[time-nuts] DIY GPS antenna...

Attila Kinali attila at kinali.ch
Sun Sep 29 13:38:19 EDT 2013


On Sun, 29 Sep 2013 19:00:54 +0200
Attila Kinali <attila at kinali.ch> wrote:

> > That sounds like a pinwheel antenna.  I've wondered how hard it would 
> > be to build one, but I've never seen a DIY version and the math is way 
> > beyond me.  The only commercial version that I've seen is from 
> > Novatel.  Do they control all the patents?
> 
> Building one should be fairly easy. Just use standard PCB fabrication
> and you can build the whole structure with high accuracy. The design
> itself is the hard thing to come up with. I think, for us mere mortals
> the only way to come up with something half decent is to use a genetic
> algorithm that goes trough various designs and selects the best one.
> Unfortunately, i'm not aware of any antenna design software suitable
> for such a task.

BTW: if you want to build your own GPS antenna, the easiest designs
are

* Patch antenna 
	If you don't use any ceramic material, it's just two rectangular
	copper plates ontop of eachother with some spacers, although
	bit big. There are IIRC closed formulas to calculate the size
	of the patch and the position of the feed. One example of this
	kind is [1]

* Cross-Dipole antenna
	Like the patch antenna, this one follows the idea of having
	two arms at a right angle, which differ slightly in length.
	Very easy to calculate, but it becomes large and you need more
	free space around it than the patch antenna and probably a ground
	plate as reflector as well. If you dont feel like building this
	one out of wires for the fear of getting the measures wrong, make
	a PCB design with a slit in the middle to put two PCBs at a right
	angle together. The traces on the PCB then form the antenna which
	you just have to solder together.

* Helical antenna
	I haven't had a look at those, but i guess they are not too
	difficult to calculate (there are at least a few ham radio
	designs out there and i'm sure Rothammel has the formulas).
	Some form of platic tube could form as a base material where
	you wind wire around it.
	The clasic 4 wire helical design with a rigid coax core is well
	known and there are quite a few desscriptions out there, e.g. [2]
	And if you are at JPL, you are free to but this antenna into
	a salad bowl ;-)


I would probably prefer the patch or cross dipol antenna made out of
PCB material, such that the bottom of the "ground plate" PCB would contain
the LNA.

			Attila Kinali



[1] http://www.express-builder.com/docs/gpsant/
[2] http://lea.hamradio.si/~s53mv/navsats/analog.html

-- 
1.) Write everything down.
2.) Reduce to the essential.
3.) Stop and question.
		-- The Habits of Highly Boring People, Chris Sauve


More information about the time-nuts mailing list