[time-nuts] GPS backup for the stationary time and frequencyuser

jimlux jimlux at earthlink.net
Fri Oct 8 16:32:07 UTC 2010

Bill Janssen wrote:
> Magnus Danielson wrote:
>> On 10/08/2010 03:35 AM, jmfranke wrote:
>>> When I said the feed would work, I was meaning it would work if LHC.
>>> The illustrations and text imply you could just place a normal GPS
>>> receiver at the feed location, but the polarization would be wrong.
>> Which was what I reacted on...
>> I am by no means a practical antenna expert, and the EM-theory is a 
>> bit fuzzy on the edges, but I do distinctly recall that signal is RHC 
>> and reflections becomes LHC so an antenna with RHC orientation will 
>> provide some first-degree damping of the LHC reflections. For this 
>> antenna setup the intended RHC signal is reflected and should become 
>> LHC... just as the interference... so it relies on the antenna gain of 
>> the dish to out-perform the other reflections for the half-space 
>> receiver that a normal GPS antenna is. The choke ring for a dish head 
>> has a distinct different pattern (forming an inner cone rather than 
>> flat space).
>> So, a normal antenna would kind of work since the antenna gain would 
>> overcome the poor LHC supression of a simple RHC antenna... yay.
>> If an LHC antenna was used instead... now we are talking.
>> Cheers,
>> Magnus
> So a dish reflector and a sub reflector and the GPS receiver at the dish 
> would work? What is that
> configuration called? I can't remember at this early hour.

Depends on the relative curvatures and focal points:

Cassegrain if the subreflector convex.
Gregorian if the subreflector is concave parabolic.
Dragonian if the subreflector is concave hyperbolic

IEEE Ant and Prop Magazine a few years back had a series of articles on 
designing them all.

All of them can be done offset or coaxial

Any would conceivably work..  It's all about what your pattern looks 
like, what sort of efficiency you need, any mechanical constraints, etc.

More information about the time-nuts mailing list