| |
ZL3AI > APRDIG 04.05.03 18:26l 260 Lines 11139 Bytes #999 (0) @ WW
BID : 1216-ZL3AI
Read: GUEST
Subj: TAPR Digest, Apr 23, 2/4
Path: DB0FHN<DB0ZWI<DB0CHZ<DB0ERF<DB0FBB<DB0GOS<ON0AR<ON0AR<7M3TJZ<HA3PG<
VK3AVE<VK6HGR<ZL2BAU<ZL3VML
Sent: 030425/1334Z @:ZL3VML.#80.NZL.OC #:64548 [Chch-NZ] FBB7.00i $:1216-ZL3AI
From: ZL3AI@ZL3VML.#80.NZL.OC
To : APRDIG@WW
Subject: Re: software to track aprs equiped buggy
From: "Curt Mills, WE7U" <hacker@tc.fluke.com>
Date: Wed, 23 Apr 2003 11:23:50 -0700 (PDT)
X-Message-Number: 15
On Tue, 22 Apr 2003 WA8LMF2@aol.com wrote:
>APRSplus will do this IF the race is going to be in the US. [ APRSplus uses
>Delorme Street Atlas Ver 9.0 or earlier which only covers the US. ]
>
>If the mobile does a good job of smart beaconing ( transmitting whenever it's
>direction changes or has traveled x distance), the resulting black track line
>drawn on top of Street Atlas is excellent, and agrees with the roads in the
>Street Atlas database very closely.
>
>APRSplus itself, running in the mobile, does a very good job of smart
>beaconing if you want to put a laptop in the car, --or-- one could use a
>TinyTrak III in the car (Also does smart beaconing).
I agree that a TinyTrak-II or III would be an excellent fit for this:
Small, light, inexpensive and does the smart beaconing thing very well.
I think most of the APRS applications will draw a track between the points.
If you use Xastir to do it, you can even make a Shapefile map (a real map!)
of the resulting APRS track for posterity. You can make a map of the track
at any time and it time/datestamps the file, so multiple saves of the same
callsign's track won't overwrite each other.
--
Curt Mills, WE7U hacker_NO_SPAM_@tc.fluke.com
Senior Methods Engineer/SysAdmin
"Lotto: A tax on people who are bad at math!"
"Windows: Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me: I picked the coordinate system!"
----------------------------------------------------------------------
Subject: Re: KISS mode PIC / stamp controller?
From: John Hansen <hansen@fredonia.edu>
Date: Wed, 23 Apr 2003 14:35:21 -0400
X-Message-Number: 16
Wes:
The project I'm currently working on may be of interest to you.
Called TNCx, it is a small cheap (~$50) KISS TNC kit to replace my current
PIC KISS TNC. The thing will be quite small (probably about half the size
of the current KISS TNC) and have very low power consumption (I expect to
run it off a 9 volt battery). But what makes this project interesting is
that it is expandable. The "core" module is a stand alone KISS TNC. But
it has an expansion header that brings out power, access to a MAX232 and,
most importantly, the data in and out lines that would normally go to a PC
via a level converter. Thus one will be able to simply plug in discrete
(small, cheap) modules that will do things like:
1. Provide a USB port.
2. Implement a full digipeter.
3. Interface with housekeeping units for satellites.
4. Plug in a HamHud type device.
5. If one wanted to take on the project, it would even be possible to
implement an entire AX.25 protocol machine on a plug in module.
6. Whatever else you might like.
The coding of this project is almost done. The receive chain is working
perfectly and all that remains is to port the transmit code from my current
KISS TNC. I expect to have a prototype of the "core" TNCx at Dayton, but I
doubt that I'll have units for sale before the summer. It seems to me that
once the "core" KISS engine is done, it would be fairly easy to build a
digipeter module that would plug into it.
I thought I would let you know about this because it seems to be a nice
platform for what you want to do.
BTW, for others on the list, I am looking into the possibility of making
the entire project (source code, object code, schematics, etc.) publicly
available and am trying to figure out what mechanism to use for this. If
someone who is familiar with this stuff could point me in the direction of
information on Open Source or GPL (written in English please, not
legal-ese) I would be most appreciative.
John W2FS
----------------------------------------------------------------------
Subject: Re: KISS mode PIC / stamp controller?
From: "Curt Mills, WE7U" <hacker@tc.fluke.com>
Date: Wed, 23 Apr 2003 11:48:30 -0700 (PDT)
X-Message-Number: 17
On Wed, 23 Apr 2003, John Hansen wrote:
>BTW, for others on the list, I am looking into the possibility of making
>the entire project (source code, object code, schematics, etc.) publicly
>available and am trying to figure out what mechanism to use for this. If
>someone who is familiar with this stuff could point me in the direction of
>information on Open Source or GPL (written in English please, not
>legal-ese) I would be most appreciative.
This site has a lot of good info:
http://www.opensource.org/licenses/
The most common choices are GPL and BSD, but there are plenty of
others.
I _think_ the software can come under these licenses but not the
schematics. I think you're just protected by normal copyright rules
there. IANAL though.
--
Curt Mills, WE7U hacker_NO_SPAM_@tc.fluke.com
Senior Methods Engineer/SysAdmin
"Lotto: A tax on people who are bad at math!"
"Windows: Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me: I picked the coordinate system!"
----------------------------------------------------------------------
Subject: Re: KISS mode PIC / stamp controller?
From: "Curt Mills, WE7U" <hacker@tc.fluke.com>
Date: Wed, 23 Apr 2003 11:56:02 -0700 (PDT)
X-Message-Number: 18
Here's another page from the FSF standpoint on some of these
licenses:
http://www.gnu.org/philosophy/license-list.html
And here's some very excellent writing on licenses as well as other
articles:
http://www.stromian.com/Open_Source_Licensing.htm
--
Curt Mills, WE7U hacker_NO_SPAM_@tc.fluke.com
Senior Methods Engineer/SysAdmin
"Lotto: A tax on people who are bad at math!"
"Windows: Microsoft's tax on computer illiterates!" -- WE7U
"The world DOES revolve around me: I picked the coordinate system!"
----------------------------------------------------------------------
Subject: Re: KISS mode PIC / stamp controller?
From: Miller Scott Contr 30SCS/FTI <Scott.Miller@vandenberg.af.mil>
Date: Wed, 23 Apr 2003 18:56:14 -0000
X-Message-Number: 19
I've been thinking of adapting my TinyTrak-style circuit into an ultra-cheap
KISS TNC - no modem chip and no MAX232. It'd be just an HC08 MCU, resistor
ladder D/A, 1-transistor inverter/buffer for RS-232 RX, and NJM2211 FSK
decoder. Should be under $10 in parts, not counting a PCB. My biggest
design problem is buffering - the MCU only has 192 bytes of RAM, so flow
control would be a problem.
The TNCx sounds interesting. Is it MX614 based?
For open source licenses, check out
http://www.asiaosc.org/enwiki/page/Comparison_of_Open_Source_Licenses.html.
I'm no GPL-pushing Linux zealot commie ;], but I think open source is a good
fit with the spirit of amateur radio. I'm sticking to the Modified BSD
license for most of my stuff, which allows it to be incorporated into other
(possibly commercial) products without having to release changes.
Scott
N1VG
----------------------------------------------------------------------
Subject: RE: KISS mode PIC / stamp controller?
From: wes@johnston.net
Date: Wed, 23 Apr 2003 15:10:46 -0400 (EDT)
X-Message-Number: 20
Thats the idea... exactly. Guess I'm not far afield as usual <grin>.
Stamps aren't the best "listeners" on serial lines... I'll probably end up
using this only as a digipeater ID'er for a while. If the proof of concept
works, heck we'll have to go right on up to a PIC WIDEn-n KISS controller.
This could certainly be a boon to mic-e encoders for repeaters.
Wes
----------------------------------------------------------------------
Subject: Re: KISS mode PIC / stamp controller?
From: Miller Scott Contr 30SCS/FTI <Scott.Miller@vandenberg.af.mil>
Date: Wed, 23 Apr 2003 19:46:19 -0000
X-Message-Number: 21
I was looking at serial SRAMs for buffer memory. I was thinking FRAM was
more limited in read/write cycles, but Ramtron's claiming something like 10
billion cycles durability now. I suppose that'll last a while on a typical
packet channel.
My goal is to build a KISS TNC that'll fit in a DB25 connector shell. It's
kind of a background level thing right now, though. With the trouble
everyone is having with serial ports these days, I might just go straight to
a pure USB TNC.
Gotta get my tracker working right now. Stupid interrupt timing jitter...
Scott
N1VG
----------------------------------------------------------------------
Subject: Re: KISS mode PIC / stamp controller?
From: John Hansen <hansen@fredonia.edu>
Date: Wed, 23 Apr 2003 15:45:14 -0400
X-Message-Number: 22
At 06:56 PM 4/23/2003 +0000, Miller Scott Contr 30SCS/FTI wrote:
>I've been thinking of adapting my TinyTrak-style circuit into an ultra-cheap
>KISS TNC - no modem chip and no MAX232. It'd be just an HC08 MCU, resistor
>ladder D/A, 1-transistor inverter/buffer for RS-232 RX, and NJM2211 FSK
>decoder. Should be under $10 in parts, not counting a PCB. My biggest
>design problem is buffering - the MCU only has 192 bytes of RAM, so flow
>control would be a problem.
Yes, flow control (KISS has none) but 192 bytes presents problems on
receive as well. Before you can send a received packet on to the Terminal
(PC or whatever) you need to check the CRC to verify that it was received
ok. That means you have to be able to hold an entire frame in memory at
one time while you check the CRC to decide whether to send it on or
not. So even if you could use all 192 bytes, you wouldn't be able to have
a packet length more than 120 bytes or so (allowing for addressing
information). But it would be even less than this because while you are
checking the CRC on the last frame and transmitting it on the serial line,
you have to be accumulating the next frame. AND at the same time, data may
be arriving that needs to be transmitted, so you have to accumulate that
until the channel is clear. As I understand it, there are some packet
radio applications that use packet lengths as long as 1K. My original TNC
design used a 32K SRAM subdivided 4K for receive and 28K for transmit. I
think, on the whole that was overkill. And it also required an awful lot
of I/O pins to manage the SRAM, which meant a 40 pin PIC. On the new
design I reduced the size by going to an 8K FRAM chip and downsized to a 18
pin PIC. This is what accounts for most of the size reduction in the unit
itself. However, that will give 2K of receive memory (should handle 1K
packets ok) and 6K for a transmit buffer. I'm pretty confident this will
be plenty. 192 bytes would probably be ok to receive most APRS beacons,
but I think it would be not be possible to do a digipeter in that amount of
space. Aside from the memory problem, though, you have a very interesting
approach. Sounds like it could be very small indeed.
>The TNCx sounds interesting. Is it MX614 based?
Yes.
>For open source licenses, check out
>http://www.asiaosc.org/enwiki/page/Comparison_of_Open_Source_Licenses.html.
Thanks for the pointer.
----------------------------------------------------------------------
Read previous mail | Read next mail
| |
