[time-nuts] position determination over short distance

Lux, James P james.p.lux at jpl.nasa.gov
Sat Dec 6 15:36:52 UTC 2008




On 12/5/08 9:47 PM, "WarrenS" <warrensjmail-one at yahoo.com> wrote:

> James
>
>
> Question
> So if the antenna is rotated in any and all of its axes, it does not should
> like from what you are saying there is really just a single point in it like
> you get for say 'center of mass'. that stays at a known spot.
> It sounds  more like a bunch of calibration points than a single point,
> similar to a calculated lookup table, depending on angle etc.
> Does the question make any since?


Yes.  What you're talking about is the "phase center" of the antenna.  Most
antennas have a fairly well defined phase center over some restricted set of
angles.

A perfect point source antenna would have the phase center at the point.
But, in real life antennas have physical extent, so the possibility arises
of the apparent phase center moving depending on what direction the incident
wavefront is coming from.

Consider a point radiator some distance above a reflective plane, and look
at the received phase at the point, as the angle of incidence of the
wavefront changes.  The point receives both the direct signal and the
reflected signal which combine to form something with different phase and
amplitude.


If you're building a feed for a parabolic dish, it's pretty easy to arrange
the phase center to be in a fixed location (i.e. The focus of the parabola),
but the field of view of the feed is only a small fraction of a hemisphere.


For GPS antennas, where you really want coverage over a hemisphere, its a
bit more challenging.  Mostly, it's a matter of suppressing the reflections
and creeping waves along the surface next to the antenna. (i.e. A patch or
quad helix in free space is pretty good.. But darn it, you've got to mount
it on something, and usually, you want the antenna on something else anyway,
like some sort of building)


There are published (online) data for lots of GPS antennas that give phase
center tolerances and properties. (it's something of interest to precision
GPS folks like surveyors and geodysists)

> In any case do you think it would help in the original question, which was to
> measure NON GPS RF signal distances to mm distances??

Yes. Most antennas can be calibrated.  In the specific case here, you'd
probably do some sort of insitu calibration.  Move the target to known
locations and build up a cal table.  We did something similar a few years
back with a form of interferometer where we were interested in knowing the
angle of an incoming wavefront to fractions of a milliradian.  We put the
system in an anechoic chamber (it was big, so doing it on the benchtop
wasn't feasible) and then had a source 15 meters away that we moved on a
precision mechanical positioner.


Jim




More information about the time-nuts mailing list