Page Index Toggle Pages: 1 Send Topic | Print
Normal Topic Astra +7 deg and Eutelsat +3.5 deg polarisation offset (Read 1095 times)
Admin1
Forum Administrator
*****
Offline

Welcome to the new forum

Posts: 103
Location: UK
Joined: 2nd Jun, 2017
Astra +7 deg and Eutelsat +3.5 deg polarisation offset
18th Jul, 2017 at 9:51pm
Print Post  
Several people have asked why Astra (+7 deg) and some Eutelsat satellites (+3 deg) have a polarisation offset. This is unusual.

The linear polarisations of Ku band satellites are normally parallel to and at 90 deg to the equator.

The reason for the offset is to minimise cross-polarisation interference during rain for sites to the extreme west, like Portugal, where the polarisation angle seen at the ground approaches a 45 deg tilt angle.

Rain drops tend to have flat horizontal lower surfaces and this significantly affects linear polarisation signals that are tilted at 45 deg, causing both polarisations to interfere with one another.

By reducing the polarisation angle for sites in western europe means that they suffer less cross-polarisation interference during rain.

The main effect of rain is attenuation and noise increase.

Rain causes attenuation by absorbing some of the signal. Rain also increases the system noise temperature since rain is much warmer (~290 deg K) than the extreme cold (~ 3 deg K) of a clear sky outer space. The increase in noise temperature makes the LNB appear noisy and degrades the C/N ratio.

Note that + 7 deg means turn the LNB clockwise by a further 7 deg, when facing forwards towards the satellite in the sky..

I would like to know exactly which Astra and Eutelsat satellites have these unusual polarisation set ups.
  

Admin1
Back to top
WWW  
IP Logged
 
chriswlan
New Member
*
Offline


Posts: 8
Joined: 4th Aug, 2017
Re: Astra +7 deg and Eutelsat +3.5 deg polarisation offset
Reply #1 - 5th Aug, 2017 at 6:32pm
Print Post  
I've also read somewhere that offset antennas are not capable of crosspols as good as non offset systems...

But I don't quite understand why... I'm also curious if this degraded crosspol is only with intermediate skews, and could there be NO degradation when the offset dish has skews near zero, or near 90 deg (VSAT near the Equator)?
  
Back to top
 
IP Logged
 
Admin1
Forum Administrator
*****
Offline

Welcome to the new forum

Posts: 103
Location: UK
Joined: 2nd Jun, 2017
Re: Cross-pol in offset antennas : Mode matched feed
Reply #2 - 5th Aug, 2017 at 8:23pm
Print Post  
Offset antennas do have a worse cross-pol performance, particularly if the F/D ratio is short, like 0.5.

Example F/D=0.65 Cross pol >26 dB within -1 dB contour.

If you have a long feed support arm arm and F/D around 1.0 the cross-pol is much better but you then have mechanical rigidity problems making side struts necessary and you also need a larger diameter feed horn to make sure the power is distributed over the narrower angle subtended towards the dish.



Having a sub-reflector is one solution as the offset subreflector creates opposite cross-pol and this cancels the cross-pol of the reflector. There have been commercial sales of both Ku and Ka band dual reflector systems. The geometry is called Gregorian when the subreflector is cup shaped. The extra reflector costs more and the precise alignment of the two reflectors and the position of the feed is a nightmare with erros needing to be under 1mm.

Example (dual ref): Cross pol >35 dB within -1 dB contour.
 
My feeling is that satellite operators have gradually compromised on high cross pol standards in order to keep user costs users down. Thus we finish up with front fed offset with F/D ratios like 0.8 for higher quality Prodelin dishes (Europe spec) and F/D=0.6 for the budget variety (Elsewhere).

An attempt that I was involved with was the 'mode matched feed' developed by Barry Watson at ERA at Leatherhead, UK. This was taken up by Prodelin and Hughes and used on many of their Ku band VSAT antennas.
The antenna was still front fed offset but used a special wide diameter tapered adaptor between the tx/rx OMT and the horn flange.

http://www.satsig.net/bentley-walker/hx/prodelin-dish-mode-matched-feed.htm



http://www.satsig.net/cgi-bin/yabb/YaBB.pl?num=1303226109

The inside of the taper tube had 3 notches that must be aligned with the feed arm, and the whole dish rotated to set the working polarisation angle. The 3 notches caused additional waveguide modes which exactly cancelled the cross-pol of the dish geometry.  It seems to have faded from use however perhaps due to cost. In excess of 30 dB cross pol was achieved.

It needs to be borne in mind that actual average achieved polarisation alignments are quite poor despite every effort by hubs (and myself!) to help installers get to +/- 1 deg. It makes you wonder why bother making Rolls Royce equipment when the installers can't get it perfect. 

I guess the costs force a trend to higher allowances for cross-pol interference in the link budgets and thus slightly lower capacities and higher outage times in rain. Let's face it, satellite powers have had to be increased to make VSATs smaller and cheaper.  It is the grand total cost of satellites plus all their ground dishes that matters.

Any installers reading this, please note that this is not an excuse for not making an effort to set the linear polarisation accurately!

Fortunately modern Ka band operation is all circular polarisation so as to remove a major skill requirement from installers. Having to install polarisers on the satellite all home dishes in an increased cost but the skill saving is critical if everyone is to be able to install their own dish without help.

More technical references:

http://www.sciencedirect.com/science/article/pii/S1434841115000217

http://www.ticra.com/papers/optimization-procedure-wideband-matched-feed-design

http://ieeexplore.ieee.org/document/7084838/

I don't know if the degraded crosspol of offset dishes occurs only with intermediate skews as there could be NO degradation when the offset dish has skews near zero, or near 90 deg.

I would like to know also.  Please tell us if you know.

Eric
  

Admin1
Back to top
WWW  
IP Logged
 
Nimrod Kapon - Oasis Networks
New Member
*
Offline


Posts: 20
Joined: 13th Jul, 2017
Re: Astra +7 deg and Eutelsat +3.5 deg polarisation offset
Reply #3 - 6th Aug, 2017 at 6:35am
Print Post  
Thank you Eric for the information, I learnt a lot!

About the polarization offset, I heard that in Australia there is (or there was, I heard it about 10 years ago), a satellite used only for TV broadcast, and they offset the polarity so the installers will need to install the LNBF in a vertical orientation instead of tilting it some degrees.
  
Back to top
 
IP Logged
 
Page Index Toggle Pages: 1
Send Topic | Print
Bookmarks:    Facebook    Google    Google+    Linked in    reddit    StumbleUpon    Twitter    Yahoo