[MNAPRS] APRS issue in western MN

[Doug Reed]

You are about 55KM from QXM and about 75KM from CBV. That may be some
of it right there.

According to aprs.fi, you are on the fringe of the CBV coverage
area and it hears a LOT of activity from Fergus Falls, Fargo,
Detroit Lakes, and all the traffic on I-94 and Hwy 10.

You are also on the fringe of the QXM coverage area, and on the far
side of I-94 to boot. But QXM only hears about 20% as many packets as
CBV so there is a lot better chance you'll be heard.

But also remember that aprs.fi only saves and stores the FIRST
reception report it receives. If one station has a faster network
connection,  it might be heard first more often. And if there are
fewer hops to the I-gate, it will be reported more often. But in this
case, the aprs.fi data seems to indicate reception by CBV is VERY
spotty.....

So the problem could be just due to collisions with other packets. It
doesn't really sound like a TNC problem, You are running a KPC-3plus,
QXM is running an OpenTracker 3, and CBV is running UIDIGI on a TNC-2
compatible of some sort. I suppose that CBV could be on a MFJ TNC and
maybe the tone decoder needs a slight center freq adjustment? But CBV
has the only TNC that has adjustments among the three TNCs. But it is
seeing plenty of activity from other stations.

But that doesn't explain why CBV repeats packets digied by your
station more often than it repeats your station's own beacons..... But
does it really?

Have you actually captured the text strings from your KPC-3 and then
actually attempted to count and compare how many times you beacon and
CBV repeats it versus how many times you digi someone elses packet and
your packet is the one repeated by CBV? It looks like you beacon once
per hour and QXM last heard 18 beacons in 45 hours, with a bias toward
hearing the night-time packets rather than daylight. There would be
less traffic on the highway in the night.

CBV only heard 4 packets in the last 4 days and there is no
significant bias toward AM or PM hours. So QXM was hearing maybe 40%
of your beacons but CBV is hearing less than 5%.

I will frequently capture a couple days of text from my TNC so I have
sample activity to look at. If every digi is properly doing no-wait
digipeating as they are supposed to, it can be very difficult to
figure out how far you are being heard, but with enough data, you can
try....  So go through the text capture and count how often ZZQ and
CBV appear in the same line of text. This is something you could do
with a simple BASIC program. There may even be some existing software
for analyzing APRS or packet traffic that might help.

Overall, I don't really see any reason CBV should be hearing and
digipeating packets that your WIDE1 retransmitted, but not be seeing
beacons originating at your station. I guess I'd make one more change
to your config just to see what happens. It looks like you are using
the proper WIDE2-1 outgoing path on your beacons but CBV doesn't hear
you well. So try changing to WIDE1-1,WIDE2-1 for a while and see what
happens. The usual rule is that a fixed station shouldn't use WIDE1-1,
but perhaps in this case it will improve your chances of being heard.
I don't like the idea, but it is worth trying in case one of the
stations in Fergus Falls hears you, or maybe KD0NEL.

Have you ever done a radio coverage map of your station? You can go
over to Radio Mobile Online, create a free account under your
callsign, then enter in your station location as a new site. (It is
probably best to drill down to your street by zooming and centering
the map until you find it.) Be sure to tag it as your home station
too, when you save it. Once that is done you can create a coverage map
based on your station and antenna parameters. What I usually use is
the existing antenna height, gain, coax loss, and TX power. Then
configure the mobile as a default quarter-wave on the car roof, 2M
height & 1dB gain, with 1 or 2dB coax loss and 1uv receive
sensitivity. I use those numbers because they are conservative and
will indicate less range than might actually be obtained. .

The resulting coverage map will show about how far a mobile should be
able to hear your station. The critical part here is that the coverage
map will show you how the terrain is effecting your antenna pattern.
If there is a ridge in one direction, your coverage will be less in
that direction. If you are looking down a river valley, your coverage
will be better in that direction.

After you get your home coverage map working, you can add the
locations of QXM and CBV and run coverage maps based on the info in
their beacons. That will show you where you are from their
perspective. If you are behind a ridge from CBV but down valley to
QXM, that pretty well explains your problem.... It may not prove
anything but the coverage maps can be fun to try.

One simple config change you can try would be to change your beacon
time to 3595 seconds or 3005 seconds, or some similar random
difference from exactly 3600 seconds. Perhaps your beacon is happening
about the same moment as someone elses beacon? I usually make a small
change of 1 to 3 seconds on my beacon times, just to be safe.
Actually, why don't you change back to 30 minute beacons, maybe 1798
seconds? 30 minute beacons for a fixed station is not excessive and
will keep your symbol on the map all the time.

Overall, my guess would be that CBV is just much busier and you are
farther away from it. If you can throw in a terrain problem, it is
likely everything is working as well as can be expected. You might
find significant improvement by raising your antenna another 10 feet.
But I think what you really need is a nice high WIDEn-n somewhere
between Wadena and Alexandria., or preferably in both cities...  :-)
You are beaconing out two repeater frequencies. What are your chances
of moving your WIDE1-1 to one of those sites?

By the way, did you try to contact N0QXM and tell him the digi is down
and ask when it will return? He may not know it is down, or possibly
it will be a permanent loss..... I don't think his beacon message is
in the correct format either, PHG is at the end rather than
immediately after the lat-lon.....  I didn't find an email address for
him, just a street address, and a white pages phone number.

Maybe someone will have some better ideas.....

73, Doug Reed, N0NAS.


On 3/27/14, Keith Lakeman <keith at lakeman.org> wrote:
> I'm looking for a little help in diagnosing an inconsistency I'm seeing in
> APRS with my digipeater node, KC0ZZQ-5.
>
> My node is located near Battle Lake MN.  It's been up since last July.
>
> My nearest digi's are N0CBV-1 - Barnsville,MN and N0QXM-1 - Morris,MN
>
> N0QXM-1 has been the node that consistently repeating my node's
> announcements.  Last weekend that node went down.
>
> N0CBV-1 very occasionally repeats packets that originate from my node.  It
> is very consistent repeating packets that my digi has repeated orginating
> from another node.
>
> I'm trying to figure out why N0CBV-1 will not consistently repeat packets
> that originate from my node.
>
> Any ideas on what could cause this would be appreciated.
>
> thanks,
> Keith
> kc0zzq
> keith at lakeman.org
>