| |
PA2AGA > HDDIG 12.11.99 12:24l 190 Lines 7281 Bytes #-9687 (0) @ EU
BID : HD_99_287C
Read: GUEST
Subj: HamDigitalDigest 99/287C
Path: DB0AAB<DB0PV<DB0MAK<DB0HOT<DB0ERF<DB0SHG<DB0BRI<PI8DRS<PI8ZWL<PI8APD<
PI8WNO<PI8HGL<PE1NMB<EA7URC<PE0MAR<PI8VNW
Sent: 991112/0821Z @:PI8VNW.#ZH2.NLD.EU #:16184 [HvHolland] FBB7.00g $:HD_99_28
From: PA2AGA@PI8VNW.#ZH2.NLD.EU
To : HDDIG@EU
Received: from pa2aga by pi1hvh with SMTP
id AA23032 ; Fri, 12 Nov 99 07:25:39 UTC
Received: from pa2aga by pa2aga (NET/Mac 2.3.67/7.5.3) with SMTP
id AA00016807 ; Fri, 12 Nov 99 07:16:53 MET
Date: Fri, 12 Nov 99 07:04:03 MET
Message-Id: <hd_99_287C>
From: pa2aga
To: hd_broadcast@pa2aga
Subject: HamDigitalDigest 99/287C
X-BBS-Msg-Type: B
:>
:> I've watched this from the start, back when TAPR "promised" 9k6 gear
:> to be available in 1985. Although a few of the RF engineering types seem
:> to have done a good job designing some interesting gear, there has been
:> essentially total failure making anything available commercially.
:>
:> I know a bit about RF design, although I've never done it for
:> a living. Enough to understand that the problems are probably
:> in the sales and marketing side and not the engineering side.
: Side note: A member of my family designed and implememnted a 56 kb.
: Spread -Spectrum cellular network five years ago from commercially available
: equipment, off the shelf. During that first year, he had over 150 customers.
: TAPR has been "developing" SS for over five years now, and has yet to market
: one (1) usable unit to a Ham.
Charles... Kindly give us the details regarding the "member o' family"
network, please so that we may independently evaluate what I suspect is
a typical Charlie Brabham "Apples vs. Oranges" analysis... Some specifics
I'd like to see are...
1) The frequencies the "member o' family" network operated at...
2) Was the "member o' family" network DS or FH (or maybe hybrid)?
3) What was the cost per node of the "member o' family" network?
: I guess that's why I, like most Hams, do not pay a lot of attention to TAPR.
:> > Is it any wonder that radio networks are dying of old age?
:>
:> Yup. The lack of gear, and the ease of using the internet.
: It's different in Europe. They have no TAPR there, but they do have several
: small companies started up by Hams that produce usable digital equipment at
: reasonable prices.
<sigh> The Charlie Conundrum...
:> > If we would stop pissing on each other, maybe we can get
:> something
:> > constructive done!
:>
:> No organization exists that could coordinate anything.
:> Hardware, software, network infrastructure: nothing there.
: Probably the single most successful program TAPR has ever gotten involved in
: is the effort to hold back and retard US digital Ham Radio. They've done a
: great job at that and a miserable job of doing anything else.
So says the Charlie...
-jw
>.
------------------------------
Date: Wed, 10 Nov 1999 08:50:55 +0200
From: Paul Keinanen <keinanen@sci.fi>
Subject: German packet radio
On Tue, 9 Nov 1999 19:38:27 -0600, "Steve Sampson"
<ssampson@usa-site.net> wrote:
>>that might be why people are using FM radio's for the FSK transport.
>>Wouldn't the capture effect help compensate for some of the drift? Or am
>>I merely displaying my appalling ignorance of RF systems?
>There's a chart in most math compendiums called a Bessel function chart.
>This is the gadget that tells you how wide your signal is going to be, based
on
>your modulation index (Deviation in kHz divided by Modulating frequency).
>I would suspect that with a 38400 Hz modulation component, that you would
>want a real small modulation index to keep it inside 100 kHz bandwidth.
Why should it be kept inside a 100 kHz bandwidth ? Is this limit in
your amateur radio regulations, if so, what has your national amateur
radio society done to remove these restrictions ?
If this is a self-imposed bandplan, is there really that much traffic
on 70 cm or 23 cm that are just outside the intended packet channel ?
Voluntary band plans are very usable when deciding where to put some
new service and also for solving disputes _when_ two or more services
are wanting use the same frequency band. However, if there is only one
service and no plans for other services in that area for many years,
why degrade the performance of the link just to meet some administra-
tive plans made originally for very different circumstances ?
_When_ there is a demand by other services, _then_ it is time to
reduce the bandwidth by reducing deviation or using better coding.
These usually require a better S/N, forcing intermediate stations to
be erected, but this distributes the costs over a longer period of
time.
Using a large deviation reduces the stability requirement and filter
accuracy requirements. A large deviation is more resistant to
multipath, when the whole channel is not wiped out by frequency
selective multipath.
Paul OH3LWR
>.
------------------------------
Date: Wed, 10 Nov 1999 09:21:31 GMT
From: nomail@pe1chl.demon.nl (Rob Janssen)
Subject: German packet radio
Hans-Peter Zorn <hpz@gmx.net> wrote:
>Rob Janssen <nomail@pe1chl.demon.nl> wrote:
>> Do you have an estimation of the ratio between stations using off-the-shelf
>> Japanese transceivers and those using faster "kit" designs?
>> (the T7F is 9k6 as well...)
>Yes, but for user access 9k6 is sufficient.
??? I don't understand what you mean...
When 9k6 is sufficient for user access, then why have telephone modem
manufacturers pushed the limit up to 56K, on a single-user channel?
> And AFAIK you can do 19k2 with
>the T7F if you modify it. Concerning the ratio between commercial and
>homemade: The T7F is available both as kit and assembled. I know lots
>of local DARC Clubs (OVs) which are currently doing T7F-assembly weekends.
There is probably an interest group for assembling kits and/or to buy
an assembled T7F, maybe it is larger in Germany than it is over here.
Typical scenario here: newbie asks "what should I buy for really fast
packet". It is recommended to buy a fast surplus x-tal controlled
transceiver or one of the German kits, plus a G3RUH-type modem kit and
an SCC card. Newbie listens and thanks you for the advice.
Then, he visits the local shiny-boxes store. Salesman talks to him, and
a few days later we see the proud owner of a Japanese "9600 ready"
transceiver (boasting a 1E-4 bit error rate in QST) and an AEA PK96,
TXDELAY 300ms.
With a sufficient installed base of this crap (which usually also jams
the surrounding channels with keyup clicks) it will be very difficult
to move up to higher speeds, even to effciently use the 9k6 speed.
>Higher speed user-access is very rare. In Darmstadt they have a 230400
>Bit/s Link on 3cm now and are planning a high speed access on 13cm.
>There are 2 or 3 76k8 access points on 70cm, the problem here is that
>there is only one channel for this bandwidth on 70cm. Interlinks with
>76k8 are quite common, there are at least 15 of them. A very nice example
>is DB0HUN which has 2 interlinks on 6cm, distance 23km and 68Km, running
>at 76k8, no user access. On weekends there are usually abt 100-160 Qso
>routed via this node, transferring about 500 Kbytes in 10 minutes.
>The problem is, that there isnīt enough traffic to get those nodes to
>their limits :-).
No wonder, when the user access speed is limited.
It should be no problem to get a 76k8 link completely saturated when
sufficient bandwidth to the users is available. Ask an Internet provider.
To be continued in digest: hd_99_287D
Read previous mail | Read next mail
| |
