## Eric's Weekly Post #3 (Warning, science enclosed. Do not fold or bend.)

Message boards : SETI@home Staff Blog : Eric's Weekly Post #3 (Warning, science enclosed. Do not fold or bend.)

Previous · 1 · 2 · 3 · 4 · 5 · 6 · Next

AuthorMessage
1mp0£173
Volunteer tester

Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
Message 530555 - Posted: 12 Mar 2007, 23:32:21 UTC - in response to Message 530428.

[quote]It shows up in both polarizations (and the polarizations are linear, but it would be possible to derive circular polarizations from the data stream.)

1. When will we search for helical/circular polarizations?

A circularly polarized signal, received by a linear-polarized antenna is only attenuated by about 3db.

A linearly polarized signal, received on a linearly polarized antenna at a 45 degree angle is also attenuated by about 3db.

So assuming that the multi-beam receiver uses linear antennas 90 degrees apart, there is no difference between a circularly polarized signal and the worst case linear polarization.

Unfortunately, if you have a left-hand circularly polarized signal, and a right-hand circularly polarized antenna, the loss approaches infinity.

... and you'd have twice as much recorded data (and a more complex multibeam receiver on the telescope).
ID: 530555 ·
ML1
Volunteer tester

Joined: 25 Nov 01
Posts: 9231
Credit: 5,994,095
RAC: 1,539
Message 530878 - Posted: 13 Mar 2007, 22:13:12 UTC - in response to Message 530555.

1. When will we search for helical/circular polarizations?

A circularly polarized signal, received by a linear-polarized antenna is only attenuated by about 3db.

A linearly polarized signal, received on a linearly polarized antenna at a 45 degree angle is also attenuated by about 3db.

So assuming that the multi-beam receiver uses linear antennas 90 degrees apart, there is no difference between a circularly polarized signal and the worst case linear polarization...

That's a very good point.

A crafted database search for simultaneous signals for each linear polarisation could infer circular polarisation without the need for clever tricks with the splitters (and all the extra data).

Much more efficient to let the database do some of the work!

Keep searchin',
Martin

See new freedom: Mageia5
See & try out for yourself: Linux Voice
The Future is what We all make IT (GPLv3)
ID: 530878 ·
1mp0£173
Volunteer tester

Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
Message 530888 - Posted: 13 Mar 2007, 22:47:39 UTC - in response to Message 530878.

1. When will we search for helical/circular polarizations?

A circularly polarized signal, received by a linear-polarized antenna is only attenuated by about 3db.

A linearly polarized signal, received on a linearly polarized antenna at a 45 degree angle is also attenuated by about 3db.

So assuming that the multi-beam receiver uses linear antennas 90 degrees apart, there is no difference between a circularly polarized signal and the worst case linear polarization...

That's a very good point.

A crafted database search for simultaneous signals for each linear polarisation could infer circular polarisation without the need for clever tricks with the splitters (and all the extra data).

Much more efficient to let the database do some of the work!

Keep searchin',
Martin

... a bigger question is: does polarization carry any additional information?

There are practical reasons for choosing a polarization -- for example, circular polarization is really good if you are trying to send (or receive) signals from a spinning satellite (big expensive satellites are stabilized, but inexpensive small satellites are not).

... and generally speaking, for terrestrial paths, horizontal polarization is quieter. Vertical polarization is better for mobile-to-mobile and base-to-mobile for purely practical reasons.

It generally doesn't change once a system is designed and installed.

I don't see how anything like that would tell us any more about ET. Maybe a real ET researcher might have a comment....
ID: 530888 ·
ML1
Volunteer tester

Joined: 25 Nov 01
Posts: 9231
Credit: 5,994,095
RAC: 1,539
Message 531208 - Posted: 14 Mar 2007, 12:59:26 UTC - in response to Message 530888.

... a bigger question is: does polarization carry any additional information?

There are practical reasons for choosing a polarization -- for example, circular polarization is really good if you are trying to send (or receive) signals from a spinning satellite (big expensive satellites are stabilized, but inexpensive small satellites are not).

And for that reason, if I were to radiate a beacon signal, I would use circular polarisation so that the signal is kept out of any natural sourced noise and to allow reception by a spacecraft antenna at any orientation.

If we took the "Star Trek" view and aligned everything to the galactic plane then linear polarisations could be used. Then again, are there any features of space that rotate/twist the polarisation plane? (I don't know of any other than star or planet magnetic fields, but...)

... and generally speaking, for terrestrial paths, horizontal polarization is quieter. Vertical polarization is better for mobile-to-mobile and base-to-mobile for purely practical reasons.

I wonder how you could have an circular polarised isotropic radiator?... :-(

Or could you have a 'phased array' for circular polarisation. Phew, that would be an interesting headache!

Or?...

...I don't see how anything like that would tell us any more about ET. Maybe a real ET researcher might have a comment....

It would say that this 'ere signal is deliberately artifical and we are trying to maximise how easily it can be received by you...

That extra 3dB over a misaligned linear polarised signal equates to a greatly increased volume of search space! (If we could recieve it with a dedicated circular polarise antenna. Or can the two linear polarised antenna signals be phased/added for zero loss?)

Keep searchin',
Martin

See new freedom: Mageia5
See & try out for yourself: Linux Voice
The Future is what We all make IT (GPLv3)
ID: 531208 ·
WinterKnight
Volunteer tester

Joined: 18 May 99
Posts: 10271
Credit: 30,621,493
RAC: 565
Message 531236 - Posted: 14 Mar 2007, 16:32:39 UTC - in response to Message 531208.

... a bigger question is: does polarization carry any additional information?

There are practical reasons for choosing a polarization -- for example, circular polarization is really good if you are trying to send (or receive) signals from a spinning satellite (big expensive satellites are stabilized, but inexpensive small satellites are not).

And for that reason, if I were to radiate a beacon signal, I would use circular polarisation so that the signal is kept out of any natural sourced noise and to allow reception by a spacecraft antenna at any orientation.

If we took the "Star Trek" view and aligned everything to the galactic plane then linear polarisations could be used. Then again, are there any features of space that rotate/twist the polarisation plane? (I don't know of any other than star or planet magnetic fields, but...)

... and generally speaking, for terrestrial paths, horizontal polarization is quieter. Vertical polarization is better for mobile-to-mobile and base-to-mobile for purely practical reasons.

I wonder how you could have an circular polarised isotropic radiator?... :-(

Or could you have a 'phased array' for circular polarisation. Phew, that would be an interesting headache!

Or?...

...I don't see how anything like that would tell us any more about ET. Maybe a real ET researcher might have a comment....

It would say that this 'ere signal is deliberately artifical and we are trying to maximise how easily it can be received by you...

That extra 3dB over a misaligned linear polarised signal equates to a greatly increased volume of search space! (If we could recieve it with a dedicated circular polarise antenna. Or can the two linear polarised antenna signals be phased/added for zero loss?)

Keep searchin',
Martin

If you use two receive antenna where the received signals are anti-phase then combining techniques can give a theoretical 6dB improvement in signal to noise ratio.

Andy
ID: 531236 ·
ML1
Volunteer tester

Joined: 25 Nov 01
Posts: 9231
Credit: 5,994,095
RAC: 1,539
Message 531261 - Posted: 14 Mar 2007, 17:56:05 UTC - in response to Message 531236.

If you use two receive antenna where the received signals are anti-phase then combining techniques can give a theoretical 6dB improvement in signal to noise ratio.

Indeed yes for signals that are deliberately transmitted that way with the same antenna arrangement. That would certainly prove whatever is recieved to be a non-natural transmission! (However, would that not actually be some sort of circular polarisation in any case?)

... Now, is there any mileage to be had by pre-processing the recorded signal data before the splitters to remove "common mode" noise betwen crossed polarisations and across all polarisations in a similar way to what is being done to remove the instrument interference at the moment?...

Regards,
Martin

See new freedom: Mageia5
See & try out for yourself: Linux Voice
The Future is what We all make IT (GPLv3)
ID: 531261 ·
1mp0£173
Volunteer tester

Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
Message 531292 - Posted: 14 Mar 2007, 19:51:51 UTC - in response to Message 531208.

... a bigger question is: does polarization carry any additional information?

There are practical reasons for choosing a polarization -- for example, circular polarization is really good if you are trying to send (or receive) signals from a spinning satellite (big expensive satellites are stabilized, but inexpensive small satellites are not).

And for that reason, if I were to radiate a beacon signal, I would use circular polarisation so that the signal is kept out of any natural sourced noise and to allow reception by a spacecraft antenna at any orientation.

If we took the "Star Trek" view and aligned everything to the galactic plane then linear polarisations could be used. Then again, are there any features of space that rotate/twist the polarisation plane? (I don't know of any other than star or planet magnetic fields, but...)

... and generally speaking, for terrestrial paths, horizontal polarization is quieter. Vertical polarization is better for mobile-to-mobile and base-to-mobile for purely practical reasons.

I wonder how you could have an circular polarised isotropic radiator?... :-(

Or could you have a 'phased array' for circular polarisation. Phew, that would be an interesting headache!

Or?...

...I don't see how anything like that would tell us any more about ET. Maybe a real ET researcher might have a comment....

It would say that this 'ere signal is deliberately artifical and we are trying to maximise how easily it can be received by you...

That extra 3dB over a misaligned linear polarised signal equates to a greatly increased volume of search space! (If we could recieve it with a dedicated circular polarise antenna. Or can the two linear polarised antenna signals be phased/added for zero loss?)

Keep searchin',
Martin

Martin,

There are a couple of ways to get circularly polarized signals, but it's usually done with crossed dipoles (or crossed yagis) fed 90 degrees out of phase.

But my point is that we have three choices:

1) Use linearly polarized antennas at 90 degrees, and accept that circularly polarized signals will be 3db down.

2) Use circularly polarized antennas (left-hand and right-hand) and accept that linear-polarized signals will always be 3db down.

3) Use four antennas (horizontal, vertical, left and right) and pick up a lousy 3db on some signals (and record four times the data, where at least two are probably redundant).

My gut feeling is that #3 is a lot more complexity with very little gain in data collected.

... but that's me.

-- Lynn
ID: 531292 ·
1mp0£173
Volunteer tester

Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
Message 531297 - Posted: 14 Mar 2007, 19:59:19 UTC - in response to Message 531236.

If you use two receive antenna where the received signals are anti-phase then combining techniques can give a theoretical 6dB improvement in signal to noise ratio.

Andy

Well, maybe closer to 3db than 6db, but there are so many ways to get more antenna gain.

... and for the subject at hand, we're talking about a second dish, not just a second antenna at the feedpoint. I don't know if we have another nearby parabolic valley or the money to build another Arecibo.

Probably easier to just make that dish bigger.
ID: 531297 ·
ML1
Volunteer tester

Joined: 25 Nov 01
Posts: 9231
Credit: 5,994,095
RAC: 1,539
Message 531724 - Posted: 15 Mar 2007, 16:57:53 UTC

Eric,

How's the interference search and fix progressing?

I've seen a comment somewhere saying that the interference is an instrumentation problem rather than external interference pickup by the antenna.

Yet another wild guess: Have you got an impedance missmatch causing reflections on a (long) signal cable?...

Keep searchin',

Regards,
Martin

See new freedom: Mageia5
See & try out for yourself: Linux Voice
The Future is what We all make IT (GPLv3)
ID: 531724 ·
Graeme Stretton

Joined: 2 Nov 01
Posts: 392
Credit: 349,012
RAC: 0
Message 545714 - Posted: 13 Apr 2007, 22:50:18 UTC - in response to Message 524924.

The antenna, then, at the focus, is only one, 1/2 or 1/4 wavelength long?

While that isn't 100% true, it's a pretty fair generalization.

Dipoles can be much shorter than 1/2 wavelength. Theoretically, a half-wavelength dipole is ideal, but practically, this turns out to be about 5/8 wavelength. I believe this avoids problems with "standing waves". A dipole of one wavelength just doesn't work. Since there is a sine wave involved, half the antenna would be electrically of the opposite potential to the other half. They would cancel each other out.

If a dipole is physically less than half-wavelength, it is usually "fooled" into behaving as a half-wavelength by the use of a "loading coil". Basically, this is a rheostat (variable resistor) which can be adjusted. When CB users adjust the "SWR" (Standing Wave Ratio) of their antenna, this is what they are doing.

Ned's right... horizontally-polarised radio suffers less from noise, but TV stations not only transmit on the horizontal for that purpose. They also assign specific areas for their broadcast, and directional, horizontal antennae limit the broadcast area. Directional antennae also deliver more power to the desired broadcast area. Having said that, the British Services TV in Germany used vertical polarisation (probably still do) to prevent the civilian population from watching British TV!!!!!

Umm, errr ... A loading coil is not a variable resistor. It is an inductor which makes the antenna "appear" longer than it really is.

Ahhh, cross polarisation won't stop you watching any TV just involves about 20dB loss os signal. And in the real world not even that much.

Who says? 40+ years of experiance ;P
ID: 545714 ·
1mp0£173
Volunteer tester

Joined: 3 Apr 99
Posts: 8423
Credit: 356,897
RAC: 0
Message 545727 - Posted: 13 Apr 2007, 23:11:35 UTC - in response to Message 545714.

Ahhh, cross polarisation won't stop you watching any TV just involves about 20dB loss os signal. And in the real world not even that much.

Who says? 40+ years of experiance ;P

In a perfect world, cross polarization is an infinite signal loss.

In the real world, there is always some signal that bounced off or bent around something and is no longer horizontally polarized, or for that matter, vertically polarized -- which is why 20db or so is a much more practical number.
ID: 545727 ·
Eric Korpela
Volunteer moderator
Project developer
Project scientist

Joined: 3 Apr 99
Posts: 1212
Credit: 17,777,546
RAC: 9,077
Message 547057 - Posted: 16 Apr 2007, 4:12:47 UTC - in response to Message 531724.

Eric

Eric,

How's the interference search and fix progressing?

I've seen a comment somewhere saying that the interference is an instrumentation problem rather than external interference pickup by the antenna.

Yet another wild guess: Have you got an impedance missmatch causing reflections on a (long) signal cable?...

Keep searchin',

Regards,
Martin

@SETIEric

ID: 547057 ·
Misfit
Volunteer tester

Joined: 21 Jun 01
Posts: 21790
Credit: 2,510,901
RAC: 0
Message 547074 - Posted: 16 Apr 2007, 5:26:58 UTC - in response to Message 547057.

Eric

Do you know which radar system?

Join BOINC Synergy!
ID: 547074 ·
Eric Korpela
Volunteer moderator
Project developer
Project scientist

Joined: 3 Apr 99
Posts: 1212
Credit: 17,777,546
RAC: 9,077
Message 547254 - Posted: 16 Apr 2007, 17:25:12 UTC - in response to Message 547074.

Eric

Do you know which radar system?

Punta Salinas, 6.2 km west of old San Juan. It's operated by the Puerto Rico Air National Guard.

Fortunately the observatory has a system that can tell you when the radar pulses happen, so we can blank those parts of the data stream.

http://www.naic.edu/~phil/rfi/rdr/puntaSalinas/puntaSalinas.html

Eric
@SETIEric

ID: 547254 ·
Michael Buckingham
Volunteer tester

Joined: 21 Aug 99
Posts: 4551
Credit: 2,870,109
RAC: 6,673
Message 547274 - Posted: 16 Apr 2007, 18:09:35 UTC - in response to Message 520343.

Like I said over a month ago...

Have you figured out what the pulse repitition interval and the pulse duration is?

Personally, my opinion is any pulse repitition less than 100usecs would or should indicate a pulse doppler radar? Or something with a fairly predictable duty cycle...it doesn't even have to be pulse constant...

Is there a weather radar near by? Perhaps your picking up sidelobes or back lobes getting into the sidelobes of the receiving antenna? Sidelobe poke through or sibelobe of a separate transmitter getting into the sidelobe of the receiver? I am not actually suggesting radar, but even perhaps unshielded equipment nearby?

http://www.mikesbawx.org
ID: 547274 ·
Clyde C. Phillips, III

Joined: 2 Aug 00
Posts: 1851
Credit: 5,955,047
RAC: 0
Message 547276 - Posted: 16 Apr 2007, 18:16:54 UTC

The Punta Salinas website table gives the duration of the pulses, 41.2 and 409.6 microseconds. I didnt check the frequencies of of the occurrences of the pulses; they might be there, too. The frequencies of the signals themselves are just below those of Seti, from about 1.2 to 1.4 GHz.
ID: 547276 ·
Neil Walker
Volunteer tester

Joined: 23 May 99
Posts: 288
Credit: 18,101,056
RAC: 0
Message 547310 - Posted: 16 Apr 2007, 20:13:40 UTC - in response to Message 524924.

the British Services TV in Germany used vertical polarisation (probably still do) to prevent the civilian population from watching British TV!!!!!

He he. I've been a little slow spotting this one. ;) The intention was to reduce interference with German domestic broadcasts. The organisation which spends a great deal of effort and money on The BBC World Service has no interest in preventing people receiving it's broadcasts. ;)

Be lucky

Neil

ID: 547310 ·
ML1
Volunteer tester

Joined: 25 Nov 01
Posts: 9231
Credit: 5,994,095
RAC: 1,539
Message 547314 - Posted: 16 Apr 2007, 20:26:24 UTC - in response to Message 547276.

...The frequencies of the signals themselves are just below those of Seti, from about 1.2 to 1.4 GHz.

Supposedly, the signals are blanked for Arecibo. At 25kW, it doesn't need much band spread or much of a side lobe or even just a geographic reflection to make a real mess of trying to observe the universe!

Anyone nearby with camouflage paint, a hacksaw, some wire cutters, and?...

:-O

So how does that problem get fixed other than suffering degraded observations?

Cheers,
Martin
See new freedom: Mageia5
See & try out for yourself: Linux Voice
The Future is what We all make IT (GPLv3)
ID: 547314 ·
ML1
Volunteer tester

Joined: 25 Nov 01
Posts: 9231
Credit: 5,994,095
RAC: 1,539
Message 547317 - Posted: 16 Apr 2007, 20:28:33 UTC - in response to Message 547074.

We tracked it down to a radar system...

Do you know which radar system?

Got yer camouflage paint and fatigues ready?...

;-O

Cheers,
Martin

See new freedom: Mageia5
See & try out for yourself: Linux Voice
The Future is what We all make IT (GPLv3)
ID: 547317 ·
Michael Buckingham
Volunteer tester

Joined: 21 Aug 99
Posts: 4551
Credit: 2,870,109
RAC: 6,673
Message 547325 - Posted: 16 Apr 2007, 20:48:25 UTC - in response to Message 547317.

We tracked it down to a radar system...

Do you know which radar system?

Got yer camouflage paint and fatigues ready?...

;-O

Cheers,
Martin

AN/FPS-117

air traffic control and air search.

http://www.mikesbawx.org
ID: 547325 ·
Previous · 1 · 2 · 3 · 4 · 5 · 6 · Next

Message boards : SETI@home Staff Blog : Eric's Weekly Post #3 (Warning, science enclosed. Do not fold or bend.)