| |
PA2AGA > HDDIG 07.10.99 22:07l 178 Lines 7164 Bytes #-9727 (0) @ EU
BID : HD_99_250B
Read: DL6KCF GUEST
Subj: HamDigitalDigest 99/250B
Path: DB0AAB<DB0ZKA<DB0CRL<DB0TTM<DB0FP<DB0SRS<DB0AIS<DB0NDK<DB0ACH<PI8JOP<
PI8ZAA<PI8HWB<PI8VAD<PI8WNO<PI8HGL<PE1NMB<EA7URC<PE0MAR<PI8VNW
Sent: 991007/1549Z @:PI8VNW.#ZH2.NLD.EU #:5796 [HvHolland] FBB7.00g $:HD_99_250
From: PA2AGA@PI8VNW.#ZH2.NLD.EU
To : HDDIG@EU
Received: from pa2aga by pi1hvh with SMTP
id AA20624 ; Tue, 05 Oct 99 16:26:09 UTC
Received: from pa2aga by pa2aga (NET/Mac 2.3.67/7.5.3) with SMTP
id AA00016231 ; Tue, 05 Oct 99 13:14:12 MET
Date: Tue, 05 Oct 99 13:09:51 MET
Message-Id: <hd_99_250B>
From: pa2aga
To: hd_broadcast@pa2aga
Subject: HamDigitalDigest 99/250B
X-BBS-Msg-Type: B
>NBFM means 20 khz bandwidth? 9600 baud
>maximum signalling rate? Three cases to consider: 1200 baud, 9600 baud,
>and 56k (thus higher bandwidth, probably 100 khz. min).
Sorry, I do not have figures for 56 k, but it was interesting to note
how close the BERR for 300, 1200 and 9600 bit/s are.
>Try it with a desired S/N of 35 db and 20 db fade margin.
>(These two numbers were derived from the desired BER and link availability.)
>(Use others if you like.)
Us I understand, the power level distribution figures include both
long and short time fades, thus for practical purposes there is quite
large fade margin in the 99 % figures.
>Repeat for 70 cm.
Troposcatter losses increase by on 5-6 dB.
>We cannot run the 56k links on 2M in the US.
>
>One useful assumption: antennas on either band have about the same
>capture area, i.e. they are about the same physical size.
Yes, this is true for the receiving antenna and with same size Tx
antenna, the ERP is higher, so you should be able to compensate for
the 5-6 dB extra loss with antennas of identical mechanical size.
However, take care not to use too narrow beams, since this increases
"coupling losses" into the scattering air volume.
>Going higher
>in frequency helps the BER because more of the signal is on-path
>and less is scattered by off-path reflectors
On the other hand the relative bandwidth for a specific deviation
decreases, thus deepening the multipath nulls. However, some
artificial spreadening (e.g. direct sequence spread spectrum) might be
useful to combat multipath.
Paul OH3LWR
>.
------------------------------
Date: Mon, 04 Oct 1999 08:41:02 +0300
From: Paul Keinanen <keinanen@sci.fi>
Subject: 56k UHF-100W / 19.2K VHF -50W
On Sat, 2 Oct 1999 12:12:55 -0700, "Hank Oredson" <horedson@att.net>
wrote:
>Tropo is different depending on the terrain, weather conditions, etc.
Apparently you are talking about normal tropospheric bending, which
very large variations with weather. In this propagation mode, partly
due to the gradual change of the refractive index (and diffraction
near the ground) the signal propagates in the few hundred meters from
the ground. In extreme cases true ducting conditions occur, but on
other cases the signal is brought down prior to the normal optical
horizon. On most hours, the propagation is somewhere between these
extremes.
On the other hand tropospheric scatter occur in irregular air masses
higher up in the troposphere. Both stations must have direct view of a
common scattering air volume, so with distances over 400 km, the
common air volume would have to be in the upper troposphere, where the
air is so thin, that not much scattering occur, thus the losses start
to climb above 400 km paths.
The losses also increase with increasing scattering angle in the
scattering air mass, it increases by 10 dB for each extra degree it
has to bend, thus it is essential that each station has a as low take
of angle as possible. A 500 m mountain at 30 km will increase the
takeoff angle with 1 degree, increasing the path loss by 10 dB, so in
some situations locations further away from the obstacle (and the
other station) would have a stronger signal.
One example of commercial troposcatter links is the link between
Scotland and the North Sea oil rigs, so troposcatter propagation is
not just a curiosity.
For amateur use the problem is that for 200-400 km paths, the path
loss of 200-210 dB for VHF/UHF has to be compensated, which either
requires full legal limit powers to be used or 4-8 long yagis at both
stations. On both cases, it is very hard to construct duplex links,
due to the huge isolation requirement for the duplexer and an array
consisting of multiple yagis is not exactly broadbanded, so it either
works well both on Tx or Rx, but on not on both.
While it is easy to construct a link which manages to forward a few
bulletins (and put the station in the node list :-) in the pre-dawn
hours of the morning due to tropospheric bending, increasing the
reliability and/or throughput is going to be very costly in order to
make it usable for other than forwarding bulletins a few times a day.
As such, the link is as usable as some HF or packet satellites, but
unfortunately this is sufficient to very few applications.
Paul OH3LWR
>.
------------------------------
Date: Sun, 3 Oct 1999 11:36:22 -0700
From: "Hank Oredson" <horedson@att.net>
Subject: 56k UHF-100W / 19.2K VHF -50W
Cathryn Mataga <cathryn@junglevision.com> wrote in message
news:hcFJ3.509$%Z.15684@nuq-read.news.verio.net...
> Paul Keinanen wrote in message ...
> >On Fri, 1 Oct 1999 18:37:13 -0700, "Hank Oredson" <horedson@att.net>
> >wrote:
> > <Link usability times below 90 %>
> >Which is MUCH better than having no link at all, where the usability
> >>time is exactly zero.
> >
> >With usability times that low, the link is usable mainly to bulletin
> >forwarders, for others, such links are much less usable.
> >
> >Paul OH3LWR
>
>
> It's just that Oregon is really rough country up there -- or at least the
> area north of Redding in California up into Oregon, It's painful just to
> drive through -- let alone climbing up those cold mountains to put in
> ham radio gear. Me, I'd get lost or fall off a cliff or get eaten by
> bigfoot, (joke joke joke) there. Most of it is all
> state parks or national forests, so could hams even get permission
> to use this land? I don't think it's so productive to criticize people
who
> live in a place like Oregon, why they don't have fast line of site
> links every 10km.
Mostly, the places one would like to put nodes is state or national forest.
No antennas allowed. At all. We have managed to get permission in
a (very) few cases. Mostly you have to look for private land, and there
is not a lot of that available. When you do find a site, it is most likely
in deep forest with a view in only one direction. There are exceptions.
The Willamette Valley is fairly open, and extends from roughly Portland
south to Eugene, about 150 km. 9600 baud links down the valley
are easy to build, do not require much power, etc. It's "everywhere
else" that is the problem.
> Maybe more interesting is whether these links are going to be
> better than HF forwarding links. Because I would guess off the top
> of my head that 90% reliabillty with an HF link of any kind is pretty
> tough to achieve. It's always seemed to me that the reliability of
> 2meters to bend over a mountain top is slightly better than the
> reliability of, say 80meters or something like this, for a similar
> path. Still, maybe after a few of these hops, that HF starts working
> better because of better range.
For a few years I ran a 1200 baud knife-edge scatter link to Eastern
To be continued in digest: hd_99_250C
Read previous mail | Read next mail
| |
