[time-nuts] Mounting GPS Antenna on Steel Roof

Bruce Griffiths bruce.griffiths at xtra.co.nz
Tue Mar 4 03:23:06 EST 2008


Matthew Smith wrote:
> Hi Folks
>
> Further to the cable splicing thread, I have been having a discussions
> off-list with Didier regarding the placement of my antennas on the ridge
> cap of my roof.
>
> There is certain concern that I may have issues with reflections - can
> anyone confirm whether or not this is the case?
>
> Here is a terrible diagram (I cannot for the life of me figure out how
> to draw solid, straight lines with the Gimp) of the proposed installation:
>
> <http://www.smiffysplace.com/files/antenna_diagram.jpg>
>
> The ridge runs just about 15 degrees of an East-West line, is the
> highest ridge on the building and is the highest point for some distance
> (there are trees 80m to the South), barring a vertical phased array VHF
> TV antenna.
>
> I can't really see how a signal striking the roof would reach the
> antenna, working on the basis of angle of refection = angle of incidence.
>
> Is this installation likely to case me any problems?  I am rather short
> on other options due to the presence of steel sheds and fences.  Getting
> the antenna clear of all of these would probably require something like
> a 100m+ cable, either buried or strung overhead, neither of which is
> practical (nor the cable length).
>
> The cable run from the ridge to my office works out at about 10m.  I
> will have a redundant antenna a couple of metres along the ridge from
> the main, disconnected.  If antenna #1 were to fail, I would just unplug
> the antenna extension cable in the loft from #1 and connect to antenna #2.
>
> I have spotted a splitter on eBay that is good for 10-1880MHz and
> includes power passing on one output - I guess that this means I am
> still able to power the active antenna as normal provided that at a
> receiver is connected to that output.
>
> Any comments on this would be much appreciated.
>
> Cheers
>
> M
>
>   
Matthew

One has to be careful in assuming that reflections are a simple as you 
depict especially when the reflectors are only a small number of 
wavelengths (~ 19cm @ 1575.42MHz) in extent.

The presence of any periodic structure in the metal roof (eg 
corrugations) may cause significant diffraction effects particularly in 
the near field.
Roughness of the finish will increase scattering over that obtained with 
a smooth surface.
Also at least some of the magnetic base antennae are intended to be 
mounted on a conducting surface.

So the preferred mount may be on a ground plane of the recommended size 
on the end of a tripod a meter or more in height.
However since you have a magnetic mount its easy to try your method and 
see if it causes any observable problems, if it does try a different 
mounting scheme.

Even the experts can get it badly wrong, mounting GPS antennas with 
choke ring ground planes mounted on concrete pillars can contribute to 
measurement instability whereas such mounts tend to be very stable 
mechanically, see:

ftp://ftp.ngs.noaa.gov/dist/Jimr/igsframe-igs06.pdf

ftp://ftp.ngs.noaa.gov/dist/Jimr/metrologia-2.pdf

http://www.geopp.de/media/docs/pdf/gppigs06_pnf_g.pdf

http://www.geopp.de/media/docs/pdf/gppigs06_pnf_g.pdf

The size of the effects detailed above are only of significance for 
geodesy and other high precision applications.
However, if a choke ring ground plane or equivalent were not used the 
effects would be much larger.

Bruce



More information about the time-nuts mailing list