Re: KPC3+ open-source code?

Re: KPC3+ open-source code?

[Christof Bodner]

Am Montag, den 22.12.2008, 13:56 -0800 schrieb Curt, WE7U:

> *) Open hardware for my shack which I could write new code for would
> definitely be on my wish list.  Whether that's doing my own
> AVR-based TNC, helping with somebody else's AVR project, or waiting
> for Scott's next ARM-based TNC, I don't know.  I'm picky about
> processors and compilers.  I'm not a PIC fan.  The stuff Scott did
> with the OpenTracker series and Motorola processors is great, but
> the compilers cost $$.  Byon with his TT4/AVR project is heading in
> the right direction, but as far as I know it's not open-source.  If
> he has a "normal" bootloader in it perhaps it can be used for other
> things via a GCC cross-compiler.
> 

There is an AVR-based APRS module already available:
http://www.ralfwilke.com/amateurfunk/aprs_tracker/
(this site is in german, but try the Babelfish:
http://babelfish.yahoo.com/ it's not perfect, but you get a feeling of
what's about)

The hardware and software is based on
http://garydion.com/projects/whereavr/ and a kit is available at
http://www.beta.df6ef.de/index.php?option=com_content&task=view&id=57&Itemid=98

Worse luck, it's not! I mailed DF6EF for a kit but he said, this boards
were hand-made and he has no time make more of them. But he said, he
will contact me in case he gets professionally fabricated ones. For this
he needs orders. So who wants to join?

OK, it's "only" an APRS-module but we could think of a processor with a
bigger FlashRAM and/or higher clock rate for more complexity. The best
of all is, that there is an already working software. And the compliler
with the complete IDE is for free. And there is also a very cheap
possibility to program the controller, it's the parallel port. If you
want to see more, have a look at http://www.avrfreaks.net

So this is a "classical" HAM-approach. I'm very much in favor of such a
piece of crap ;-)

73 de Christof, OE8BCK

-- 
             (0-0)
---------oOO--(_)--OOo-----------------------------------------------
 Christof Bodner <oe8bck@gmx.net>
   Linux - Life is too short for reboots
         - First they ignore you. Then they laugh at you. Then they
           fight you(*). Then you win. -- Ghandi
---------------------------------------------------------------------
email: mailto:oe8bck@gmx.net
Locator: JN76DO
---------------------------------------------------------------------
GnuPG public key: 0xB15DA60F (http://wwwkeys.eu.pgp.net)
Fingerprint 5222 4DEA 773A 60B3 22F9  D244 15EF 6645 B15D A60F
---------------------------------------------------------------------

Re: KPC3+ open-source code?

[Chris Kantarjiev]

> Here's the platform for your new KPC3 replacement TNC..

Just say no.

There's really no reason today to not use something that runs OpenWRT
or even full Debian. Why hinder yourself so much?

Re: KPC3+ open-source code?

[Chris Kantarjiev]

I'd have to ask what the goal of such a project is? 9600 bps activity?
Standlone APRS? Something else?

For standalone APRS, my current approach is to use the aprs4r digi
software running on the OpenWRT distro with an OpenTracker 1+ in
KISS mode built into the case. What else are you after, Curt?

73,
chris

Re: KPC3+ open-source code?

[Curt, WE7U]

On Fri, 19 Dec 2008, Chris Kantarjiev wrote:

> I'd have to ask what the goal of such a project is? 9600 bps activity?
> Standlone APRS? Something else?
>
> For standalone APRS, my current approach is to use the aprs4r digi
> software running on the OpenWRT distro with an OpenTracker 1+ in
> KISS mode built into the case. What else are you after, Curt?

Well, considering I don't personally own any KPC-3+'s, I have to
think hard about it.  I want:

*) KISS mode that works properly in multiple TNC's, and not just for
APRS.  Considering the KISS delays that have been reported recently
in long-running TNC's, plus the lack of long packet or back-to-back
packet support in other devices, we could use some updating here and
there.  Back-to-back packets with full window size of 7 1024-byte
packets would be a good start.  Since there are a lot of KPC-3+'s
out there in shacks and on mountaintops, I think an open-source KISS
implementation would be a good start.  Once packets were flying and
the source code was available, I suspect others would add to the
code to implement smart digipeating, dupe-checking, remote
configuration, etc.  This would extend the life of the mountaintop
hardware plus extend the feature set in directions _we_ want to take
it.

*) Open hardware for my shack which I could write new code for would
definitely be on my wish list.  Whether that's doing my own
AVR-based TNC, helping with somebody else's AVR project, or waiting
for Scott's next ARM-based TNC, I don't know.  I'm picky about
processors and compilers.  I'm not a PIC fan.  The stuff Scott did
with the OpenTracker series and Motorola processors is great, but
the compilers cost $$.  Byon with his TT4/AVR project is heading in
the right direction, but as far as I know it's not open-source.  If
he has a "normal" bootloader in it perhaps it can be used for other
things via a GCC cross-compiler.

*) I'd like to get back into TCP/IP over RF from my Linux box at
some point.  I used to have a lot of fun doing that with long
packets on 2m/1200baud, 2m/9600 baud, and on the 220/1200 baud
full-duplex repeater.  That was with AEA PK-88's in KISS mode:  All
three are dead now.

-- 
Curt, WE7U.				archer at eskimo dot com
http://www.eskimo.com/~archer
   Lotto:  A tax on people who are bad at math. - unknown
Windows:  Microsoft's tax on computer illiterates. - WE7U.
The world DOES revolve around me:  I picked the coordinate system!"

Re: KPC3+ open-source code?

[Bill Vodall WA7NWP]

Hey Curt,

Here's the platform for your new KPC3 replacement TNC..

  http://www.sparetimegizmos.com/Hardware/Elf2K.htm

> *) KISS mode that works properly in multiple TNC's, and not just for
> APRS.  Considering the KISS delays that have been reported recently
> in long-running TNC's, plus the lack of long packet or back-to-back


Bill

Re: KPC3+ open-source code?

[Andrew Errington]

On Tue, December 23, 2008 10:56, Curt, WE7U wrote:
<snip>  I'm picky about
> processors and compilers.  I'm not a PIC fan.

One of the projects I forgot to mention in my last roundup was the
WhereAVR- an AVR based modemless APRS thingummybob:

http://garydion.com/projects/whereavr/

It's not a PIC, and it does audio decoding and generation.  Huzzah!

73,

Andrew
ZL3AME

KPC3+ open-source code?

[Curt, WE7U]

Sending this to the linux-hams list as well.  Any interest in the
project?  Perhaps development of a KISS-only ROM for the KPC-3+ as a
start?

I don't have a lot of time to help, but have HC11 hardware to play
with (MIT Handyboards where I can download/test code via serial
port) and can get the compiler and toolchain going again on my home
system.  I can also burn EPROM's to test with.  I'd pitch in where I
could.

---------- Forwarded message ----------
Date: Fri, 19 Dec 2008 10:52:29 -0800 (PST)
From: "Curt, WE7U" <archer@eskimo.com>
Reply-To: TAPR APRS Mailing List <aprssig@tapr.org>
To: TAPR APRS Mailing List <aprssig@tapr.org>
Subject: Re: [aprssig] KPC-3+ KISS mode delayed packet issue?

On Fri, 19 Dec 2008, Curt, WE7U wrote:

> KD7NM near me has the same problem periodically.  Power-cycling the
> TNC fixes it for him too.  He's running it in KISS mode with Xastir.
> It's happened at least 2 or 3 times that I can recall, but seems to
> take months to reappear.
>
> It could be that a KISS frame END character gets missed or
> corrupted, requiring a new KISS frame to come through to supply it,
> but then that one gets stuck because of not enough framing
> characters, awaiting the next packet to come through, etc.
>
> It could also be a memory leak or counters overflowing and losing
> track.  Could be all kinds of things.

Is anyone interested in an open-source replacement that would run on
KPC3+ hardware?  I happen to know where to find an HC11
cross-compiler for Linux and a GPL'ed library to get started with...

      ftp://ftp.eskimo.com/u/a/archer/hc11/

I don't have time/energy to do this myself, but perhaps a largish
group could accomplish it, perhaps with GPL'ed code thrown in from
the digi_ned and opentracker/tracker2 projects?

-- 
Curt, WE7U.				archer at eskimo dot com
http://www.eskimo.com/~archer
    Lotto:  A tax on people who are bad at math. - unknown
Windows:  Microsoft's tax on computer illiterates. - WE7U.
The world DOES revolve around me:  I picked the coordinate system!"

Re: KPC3+ open-source code?

[Bill Vodall WA7NWP]

On Fri, Dec 19, 2008 at 12:08 PM, Curt, WE7U <archer@eskimo.com> wrote:
>
> Sending this to the linux-hams list as well.  Any interest in the
> project?  Perhaps development of a KISS-only ROM for the KPC-3+ as a
> start?
>
> I don't have a lot of time to help, but have HC11 hardware to play

What would be the advantage of investing all this effort in the KPC3
instead of supporting the existing new TT and OT hardware?

Bill

Re: KPC3+ open-source code?

[Curt, WE7U]

On Fri, 19 Dec 2008, Bill Vodall WA7NWP wrote:

> What would be the advantage of investing all this effort in the KPC3
> instead of supporting the existing new TT and OT hardware?

Lots of them in the field?

As far as the Opentrackers, the compilers still cost $$.  To me
that's an impediment to writing open-source code to go on them.  The
up-and-coming ARM-based Opentracker may be a suitable candidate
though.  Also, the Opentrackers up through the Tracker2 and T2-135
don't have a lot of RAM, so back-to-back packets and a full window's
worth of packets is impossible.

TinyTrak's are much in the same boat except for the TinyTrak4.  I
have a cross-compiler for the Atmel Mega processors running on
Linux, and development boards with some Mega processors on them.  I
suspect there's a special bootloader on Byon's systems that'd have
to be replaced if one wanted to start from scratch with his
hardware, but that'd be the only tricky part (for people without a
burner).  I haven't looked to see how much memory and flash comes
with the TT4.

-- 
Curt, WE7U.				archer at eskimo dot com
http://www.eskimo.com/~archer
   Lotto:  A tax on people who are bad at math. - unknown
Windows:  Microsoft's tax on computer illiterates. - WE7U.
The world DOES revolve around me:  I picked the coordinate system!"

Re: KPC3+ open-source code?

[don_reid]

> On Fri, 19 Dec 2008, Bill Vodall WA7NWP wrote:
>
>> What would be the advantage of investing all this effort in the KPC3
>> instead of supporting the existing new TT and OT hardware?
>
> Lots of them in the field?
>
> As far as the Opentrackers, the compilers still cost $$.  To me
> that's an impediment to writing open-source code to go on them.  The
> up-and-coming ARM-based Opentracker may be a suitable candidate
> though.  Also, the Opentrackers up through the Tracker2 and T2-135
> don't have a lot of RAM, so back-to-back packets and a full window's
> worth of packets is impossible.
>
> TinyTrak's are much in the same boat except for the TinyTrak4.  I
> have a cross-compiler for the Atmel Mega processors running on
> Linux, and development boards with some Mega processors on them.  I
> suspect there's a special bootloader on Byon's systems that'd have
> to be replaced if one wanted to start from scratch with his
> hardware, but that'd be the only tricky part (for people without a
> burner).  I haven't looked to see how much memory and flash comes
> with the TT4.

I have been working on an APRS design using an ATMega8 (maybe upgrading
as more resources are needed).  It transmits, but doesn't have receive
ability yet.  One reason for picking that processor is the gcc support,
it is great.

I'd be open to working with others to expand this project.

-- 
Don - W7DMR

Re: KPC3+ open-source code?

[Curt, WE7U]

On Fri, 19 Dec 2008, don_reid@comcast.net wrote:

> I have been working on an APRS design using an ATMega8 (maybe upgrading
> as more resources are needed).  It transmits, but doesn't have receive
> ability yet.  One reason for picking that processor is the gcc support,
> it is great.
>
> I'd be open to working with others to expand this project.

I already have a Mega103 working here with serial and LCD support,
but haven't tackled sending/receiving yet.  Was thinking about going
to a Mega128 and doing DSP for the receive, since the MAC
instructions are pretty fast.

BTW:  The TinyTrak4 uses a Mega processor.  If Byon used a standard
bootloader without modification, it's possible that hardware could
be used by others.  If not, the chip would have to be pulled,
programmed with a standard bootloader, then plugged back in.  Still
not _that_ big of a deal.

As far as I know the TT4 doesn't have open-source code, but I
haven't looked for a while.

-- 
Curt, WE7U.				archer at eskimo dot com
http://www.eskimo.com/~archer
   Lotto:  A tax on people who are bad at math. - unknown
Windows:  Microsoft's tax on computer illiterates. - WE7U.
The world DOES revolve around me:  I picked the coordinate system!"

Re: KPC3+ open-source code?

[Andrew Errington]

Hi,

It's true that an OpenSource drop-in for the KPC3+ would be great... for
KPC3+ owners.  If you have one and want to improve it- go for it.

I'd support the design of an alternative TNC with open hardware and software.

One example is the range of Fox Delta kits-

http://www.foxdelta.com/

The schematics are simple and the PCBs are very affordable.  Reprogramming
a PIC chip to perform different functions is easy (and for some projects
there are alternative firmware files available, tailored to different
applications).

Looking specifically at TNCs, there is a PIC-based modemless design by Bob
Ball WG8WBA here:

http://www.n9cx.com/BobBall.htm

And the same modemless zero-crossing detector technique is also used by
Mike Berg here:

http://www.ringolake.com/pic_proj/zcd/zcdmodem.html

And in Fox Delta's FoxDigi product.  Perhaps you could use the same
techniques for audio input in your projects?

Bob Ball's design is mostly intended as a demonstration of the modemless
receiver technique, but it has a bunch of useful features and the source
code is available.  I have built one, and it works.  I'm going to build
two more and experiment with a three-node APRS setup (one mobile tracker,
one digi and one TNC).  One immediate limitation that comes to mind is the
buffer space for an APRS packet is limited (due to limited RAM in the
PIC), but the limit does not make the design impractical.  It also doesn't
support KISS mode, but that's just software...

If PICs are not your cup of tea then the same techniques could be used on
any micro.

73,

Andrew
ZL3AME



On Sat, December 20, 2008 11:01, Curt, WE7U wrote:
> On Fri, 19 Dec 2008, don_reid@comcast.net wrote:
>
>
>> I have been working on an APRS design using an ATMega8 (maybe upgrading
>>  as more resources are needed).  It transmits, but doesn't have receive
>>  ability yet.  One reason for picking that processor is the gcc
>> support, it is great.
>>
>> I'd be open to working with others to expand this project.
>>
>
> I already have a Mega103 working here with serial and LCD support,
> but haven't tackled sending/receiving yet.  Was thinking about going to a
> Mega128 and doing DSP for the receive, since the MAC
> instructions are pretty fast.
>
> BTW:  The TinyTrak4 uses a Mega processor.  If Byon used a standard
> bootloader without modification, it's possible that hardware could be used
> by others.  If not, the chip would have to be pulled, programmed with a
> standard bootloader, then plugged back in.  Still not _that_ big of a
> deal.
>
> As far as I know the TT4 doesn't have open-source code, but I
> haven't looked for a while.
>
> --
> Curt, WE7U.				archer at eskimo dot com
> http://www.eskimo.com/~archer
> Lotto:  A tax on people who are bad at math. - unknown
> Windows:  Microsoft's tax on computer illiterates. - WE7U.
> The world DOES revolve around me:  I picked the coordinate system!"
> --
> To unsubscribe from this list: send the line "unsubscribe linux-hams" in
> the body of a message to majordomo@vger.kernel.org More majordomo info at
> http://vger.kernel.org/majordomo-info.html
>
>