Posts by pe1chl

    My experience with a freerunning StarCom SR-3602 PLL LNB so far is that it is acceptable for "amateur SSB use" under quiet weather (and hand-on-dial) but when it is rainy/windy or when it is sunny and the sun has just raised or set, not so good.

    Of course it is good for first tries. And I just received my 4-output LNB that I will convert to external reference.

    Is that an SDR-based software approach? There is that idea of locking the SDR to a carrier within the received band that is known to be stable. The beacons could be used for that purpose (after removing the BPSK modulation of course).

    The user would then see a stable tuning dial / waterfall while the software actually adjusts its parameters to keep the beacons at a fixed frequency.

    (it would be best to do this in de demodulator and not by varying the LO frequency of the receiver, as that usually leads to momentary loss of lock)

    I also wasn't too disappointed by the stability of my PLL LNB (a much cheaper StarCom SR-3602, about 8 euro including shipping from China).

    Not good enough for digital modes but for SSB during quiet weather and with the hand on the dial it should so, better than LEO sats!

    In the meantime I have ordered an SR-3604 (4 outputs) to convert it to external LO injection, much more convenient to use one of the extra F connectors than to fix some extra input to this one.

    Peter, do you know if the beacon is transmitted on an antenna with that pattern, or of a broader beam (maybe a hemi like we will get) is being used?

    I was amazed to get 30dB S/N on the beacon, expecting an antenna pattern like the above.

    Or maybe that was during a test of the antenna pattern as you indicated above, when I listened earlier today it was a lot weaker.

    Yes, now it is updating on Celestrak as well!

    ES'HAIL 2               
    1 43700U 18090A   18332.22470325  .00000137  00000-0  00000+0 0  9994
    2 43700   0.0925 247.3206 0001517 355.4163 289.2452  1.00275907   268

    Of course I understand... just a bit impatient to see the signal levels we will get from the transponder ;)

    If the beacon is any indication, it will be very good. I get its carrier at 30dB over the noise on my DTH TV dish which is not really optimal in performance.

    (of course the antenna gain for the beacon may be quite different from the antenna we get, but I am guessing it is also a broader pattern not like the pattern for the DTH transponders, which are quite weak here from the typical satellites operated for that region)

    Please understand that this is a multi-million commercial satellite and we are only a small part of it.

    Apparently those satellite operators are not like us. When we get a new box on the doorstep, we want to try it out, turn the dial, listen for the noise from the receiver, not run the extended selftest that takes a month to complete :)

    Perhaps, nothing has changed in the data of the orbit, so the TLE is not updated all this time?

    The TLE is not updated for some reason, but that reason is not that the orbit hasn't changed! The orbit has changed a lot and the TLE is totally invalid. And so is the tracking offered by sites like N2YO that use the TLE for their calculations.

    Nicest would be a project to construct a 13cm feed that can somehow fit around a satellite LNB. My LNB (and I think most of them) has a feedhorn diameter of 50mm which is a bit large to fit 13cm feeds.... but maybe someone has an innovative idea?

    See my solution:

    Yes it should be possible to solve it using locking, however what we need (in the software) is some possibility to lock on the beacon and then tune to another frequency where the locking that tracks the beacon also makes the tuned qrg track in the same way. I don't think that is already available in one of the SDR programs, but it probably will be in the future when lots of people face this issue and some competent SDR programmer picks it up.

    Is it also a bit more stable than the standard crystal?

    I have tested with an unmodified standard PLL LNB and in steady weather and not around sunrise/sunset it is bearable, but realistically you'll always have to have your hand on the dial for SSB.

    Of course it is possible to lock it to a GPS reference, but I also consider some form of automatic locking (at the IF frequency) to the beacon.

    A calculator is nice to pre-calculate the elevation in case you don't have experience with local TV sats at about the same position, to know what kind of angle (and thus obstacle clearance) you can expect.

    However in my experience it is kind of useless for pointing the antenna, because the inaccuracy you get from not knowing the exact direction of south and not easily measuring the pointing direction of an offset dish is far larger than the pointing accuracy required.

    So in practice it is easiest to use a TV transponder and a settopbox... and then just slowly pan around the expected direction until you find the signal, then optimize on it. A "satfinder" (a beeper that indicates the overall received signal strength) can also be helpful to find the satellite belt and then pan along it.

    In Europe you can always start from a strong transponder on the Astra3 system (23.5 east) and once you have found that turn east finding the next satellite. In Northwest Europe Astra2 (28.2 east) can be used the same way.

    In the past it all was far easier with analog TV than now with digital.

    What we need is a slim LNB (picture 3) with a PLL (and TCXO, OCXO or an external referenz). But until now I could not find such an LNB.

    It looks like the technology for LNBs has changed. I ordered a couple of PLL LNBs in China and received a box that was not even mentioning PLL on it. I mailed back the sender that I got an ordinary LNB while the website said "PLL" and the reply was that it really was PLL. I opened one and he is right, it has a 25 MHz crystal and no DRO.

    So maybe I try to order some slim LNBs and see if they happen to be PLL as well. It could well be that the cost for manufacturing a PLL LNB is now below a DRO type and they just silently switched over. My LNBs were only 7 euro each...

    Do you think it would be feasible to construct a 13cm feed that would fit around a standard TV LNB? I will use my TV dish with a PLL LNB for receive (it is mounted on a H-to-H motor) and I currently have an LNB that has (with the plastic removed) a neck of 20mm and a horn of 50mm diameter. That could be a bit large to accomodate your 13cm feed, isn't it? Another difficulty is that the horn cannot be removed so it will be tricky to get the 13cm feed fitted around the neck.

    That is great news! I wish you good luck with the launch and commissioning!

    And let's remember - no transmissions from amateurs until the transponder has been officially released.


    Please stop taking this discussion into "personal" affairs and as "framing/bashing/judging".

    It does not surprise me that you do it, after all it was you that started that row long ago, but I will not discuss such matters here. Nor will I discuss other remarks you have made about your plans.

    My question to the builders is just about the preference and operating practices to be used on the satellite, no matter by whom, and the reply from Peter and many others has been clear. It is only to serve as a guideline for anyone who is designing a satellite station or copying another design, and has to make decisions about simplex or duplex operation, not as a personal attack to anyone.


    Perhaps someone will write an SDR code which will automatically track the beacon while receiving other signals ;-)

    That is a great idea, it would enable receive using a standard PLL LNB without external GPS reference, which normally is a little too unstable for SSB (and much too unstable for digital modes). When you could point at the beacon and tell the software "keep this at fixed frequency" and then tune around the passband relative to that, I think the free running PLL LNB would be good enough and the mod to feed up 25 or 27 MHz from a GPSDO would not be required.