[amsat-bb] Re: S band and Eagle: an appeal for a higher level discussion

Tom Clark, K3IO K3IO at verizon.net
Thu Sep 7 22:04:15 PDT 2006

At the risk of bodily attack, let me offer some comments on the topic
that has produced so much AMSAT-BB traffic. The Eagle Mode-S decision
was made at the conclusion of an intense, but very productive design
meeting held in San Diego, just 2 months ago, at the end of June.

In case you feel that we haven't told you what went on and why the
radical changes, please refer to the very latest (July/August) AMSAT
JOURNAL for some of the details of the meeting. In particular, I ask you
to read Rick's (W2GPS) editorial on Page 3, and Bob's (N4HY)
"Engineering Notebook" beginning with "Warning: Honesty Follows" on
pages 9/10, and the Jim's (WB4GCS) Eagle Update on pages 24/25. Also,
you can glean some details of the meeting and its conclusions in the
K3IO/N4HY presentation at the Central States VHF Society (CSVHFS)
Meeting in July; that material is on my personal website at
<http://mysite.verizon.net/%7Ew3iwi/EAGLE_CSVHFS.pdf> (~3.9 MB in size)
or in its full-blown (a huge 19MB in size) PowerPoint version as

As you can see from the pictures in Bob's Engineering Notebook, the
meeting included your AMSAT President (W2GPS), VP of Engineering (N4HY),
Eagle Project Manager (WB4GCS) and two Board members (W2GPS & K3IO). To
answer one oft-posed question, the entire Board has not met since the
June meeting; the material constitutes the current "best guess"
engineering forecast, and the Board has been told of the conclusions
(and they have read the Journal); but as of now no votes have been
taken. Feel free to contact the Board members if you feel that your
desires have not been taken into consideration. A full list appears on
Page 3 of the Journal, and all can be reached by Email addressed to
call at amsat.org.

After the design group went through careful estimates of the link
budgets for each of the classes of users we considered (see Jim's paper
on Page 24 and/or my CSVHFS presentation), and considered the
feasibility of making the spacecraft work, we came to the conclusion
that my C-C Rider concept wouldn't fly; there was no way we could obtain
enough T/R isolation for full-duplex operation, and using half-duplex
with the time delays associated with a satellite 40,000 km away was
impractical. C'est la vie.

We considered the RFI environment carefully (see my CSVHFS paper for
details) as it would exist 10 years from now. Why 10 years? Well, it
will take us 3-4 years to fly Eagle, and then we hope it will last and
be useful for at least a decade, so a 10 year forecast seemed about
right. In that time frame, many of us (me especially) feel that the
European Galileo GNSS (Global Navigation Satellite System, i.e. GPS)
(and possibly similar systems being designed in Japan & China) will have
prime DOWNLINK signals overlaying our 1260-1270 MHz UPLINK band. The
signals being sent down may well be deemed critical for "Safety of Life"
services; it is not unlikely (double negative used intentionally) that
administrations may cancel out our L-band uplink so those "damned Hams
can't cause an A380 with 500 people on board to crash!". Since weight,
power and volunteer engineering talent are all finite resources, we came
to the conclusion that it was not prudent for us to invest in an L-band
uplink.  C'est la vie.

We  then considered the S1 (13 cm) and S2 (9 cm) allocations.  It's a
pity that amateurs cannot use the S2 = 9 cm (3400 MHz) band in Region 1;
it would be PERFECT as an alternate to L-band. I argued long and hard
for it and was beaten down. However, it might find some use for
telemetry downlinks. The 10 GHz = X-band and higher microwave
frequencies were considered as possible interesting channels for
experimental use, but not really practical for "heavy lifting"
transponders. C'est la vie. I'll come back to S1 & C-band in a bit.

During the discussions we put in hard numbers for each of the possible
bands for each of the classes of users we envisioned (See Jim's article
on Page 23 and my slides). We were already planning on the use of a
Mode-B (70 cm up, 2M down) Software Defined Transponder (SDX) which
would be PERFECT for the legacy "Class-0" users. We then discussed our
desire for a new Class-1 service. This is envisioned as a service which
can send short text messages from portable hand-held terminals, rather
like the SMS messaging service in wide use by teen-agers and similar to
Bob Bruninga's APRS messaging. The idea is based in a desire to have
Amateur Radio ready for the next Katrina or Christmas Tsunami disaster.
When we put in the link budget numbers for hand-held terminals
(including antennas) with a satellite at 40,000 km range, it also
dictated the use of Mode-B. (Detailed spreadsheets justifying this
conclusion are posted on EaglePedia). In fact, Fran Brickle (AB2KT) is
already hard at work on the protocol and SDX code and plans to present a
paper at San Diego.

Now let's turn our attention to the controversial S-band topic. Jan King
(W3GEY/VK4GEY) presented quantitative results that he and Kerry Banke
(N6IZW) had obtained in San Diego, Palo Alto & Queensland on the 2.4 GHz
RFI environment. He indicated that, even in the midst of a field
hundreds of meters from the nearest building, the RFI levels were 20+ dB
above thermal as seen with a non-directional antenna. Others in the
meeting reported anecdotal data that noise levels in some small towns
had increased 60+ dB in three years since WiFi systems were installed.
Several people in the meeting told the effects of the newer (802.11G and
802.11DraftN) 50-100 Mbit/s systems and discussed plans that they new
industry had in the pipeline that will make the situation even worse.
This is especially true on the 10+ year time scales that Eagle must be
designed for. In my CSVHFS paper, you will see a slide that says "S-Band
is a Sewer". Part of this is our own inadvertent fault. With AO-40 &
AO-51, we use S-Band in a receive mode. The non-licensed users have no
idea that we are occupying the frequency. What better way to clear them
out (now here's a radical thought) than by TRANSMITTING instead of
receiving! And we noted the easy availability of surplus S-band power
amplifiers at the 5-30 watt level. Thus S-band lost its "beauty" for
downlink and became a band for uplinks. And we noted that in locations
where a fairly high power ATV repeater was operating on S-band, the QRM
was a LOT less.

Yes, this decision was based on incomplete data. It would be wonderful
if many of you took the challenge to make quantitative measurements of
the S-band RFI levels in multiple locations. Just take your S-band LNA,
put on a low-gain antenna and measure the difference in background noise
level between that antenna and a 50 ohm dummy load as accurately as you
can. You will get better accuracy if you can  use a wide measurement
bandwidth (10-100 kHz or so). Remember to disable your receiver's AGC. A
good spectrum analyzer would be an excellent choice to use as the
detector. Please feel free to send me any such measurements along with
notes on the equipment you used and the environment where you made the
tests (urban, rural, etc). My Email address is mailto:K3IO at verizon.net.

Since L and S2 had been ruled out for reasons described above, this left
us with the possibility of using C-band as a DOWNLINK. There are some
concerns about the 802.11A/cordless phone/etc QRM, but we noted that the
Amateur Satellite DOWNLINK band is above most of the crap (5830-5850
MHz). The work I had done on C-C Rider had shown that, with a phased
array antenna and a power amplifier at each element, we could generate
adequate power so that a Class-A user with a 50 cm diameter (i.e.
DirecTV) dish would have good performance.

The S/C links were designed as digital links. The uplink signals would
have error correction coding that would be decoded at the spacecraft,
and then sent down with "fresh" codes on the downlink. This makes the
satellite links much easier (by tens of dB) than if they were "bent
pipe" linear links. The users would be able to use any type of signal
they desire (Voice, SSTV or even CW). In all this design work, we took
into account that the typical user does not have the digital and S/C RF
widgets. The plan is to develop the user's hardware at the same time, in
parallel with the spacecraft hardware; AMSAT would then make this
available to all. We envisioned using the TAPR TNC-2 model -- TAPR
"seeded" the community by distributing a few hundred kits, and then
licensed the design to commercial manufacturers.

In conclusion, I ask you, the satellite users to consider what will be
the state-of-the-art and the state of amateur radio in the window 5-15
years from now. It may be painful to admit that you invested in an
S-band down converted and barbecue grill dish for a now-defunct AO-40.
Just consider the big picture -- I've been around long enough to think
back on the early 70s: People had cobbled together 2M CW transmitters
for use on Mode-A with AO-6. Then AO-7 came along with Mode-B and the
loud cry went up "Now we have to throw away our 2M transmitters and then
figure out how to make a transmitter that can work on UHF!!". And when
the lower frequency transmitters failed on AO-40, people had to listen
on S-band. Admit it -- it wasn't all that hard to accept new technology.

73 de Tom, K3IO

More information about the AMSAT-BB mailing list