some electronic bandgab (EBG)[3] around the antenna is an easier
design. Especially if you are building your antenna on a PCB material
anyways. Then you can pattern the surrounding area in such a way that
you get the same high impedance surface effect as you would get with
a choke ring, that surpresses the low angle incident signals and
the back lobes.

You can find in [4] and [5] two examples on how to do that,
applied to GPS antennas. There is no specific reason why I
choose these two beside that I just had them open. There are
many more designs as the idea has been around for quite some
time. It also has been incorporated in the pinwheel antenna
as can be seen in [6]

I left out the whole economic side of this question, mostly
because I know too little about it, to give you more than
a rough guess. But you can assume that for any mold to be
cheap you have to use it for at least 10k pieces. Before
that the mold manufacturing will dominate your cost of
the molded plastic.

The PCB based design cost is limited by the size of the plate you
are using. from around 15-20cm upward, the cost of the PCB goes linearly
up with the area (or quadratic with the longest length). You also
have to have someone who can either produce FR4 PCB's with very high
reproducability, or you need to choose one of the high frequency substrates.
Ie even this isnt that cheap as you would wish. But i would guess it
has the higher potential to get cheap at high (>1k) volumes

I hope that answers your question

			Attila Kinali

[1] "Aperture Coupled Slot Array Antenna", US Patent 6,445,351
by Kunysz, 1999

[2] On an antenna, whether you place a wire far away from any metal
or a slot into a grounded plane has (almost) the same effect on the
resonance and radiation characteristic.

[3] The choke rings are actually an EBG design. A very simple one, but
one nonetheless.

[4] "A Novel GPS Patch Antenna on a Fractal Hi-Impedance Surface Substrate",
by Bao, Ruvino, Amman, John, 2006

[5] "Tri-Band GPS Antenna With Low-Multipath Ground Plane", 
by Wang, Deng, Gu, 2010

[6] "Leaky Wave Antenna With Radiating Structure Inclduing Fractal Loops",
US Patent 7,850,916, by Kunysz, 2005

-- 
< _av500_> phd is easy
< _av500_> getting dsl is hard



More information about the Time-nuts_lists.febo.com mailing list