Circular polarisation : How-to/FAQ

2old4this

Honorary Admin
Joined
Jan 1, 1999
Messages
1,658
Reaction score
0
Points
0
My Location
Cloud Cuckoo Land
Here's my FAQ as originally posted on this board, here:
http://www.satellites.co.uk/php-bin/forum/showthread.php?s=&threadid=286


-------------------------------------------
FAQ: CIRCULAR POLARISATION
2old4this
v1
29-Aug-2001
--------------------------------------------

WHAT IS IT?
-----------
Polarisation is a property of any electromagnetic wave, including the microwaves employed in satellite tranmissions. We are all familiar with Polaroid sunglasses which block out light in a given plane to (eg) reduce reflections off water (reflecteed light itself being polarised in a single plane). Satellite signals are either linearly or circularly polarised.

The whole purpose of polarisation of satellite transmissions is to enable twice as many channels to be squeezed into and distinguishable within a given frequency. Why twice as many? Well simply because each form of polarisation has only two states which will not interfere with each other. The two states of linearly polarised signals are H (horizontal) and V (vertical), indicating the plane in which the signal is polarised. The two states of circularly polarised signals are R (Right) and L (Left), those being analogous to the direction of rotation of the signal.

In actual fact, Circularly polarised signals can be viewed as comprising two constantly varying linear components.
Imagine standing in the path of the incoming signal, looking up into its beam. If you could see the signal it would be in a sense similar to looking at a hand moving around a clock face. Right-polarised signals moving clockwise, left anticlockwise. At 12 and 6 o'clock, all the signal's energy is in the vertical component. AT 3 and 9 o'clock, all of it is in the horizontal component. At other positions, there is a non-zero vertical AND a non-zero horizontal component.

In fact, the signal is "rotating" very many times per second so you don't see really this variation - you see an average strength of the vertical component, and an average (equal) strength of the horizontal. This strength averages out in each case to be about 3db less than the maximum were it to stay at the 3/6/9/12 o'clock position.


HOW DO I RECEIVE IT?
--------------------
Well the first thing to note is that since the signal is effectively a combination of horizontal and vertical components, it can be received in principle by any LNB with a standard H/V polariser - such as a Universal LNB with fixed probes. Nor do you need any special receiver since the receiver only looks at whatever the LNB passes to it, and the LNB is always passing either H or V signals.
However, there are some provisos...
For a start, the H or V probes will only be picking up that weak (average 3db less than total energy) signal. And 3db is a lot - some 50% of the signal's total strength. Effectively reducing it by this much will usually mean (unless you have a massive dish) that there is not enough left to lock on to.


I'VE HEARD ABOUT SKEW. CAN'T I TWIST THE LNB TO MAXIMISE THE SIGNAL?
--------------------------------------------------------------------
When receiving a circularly polarised signal using a standard (eg Universal) LNB, it will make no difference whatsoever how you angle the LNB (or the polarisor in it). The reason why is found in the description above - the "hand" of the "clock-face" is moving round so rapidly that the peaks of reception at 0, 90, 180 and 270 degrees (12/3/6/9-o-clock) are simply smeared in and averaged out with the strengths of the H/V components at all other positions. So however you orient the polarisor, it will always see the same average strength of H signal, and at equal strength to that, the V signal. It can also now be understood that the choice of whether to view that signal as a H or a V signal is arbitrary.


BUT I'VE HEARD OTHERWISE. SOMETHING ABOUT ELLIPTICAL SIGNALS...
----------------------------------------------------------------
Yes, there are reliable folk that insist some circular signals have a linear bias, I.E. are stronger in one "plane" than the other. As if the signal were not circular but rather elliptical (the bias then being the long-axis of the ellipse). In such a case skewing the LNB/polarisor WOULD make a difference. This was often reported for a bunch of signals previously carried on the Thor satellite.
However, other folk deny this...


OK, SO NO SKEW ON STANDARD LNBs. THEN HOW DIO I AVOID THAT 3db LOSS?
--------------------------------------------------------------------
By inserting into the waveguide of the LNB a small piece of dielectric material (such as a specially machined piece of Teflon). That acts as a depolarisor. A depolarisor causes one or other of the linear components to be retarded by a quarter-wavelength. The result is either destructive or constructive interference. I.E. the linear components either cancel each other out, or are added together - the latter forming a stronger signal polarised in either the horizontal or vertical plane.


WILL ANY OLD BIT OF TEFLON DO THE JOB THEN?
-------------------------------------------
No! The dielectric has to have very precise thickness otherwise the retarded component of the signal will end up being retarded by more or less than the essential one-quarter-wavelength (rendering it useless). The required thickness to achieve this is dependent on a property (actually, the "permittivity") of the material.


CAN I INSERT A DEPOLARISOR INTO MY STANDARD UNIVERSAL LNB?
----------------------------------------------------------
Yes, you can. But it may not have the desired result...
To understand why, we need to know more about the effect of the depolarisor.
Yes, it constructs a linear signal from the incoming circularly polarised signal. But it also has an equivalent effect on the incoming linear signals. So assuming you still want your normal H & V channels, the trick is to set it up such that there is minimal affect on the incoming linear signals. This is achieved by carefully aligning the depolariser/Teflon to be exactly horizontal or vertical. In this way, the reconstructed signal is polarised at 45 degrees. But this means it will now not be picked up by the fixed H/V polarisor (probes) of a universal LNB.


CAN'T I FIT THE TEFLON SO THE SIGNAL IT CONSTRUCTS IS AT 0/90 DEGREES?
----------------------------------------------------------------------
Yes, you can do that. And indeed the signal constructed from the incoming circularly polarised signals WILL then be picked up by the H or V probe of the universal LNB. So your receiver will have a strong H or V signal arriving at it and you will thereby have good reception of the channels carried on that L/R beam.
But you will now find that the incoming linearly polarised signals have been negatively affected by the Teflon and THEY are now aligned at 45 degrees to the H/V probes. So you will have circularly polarised channels coming through but not the linear ones!
And yes, you can choose something midway. At least then you would get partial reception of both - but the effective signal strength ariving at the receiver would be reduced in all cases.


WHAT'S THE SOLUTION TO THIS DILEMMA THEN?
-----------------------------------------
The problem is arising because the polarisor of the universal LNB is fixed. So the solution is simply not to use a universal LNB, but instead to use an LNB with a polarisor that can itself be adjusted (or "skewed") - or, of course, a separately controlled polarisor. In either case, we are talking about a magnetic or mechanical (polarotor) polarisor. The receiver will also need to be capable of driving such a device, and that means usually an exensive high-end model is required.
So in summary, we insert and align the depolarisor so that incoming circularly polarised signal are reconstructed at 45 degrees, and then have the receiver apply a 45 degree skew whenever switching to that transponder. Note that whether the signal was originally circularly or linearly polarised, only linear(H or V) signals are arriving at the receiver. So the transponders carrying circularly polarised signals are identified in the receiver's settings as H or V - just like the linearly polarised ones. Only difference is that the skew is applied when switching to channels originating on the circularly polarised beam.


SO I WON'T USE A UNIVERSAL LNB.
BUT WON'T I THEN HAVE PROBLEMS SWITCHING H/V or HIGH/LOW?
---------------------------------------------------------
There are always alternatives for that. Either there will be alternative switching methods, or else LNBs with multiple different outputs passed to an external switch/distribution system. So the theoretical answer is: no, you won't lose functionality.
However, in practise many modern "standard" consumer receivers ONLY operate with universal type LNBs. And since we had already established you need a high-end (or at least non-standard) receiver to drive the polarisor, you had already given up on the idea of using that Murdoch Kiddybox, right...?


I'VE DONE ALL THIS AND STILL I'M NOT GETTING A SIGNAL
-----------------------------------------------------
The final thing to be aware of is that circularly-polarised signals tend to be weaker anyway, and this is especially true in C-Band, where they are most commonly found. C-Band (the "C" has NOTHING to do with "Circular", by the way) is a lower band of frequencies than the Ku band most common in Europe. And one of the properties of microwaves is that the signal concentration at the target area rapidly reduces with reducing frequency. It is simply not possible to focus low-frequency signals as tightly as high-frequency ones. C-band footprints typically cover areas as large as an entire hemisphere, but the signal strength at any given spot is very low. This is why they are ideal for covering the US, but also why such big dishes (2m or more) are needed to reliably receive them.

---------------------------------------------------------------------
<END>
 
Top