I worked FalconSat-3 reliably for many years using an "arrow" antenna.  It was the 4 element "alaskan arrow" and I did mount it on an AZ-EL rotator on the roof, but it worked fine.  I had a homebrew masthead preamp with the Down East Microwave (DEMI) board but that was mainly because the FT-736R is deaf on 70cm.  The antennae were linearly polarized and it worked pretty well.  I did get some slight increase when I later upgraded CP antennas using the LEO pack, but it was workable without.  We should remember though that FalconSat-3 is pretty large with 1.25 watts of power, but it only has the same sort of 1/4 wave whip antennae that a cube sat would have.

I was using an old 9600 baud TNC from the 1990s.  We should note that software TNCs like Direwolf are considerably better at decoding noisy frames.  See WB2OSZs test results: https://github.com/wb2osz/direwolf.  We can also get another coding gain by using FEC, see https://github.com/wb2osz/direwolf/blob/master/doc/AX25_plus_FEC_equals_FX25.pdf.  I saw a comment early that FEC is compute intensive, but we ran RS(255/223) on all the FOX cubesats.  It did not cope with FEC on receive (unless it is used in the software commands?) but it did perform DSP filtering for the DUV signal.  So we can likely make that work.

With that said, 9600 baud packet is only going to be about twice as fast as 1200 baud unless we make some changes to the protocols.  There is simply too much overhead in each packet and the packets are too short to get the advantage of the additional bit rate.  I intend to do some actual measurements and calculations for this, but the Direwolf author has yet another paper on this here: https://github.com/wb2osz/direwolf/blob/master/doc/Why-is-9600-only-twice-as-fast-as-1200.pdf.  Some of that is not true for our link because we are full duplex, but most of it is true.  The combinations of very short frames, 20 bytes of overhead per frame and TX delays mean that the actual data transmitted does not scale as you would expect.

73
Chris

On Sat, Oct 29, 2022 at 9:18 PM Mark L. Hammond <marklhammond@gmail.com> wrote:
Hi Cory, a few responses below. 

On Sat, Oct 29, 2022 at 8:51 PM Corey Minyard <minyard@acm.org> wrote:
On Sat, Oct 29, 2022 at 07:09:05PM -0400, Andrew Glasbrenner wrote:
> >It's difficult for me to believe that 9600 baud would work on a satellite without FEC.  But, maybe so.
>
> UO-22, KO-23, KO-25, TO-31, AO-51, GO-32, MO-36, Falconsat-3, and literally dozens of cubesats have and do utilize 9k6 without FEC.

What kind of ground station is required? 
As much as you can afford ;).  You can copy a bit of data using a ground plane, but not a good solution for actually using digital birds. 

Do you need circular polarized antennas, antennas with serious gain
Gain yes, circular helps but isn’t essential 
, tracking? 
tracking generally, yes, because of gain
Or can this be done
with more normal equipment, like a 3-element yagi like you can use for
FM voice on a satellite?
FS-3 has a 70cm downlink, so lots of gain is still pretty short. It’ll take several elements. Probably a low noise preamp, too. 


I assume the satellites are stabilized somehow?
Yes, often using just a bar magnet in the bird. 
  Fading will kill 9600
baud.
Many of us have been doing it via satellites for 30 years, so it works ;-)

I ask because I don't have any direct experience with this sort of
stuff.  Maybe I should try to work Falconsat-3, just for experience.
Or if someone in the Dallas area plans a contact I can tag along.

You might consider getting AMSAT’s Getting Started guide.  It has lots of good stuff in it.  

-corey

73, Mark N8MH


--

Mark L. Hammond [N8MH]

-----------------------------------------------
pacsat mailing list -- pacsat@amsat.org
View archives of this mailing list at https://mailman.amsat.org/hyperkitty/list/pacsat@amsat.org
To unsubscribe send an email to pacsat-leave@amsat.org
Manage all of your AMSAT-NA mailing list preferences at https://mailman.amsat.org


--
Chris E. Thompson
chrisethompson@gmail.com
g0kla@arrl.net