| |
ZL3AI > APRDIG 11.07.03 09:09l 171 Lines 7661 Bytes #999 (0) @ WW
BID : 1593-ZL3AI
Read: GUEST
Subj: TAPR Digest, Jul 08, 2/2
Path: DB0FHN<DB0RGB<OK0PPL<RZ6HXA<HA3PG<KD4GBA<ZL2BAU<ZL3VML
Sent: 030710/1332Z @:ZL3VML.#80.NZL.OC #:4073 [Chch-NZ] FBB7.00i $:1593-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: Re: hidden transmitter problems
From: "Mark Fellhauer" <sparkfel@qwest.net>
Date: Tue, 08 Jul 2003 17:50:30 -0500
X-Message-Number: 7
At 05:17 PM 7/8/2003, Richard Amirault wrote:
>I don't understand .. what has all this to do with "hidden transmitter
>problems"?
I too, was confused... The logic that dictates if the digipeater didn't
hear you (when you thought it should) so you must have been "stepped-on" is
flawed. I've notice that with repeaters and digipeaters, sometimes
moving a few inches means the difference between being heard and not - when
you are dealing with urban and suburban ground clutter. I used to live LOS
from the White Tanks APRS Digi in Phoenix. I could step out on my balcony
- look at the radio towers, but not be able to hit the digi with a 5 watt
HT. Moving six feet either way and I could hit it, no problem. And I
noticed similar conditions with my mobile tracker.
73,
Mark
KC7BXS
----------------------------------------------------------------------
Subject: Re: hidden transmitter problems
From: "Wes Johnston" <wes@johnston.net>
Date: Tue, 8 Jul 2003 20:19:48 -0400
X-Message-Number: 8
If another transmitter is sending a packet and you don't hear him, then you
try to transmit, you screw up his packet and your's doesn't get copied
either. One hidden transmitter just killed two packets.
If kc4pl-9 can't get into the digipeater for over _consecutive_ 5 minutes
while riding down the interstate (and within 1 mile at one point of a
RELAY), then I'd say hidden transmitters are a problem.
Wes
----------------------------------------------------------------------
Subject: Re: hidden transmitter problems
From: "Scott Miller" <scott@home.miller.org>
Date: Tue, 8 Jul 2003 17:39:50 -0700
X-Message-Number: 9
Speaking of collision problems, is there any soundcard modem program that
will output raw, invalid frames? Both for analyzing colliding packets (I
suspect we have some hidden transmitter problems between digis here) and
debugging my own tracker designs, it'd be nice to have something that just
displays raw hex output, without trying to do any FCS checking or anything.
Does such a beast exist?
Scott
N1VG
----------------------------------------------------------------------
Subject: Re: hidden transmitter problems
From: "Mark Fellhauer" <sparkfel@qwest.net>
Date: Tue, 08 Jul 2003 19:59:50 -0500
X-Message-Number: 10
At 07:19 PM 7/8/2003, Wes Johnston wrote:
>If another transmitter is sending a packet and you don't hear him, then you
>try to transmit, you screw up his packet and your's doesn't get copied
>either. One hidden transmitter just killed two packets.
>
>If kc4pl-9 can't get into the digipeater for over _consecutive_ 5 minutes
>while riding down the interstate (and within 1 mile at one point of a
>RELAY), then I'd say hidden transmitters are a problem.
Or every time you transmit, you're being obstructed by physical
objects. Funny thing, but sometimes if you're close to a repeater it
can't hear you because you're below the antenna's reception pattern (or
it's above your TX pattern) - something I observed every day in driving in
Phoenix. Again, flawed logic to assume all reception/transmission
problems in a mobile environment are caused by unintentional
interferece/network congestion.
On numerous occasions, I've found that my one minute TX rate would coincide
with hitting overpasses at one mile intervals on urban highways. Summer
and fall markedly impact digipeating around here due to trees, or more
accurately, their leaf canopies.
The only way to accuately test your theory is to monitor at the digi point,
and even then it's not a cut and dry issue. I know for a fact that I can
hear more APRS traffic at the Sioux Falls Digi than actually gets
digipeated. Some of the packets are very distinct but just too far "down
in the noise" for the KPC-3+ to pick it up and re-transmit it. Now those
packets may be coming in from Rochester, Minnesota, but then again it may
be a 5 watt tracker in Brandon - a dozen miles away. I assure you around
here it's not a congestion issue, but a complex mix of terrain, motion, and
man-made ground clutter. BTW, I've never had a problem with too much
network congestion, even in Southern California.
I still don't understand the point, yes, it's quite possible for multiple
stations to key up at the same time, but you seem to infer in your 1st post
that someone here is proposing to run APRS with acks, - and I seemed to
have missed that request/comment.
Mark
KC7BXS
----------------------------------------------------------------------
Subject: Re: hidden transmitter problems
From: <BobDonnell@arkalmus.com>
Date: Tue, 8 Jul 2003 19:27:05 -0700
X-Message-Number: 11
Not necessarily so - all it takes is any cause for a single bit error in
the packet, as received by the RELAY digipeater. At one mile, signal
strength shouldn't be an issue - unless there's another station with
comparable signal strength at that digipeater site. But things like
multipath, and probably more significant, signal fading (a.k.a. flutter)
damaging the packet signal. As others mentioned, if the elevation of
the digipeater and the mobile are significantly different, their
antennas may not radiate and receive well. This is a function of the
pattern of both of the antennas involved - and high-gain base-station
antennas are normally optimized for strong radiation on the horizon, not
below it.
Other issues in the transmitted signal frequently involved, like over-
or under-deviation. As I learned while working as a product support
technician at AEA, many packet users never adjust the transmit audio
level coming going from the TNC to the transmitter to properly deviate
the transmitter. The default settings for all the the TNC's I'm aware
of result in driving the deviation limiter in the transmitter into hard
limiting, usually significantly distorting the transmitter signal. I've
also heard APRS stations on the air that have such low transmit audio
that other TNCs don't recognize the channel as being busy, and transmit
over the top of these low-audio stations.
Another problem is significant mismatch in equalization between the
receiver its TNC, and transmitter and its TNC, the signal. Poor
equalization makes it difficult for the receiving modem to properly
recognize the transitions from one tone to the other. That difficulty
can cause the receiving TNC to have difficulty in recovering data and
clock from the transmitted signal. Any TNC using software or true DCD
depends on this timing data both for data recovery and for determining
if the channel is busy.
So, using the station in your example, and the digipeater you're
refering to, it could be a problem that kc4pl-9 has, or the digipeater
has - not just that there are hidden transmitters. The problems I
mention above are exacerbated by mobile operation, where there's no
assurance that the packet will remain strong enough to properly decode a
packet - i.e. be received entirely error-free. Another factor can be
the type of encoding used by the tracker. If the station is sending
NMEA data strings, these packets are much longer, with many more bits,
than are MIC-E-type packets. Fewer bits to be transmitted results in
fewer opportunities for errors to happen. Mobile packet, even on a
clear frequency, will have a high percentage of lost packets, unless the
stations are close enough to see on-another, and are in eachothers
antenna patterns.
Hope that introduces you to some alternatives.
73, Bob, KD7NM
---
END OF DIGEST
Read previous mail | Read next mail
| |
