Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.naic.edu/~phil/obsproblems/faaharmonics.html
Äàòà èçìåíåíèÿ: Fri Jun 7 01:16:13 2002
Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 13:30:42 2012
Êîäèðîâêà:
Ïîèñêîâûå ñëîâà: orion nebula

FAA radar harmonics

june,2002

        The during may02, experiment A1578 (wim van driel) had problems with radar like rfi between 1360 and 1374. These were being caused by the FAA radar at 1350 Mhz. For those that aren't familiar with how the correlator works, I thought i'd give a little more detail.

        The sky signal is normally mixed to 750 Mhz with a 500 Mhz bandpass filter limiting the band. There are 4 separate mixers that can mix portions of this first IF down to 260. There is a 500 Mhz lowpass filter around each of these bands.

        The 4 260 Mhz signals (actually 8 if you count both polarizations) are sent to the correlator where they pass through a 50 Mhz bandpass filter centered at 260 Mhz.The data is then down converted and sampled with an 8 bit 100 Mhz A/D converter. The digitized data is then sent through digital filters that can provide bandwidths of 25Mhz,12.5Mhz,6.25Mhz.. all the way down to 195 Khz.

        The digital filtering provides excellent out of band rejection for signals that are out of the digital filter band. The only thing you need to be careful with is that the digitizers are never saturated (this is what happened to wim).

        An example from wim's experiment..

Let sbc1 be centered at 1369 with a 12.5 Mhz bandwidth digital filter.

the freq range of the sbc	1362.75 to 1375.25
The Faa radar frequency	1350 Mhz
The 50 Mhz digitized band sky range	1344 to 1394 but it is flipped by 3 hi side lo's so Dc=1394, 50Mhz=1344
Offset 1350 from bottom of Dc band	1394-1350= 44Mhz

When the radar points at the observatory, the a/d converter will see a gigantic spike at 44 Mhz. It will go non linear and create harmonics at n*44 with the odd harmonics probably stronger. The birdies will alias at 50 Mhz so in band they will be at:
      44,100-44*2=12, 44*3-100=32,200-44*4=24,44*5-200=20Mhz
The sky frequencies would be:

    - 1350,1382, 1362, 1370, 1374 sky freq
        44, 12,   32,   24,   20 Offset in a/d 50Mhz band.

The 1370 and 1374 birdies will then lay inside the 12.5 Mhz digital filter band so they will not be filtered out.

Moving the center of the sbc will change the location of the 1350 birdie in the 50 Mhz band and cause the birdies to move.
    An example of the faa harmonics has a birdie at 1366 Mhz. This is the 3rd harmonic of the 1350 Mhz band. There are also birdies at 1314 from the 1330 Mhz radar. A 12.5 Mhz digital filter was used and the radar was not within the bandwidth of the digital fitler but it was within the bandwidth of the A/D converter.

The "New " correlator will use a 100 Mhz 12 bit a/d converter instead of the 8 bit 100 Mhz a/d converter used on the "old" correlator. This will give more dynamic range before signals will create harmonics. Tests done with the 12 bit a/d on the ri show that there will still be saturation of the a/d with 12 bits.

So...
        If the center frequency of your subcorrelator is within 25 Mhz of 1330 or 1350, then you should use the radar blanker. The number 25 is a bit soft.... At 25 Mhz the filter is probably only down 3db. We need to do some tests to see how far away we need to be from 1330,1350 before the digitizers don't saturate.. this test needs to also measure this as a funtion of azimuth and za.

        These tests will also be affected by the power level going into the a/d converter. We normally run with a few bits on the noise. This lowers the dynamic range of the 8 bit a/d.

 home_~phil