Discussion Forums

baudline analysis of AMC-07

9 replies [Last post]
sigblips
sigblips's picture
Offline
Joined: 2010-04-20
Posts: 733

I downloaded the AMC-07 data file parts, combined them, and streamed them into the baudline signal analyzer. I wrote up a blog post as a sort of tutorial of the steps I used. There are several screenshots too:

http://baudline.blogspot.com/2010/04/setiquest-amc7-36934464-mhz.html

The wandering sidebands are very interesting. I think they are caused by drift in the sampler clock. Does anyone have any other ideas?

Peter
Offline
Joined: 2010-01-26
Posts: 15
The AMC-7 geosynchronous

The AMC-7 geosynchronous satellite broadcasts some information so the wandering sidebands may be related to the modulation. It may be a slow drift in the frequency/clock reference on the satellite. If so, I would expect the carrier to wander with the sidebands.

It is not due to equipment at the observatory. When we observe deep space probes like Voyager (with a good on-board reference) we only see the frequency drift caused by the rotation of the Earth.

Another possibility is the motion of the satellite. Geosynchs do wander a bit, but I would expect the acceleration (causing a changing Doppler shift) to be very small and slowly changing. The strong carrier would also be shifted with the sidebands by that motion.

Kikuchiyo
Offline
Joined: 2010-04-23
Posts: 3
Hello, I'm still a Windows

Hello, I'm still a Windows (Xp and Seven) user... What software I can use to analyse the data? Thanks in advance ^^

ncc1701zzz
Offline
Joined: 2010-04-22
Posts: 25
I think that the idea is that

I think that the idea is that people with signal processing knowledge could develop better algorithms and software to analyze the data.

In the future, SetiQuest will release a analysis software as open source (I think) so people can look at it and try to improve the algorithms and the implementation.

Regards.

avinash
Offline
Joined: 2010-01-26
Posts: 278
Yes, that is correct. We will

Yes, that is correct. We will release the software as open source. Working on it. This open sourcing will happen in phases. The target date for first phase is in Q3 of 2010.

sigblips
sigblips's picture
Offline
Joined: 2010-04-20
Posts: 733
You could run baudline with a

You could run baudline with a Linux live-CD, Ubuntu 9.10 works well. Since the live-CD is non-destructive and doesn't have write access to your hard drive you could use a big USB flash drive to hold the downloaded setiQuest data files.

Kikuchiyo
Offline
Joined: 2010-04-23
Posts: 3
I see, thanks for the tip

I see, thanks for the tip ^^
Starting to download linux and the data to learn!

sigblips
sigblips's picture
Offline
Joined: 2010-04-20
Posts: 733
Will there be any Doppler

Will there be any Doppler shift due to the earth's rotation since the satellite is in geosynchronous orbit? I'm thinking that answer is no, so would the Doppler shift you mention be due to the satellite wobbling, vibrating, or tides?

I updated my blog post with some more screenshots and a couple new sections. See the new "Decimate by 4096 and 32768" sections. I didn't see it in my original "Decimate by 512" analysis but now I can measure that the carrier has about a +0.4 Hz drift over the 326 second signal duration. The random wandering of the carrier in the "Decimate by 32768" section is really interesting. Do you have any ideas what might be causing that?

robackrman
Offline
Joined: 2010-04-15
Posts: 235
The AMC-7 satellite is



The AMC-7 satellite is slightly inclined and also has a not quite circular (elliptical) orbit. For these reasons, although the satellite position has a stationary mean, it moves in a small figure-eight in the sky and also towards-and-away from the observatory during a sidereal day. Using orbital Keplerian elements (TLEs) and the SGP4 model, I tracked an ECEF x,y,z vector from the observatory (ATA) to the AMC-7 satellite over a recent sidereal day. The maximum d2r/dt (acceleration) of the satellite results in a dΔf/dt of ~ 0.001 Hz ⋅ s-1 by my calculation. Your analysis yields a value a bit higher (0.4/326) but within error of the expected Doppler shift rate-of-change if we happened to catch the satellite at its largest acceleration. I think it is reasonable to claim that you are measuring Doppler shift in your analysis especially given that TLE-based satellite positions can be off by hundreds, and even thousands, of meters. Also, my calculations, which were done quickly, could be wrong. Nice work! To gain more confidence, it would be necessary to observe the satellite a few different times over the course of a day.

robackrman
Offline
Joined: 2010-04-15
Posts: 235
I very much enjoyed viewing

I very much enjoyed viewing the AMC-7 analysis posted in your blog!

The observatory mixers and sample clocks are driven by Agilent frequency generators with extremely low phase noise and 0.001 Hz resolution locked to a 10MHz reference derived from a rubidium standard and long-term trained by GPS 1PPS. Unless we have an equipment malfunction, the "wandering sidebands" should be considered real measurements.

Note that a poor sample clock should affect other straight lines in your spectra at similar resolution, but that is not seen.

I don't know about this case; however, we often see similar erratic and drifting signals associated with harmonics from lesser quality crystal oscillators found on motherboards and other local electronics equipment that leak into the antenna sidelobes and signal chain. A catalog is maintained for some of these signals so that they can be ignored during the search for ET.