[aprssig] The ideal APRS digi

[Wes Johnston]

Tie all of this back to the 12v pc mentioned this weekend.

Since I haven't had any luck (ie no experience) with sound card modems 
and AX.25 kernels under linux, we can run win95/98, a 500 meg HDD, 12v 
PC, sound card modem and AGWPE with the new AGWdigi software.  Make sure 
it has a watchdog timer of about 10 minutes watching for PTT to assert 
to reboot the thing.  A 28Ah battery would run it for probably a day w/o 
commercial power.  I have booted my 12v pc from a CF card but worry 
about windows constantly writing log files and wearing it out.

Bob, I have known this all along but have not been able to say it so 
clearly!  "But only the digipeaters in APRS can
hear each other. Hence they can bring in EXTERNAL UNWANTED QRM packets 
more reliably, but this does absolutely nothing to improve the 
reliability of the local user to be heard."  That is the crux of the 
matter. 

In your digi.gif file though, you missed one thing... you need to diode 
OR the CD line and PTT lines to the XCD pins of the opposite TNC.  I'd 
hate to see one of them in the middle of sending a packet and have the 
other decide it wanted to key up and send two packets into the mic jack 
of the radio at once.  When either TNC has it's own PTT asserted it 
needs to assert the XCD of the other TNC.  Of course in most 
applications when one TNC PTT's it will desense the other's RX so in 
reality both could never key up at the same time because both would 
never have complete packets at the same time.

Wes

John Kraus wrote:

> This could also be a prime use for Digi_Ned.  If I am correct Digi_Ned 
> can receive on multiple ports and cross connect the ports and packets 
> while doing n-N. It may even be possible to have the packet identified 
> by port on the output.
>
> Perhaps Henk will add some insight soon.
>
>
> At 07:52 PM 9/26/2004, you wrote:
>
>> Robert Bruninga wrote on 9/26/2004, 4:18 PM
>>
>>  > To clarify what I am proposing for APRS alt-input digis:
>>  >
>>  > The ideal APRS digi should consist of:
>>  >
>>  > 1) RX/TX on 144.39 connected to good APRS style TNC
>>  > 2) RX on alt input channel for locals using any RELAY-only TNC
>>  > 3) External Carrier Detects cross connected between TNC's
>>  >
>>  > See the drawing at: http://www.ew.usna.edu/~bruninga/digi.gif
>>
>> You might note that this setup has EXACTLY the same
>> configuration/integration/DCD issues as adding a Mic-E function to a
>> voice repeater; i.e. additional RX on a non-144.39 channel while
>> providing TX holdoff until 144.39 is clear.
>>
>> I have wrestled with this issue for some time now since no
>> currently-available dual-port TNCs support 1200 baud on BOTH ports.
>>
>> One thought I have had about an inexpensive way to implement this would
>> to use a TNC-X with the serial RX-data-out line connected directly back
>> into the serial TX-data-in line. (Since the TNC-X exclusively uses KISS,
>> the retransmission would be an exact duplicate of the transmission
>> received.  The CD and XCD lines would be cross-connected between the two
>> TNCs as in Bob's diagram
>>
>> Unfortunately, it would also not have the digi's call inserted, or
>> WIDEn-N decremented, but it would probably be the cheapest way to go
>> (except possibly an old TNC-2 clone with a UI-Digi firmware brain
>> transplant).  Presumably the digi's primary on-channel TNC beacons would
>> statisfy the legal requirements for ID.
>>
>>
>>
>>
>> Stephen H. Smith                   wa8lmf (at) aol.com
>>
>> Home Page:                            http://wa8lmf.com
>>
>>
>> _______________________________________________
>> aprssig mailing list
>> aprssig at lists.tapr.org
>> https://lists.tapr.org/cgi-bin/mailman/listinfo/aprssig
>
>
>
>
> _______________________________________________
> aprssig mailing list
> aprssig at lists.tapr.org
> https://lists.tapr.org/cgi-bin/mailman/listinfo/aprssig
>
>