| |
ZL3AI > APRDIG 20.12.00 15:51l 181 Lines 7740 Bytes #999 (0) @ WW
BID : 35143_ZL3AI
Read: GUEST
Subj: TAPR Digest, Dec 18, 5/10
Path: DB0ABH<DB0RGB<OK0PPL<OK0PHL<OK0PBB<OK0PAB<HA5OB<HA3PG<PP5BLU<PY3KW<
ZL2TZE<ZL2TZE<ZL2WA<ZL2BAU<ZL3VML
Sent: 001219/0022Z @:ZL3VML.#80.NZL.OC #:51016 [Chch-NZ] FBB7.00g25 $:35143_ZL3
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
to things like multipath distortion.
Debating this issue now is like revisiting the Lincoln/Douglas
Debates. It's time to start rethinking high-speed mobile data
transmission protocols rather than arguing about technology that never left
the starting gate 10 years ago. Let's move forward and forget about what
is largely in the past.
[possible commercial content deleted]
73,
Mark
----------------------------------------------------------------------
Subject: Re: Changeing callsign in APRSdos
From: Arte Booten <n2zrc@webspan.net>
Date: Mon, 18 Dec 2000 13:09:33 -0500
X-Message-Number: 26
Hi all,
Bob B. wrote, in response to Stephen K's:
>>I'd like to add a ssid to my call. How do I change the callsign?
with:
>Edit the first two lines in the C:\CFIG8??.APR file
When you do this, be sure you overwrite the empty spaces between
the callsign and datestamp. 73
+----------------------------------------------------------------------+
|Arte Booten <n2zrc@arrl.net> AEC for Digital Services, NYC ARES/RACES|
|Riverdale, New York [FN30bu] !4052.71N/07354.06WNPHG5370/A=00240|
|Packet: N2ZRC@KB2VLX.#BRONX.NY.USA.NA HTX-202/RS Brick/KPC-3 6.0/X-510|
|PGP Key Fingerprint: D73E B889 C630 6F4A F31F 3083 56BD 0AAD 9996 3B03|
+----------------------------------------------------------------------+
----------------------------------------------------------------------
Subject: Re: 9600bps radios
From: "Robert Donnell" <kd7nm@pugetsound.net>
Date: Mon, 18 Dec 2000 14:02:19 -0800
X-Message-Number: 27
Bob and all,
I think we would all have been better off if Kenwood hadn't even provided
9k6 packet operation at all - just because so far the only radios that the
major manufacturers have made that have a both a well executed receiver and
a well executed transmitter for 9k6 packet, that do not require modification
have been the top-line satellite radios. I do not include the new Alinco or
the D700 in this category, just because there hasn't been any definitive
testing that I've seen. With the exceptions above noted, due to lack of
data, none of the FM-only radios that claim 9k6 operation appear able to
transmit FSK properly.
The manufacturers want to claim the capability, but are unwilling to expend
the engineering effort to do the job correctly.
None of the radios I've seen that perform FSK modulation by swinging the PLL
VCO do it right because the self-correcting characteristics of the PLL
remove the low-frequency components of the 9k6 transmit waveform. This has
the effect of causing the center frequency of the transmitted signal to
wander around as much as 2-3 kHz. This causes the signal to be wider than
it should be, assuming that the FSK frequency changes produce the proper
5.5-6.0 kHz change in frequency (double the correct +/- 2.75-3.0 kHz
deviation). That also means that much too often, the received signal will
deviate out to the edge of the IF (or beyond), where filter group delay
issues rear their heads. But even more important is that the demodulating
modem can get a signal that can't be decoded, because the voltage decision
points that the modem has to use to decide whether an FSK jump occurred or
not can lie outside the voltage swing of the signal received from the radio.
The effects of a transmitter like I'm describing can easily be viewed by
connecting a digital storage oscilloscope (the most common variety, these
days) via DC coupling directly to an FM receiver's discriminator - and it
has to be a DC-coupled discriminator - many chip-based FM demodulators are
not. Observing the signal during those periods where the 9k6 signal has no
transitions for 4 or more bit times will tell the story - if the signal
starts returning to a zero-volt offset from the upper and lower excursions
of the transition at the beginning of that time, then the phase-locked loop
is "correcting" the frequency error imposed by the modulation signal from
the modem in a sufficiently aggressively way that the signal is probably too
distorted to achieve a decent bit-error rate.
And the cost to "do it right" - at the volume that these manufacturers buy
parts, the trim pot (if needed), varicap diode, two or three resistors, and
maybe a choke - should be around $0.50 per band or less. This would be the
price to add parts to 'rubber' the PLL reference crystal around to keep the
PLL believing it was locked, and not needing to perform frequency
correction. Perhaps the biggest expense would be on the alignment bench,
where a square wave (or bandwidth-limited square wave) in the 20-100 Hz
range would need to be applied to the 9k6 modulation input, and the
aforementioned trim pot adjusted to reproduce the square wave at the bench
test receiver's discriminator.
As soon as I get settled in a new home, I'm looking forward to trying mobile
9k6 aprs by way of our local 2 meter 9k6 bit-regenerating repeater. I'll
probably try to get some of the local D700 owners to try their luck too.
And we might also try the 70 centimeter 9k6 repeaters too. I'm very
interested in seeing if the BER is low enough, and/or we're able to get more
packets through due to shorter packets providing less opportunity for mobile
flutter to damage packets.
I'm also hoping that a request I've heard about of Tom McDermott to do an
engineering comparison of 1k2 AFSK over FM vs. 9k6 FSK as data transmission
techniques is able to surface soon.
-----Original Message-----
From: bounce-aprssig-17768@lists.tapr.org
[mailto:bounce-aprssig-17768@lists.tapr.org]On Behalf Of Bob Bruninga
Sent: Monday, December 18, 2000 10:53 AM
>Could this possibly end the Great 9600 baud Debate?
--- rest of message not included ---
Looks like the answer is not yet.
73, Bob
-----------------------------------------------------------------
Robert Donnell, KD7NM (206) 542-7661
18828 Midvale Ave N mailto:kd7nm@pugetsound.net
Shoreline, WA 98133
Western Washington Amateur IP Address Coordinator
-----------------------------------------------------------------
----------------------------------------------------------------------
Subject: Re: 9600 Baud Cxxx
From: Jeff King <jeff@aerodata.net>
Date: Mon, 18 Dec 2000 17:15:57 -0500
X-Message-Number: 28
Mark Fellhauer wrote:
>
>
> The problem being that current amateur radio 9600 BPS technology (even with
> data radios) is not well suited for terrestrial mobile applications. Both
> Phil Karn (KA9Q) and Dale Heatherington (WA4DSY) have written about the
> severe shortcomings of conventional packet schemes especially when it comes
> to things like multipath distortion.
Your 100% right, if your talking bit smearing, at speeds around 56kbaud and
higher
it starts to become a real issue, in particular with omni antennas in urban
canyons.
This is one advantage of DSSS. It is not much of a practical issue at 9600bps,
in
particular if you apply some mild FEC that a PIC could handle.
> Debating this issue now is like revisiting the Lincoln/Douglas
> Debates. It's time to start rethinking high-speed mobile data
> transmission protocols rather than arguing about technology that never left
> the starting gate 10 years ago.
Says you. Quite a few commercial operators are selling 9600bps equipment in
the states, some of it even has decent performance at a good price. And has
Henk has pointed out, the majority of Europe is at 9600bps
or higher. Hardly not leaving the "starting gate", even though things could be
better.
> Let's move forward and forget about what
> is largely in the past.
I'm starting to think I'm in a episode of back to the future. Ahh... Scotty,
Read previous mail | Read next mail
| |