Order Tray | Contact Us | Home | SIG Lists

[wxsig] T238/T238+ 1.15.8 Firmware release

Will Beals will at beals5.com
Sat Mar 18 19:54:44 UTC 2006


Folks:

I have just posted the 1.15.8 release for the T238 and T238+ on
http://www.beals5.com/wx.  Select downloads from that menu for the code.
Release notes are copied below.  This release was mainly a big cleanup of a
lot of minor issues and some compatibility issues discovered by George
(SV2DOF).

1) Fixed and E/W bug in position string setup.
2) Fixed a rounding error in metric rain gauge calculations.
3) Updated copyrights to 2006.
4) Changed T238+ source ID from WX1W to AP1WWX for Kenwood receivers (OKd by
Rob).
5) Added 12th Peet field for avg wind speed (but is 5 min, not 1 min!).
6) Finally found the RH calculation error! (error in table lookup on error).
7) Fixed bug that erroneously turned on T238+ transmitter in Computer mode.

Of particular note, I have always been wondering why T238+ data never showed
up on Kenwood radios.  It turns out the default settings are that those
radios ignore any packets who's destination callsign don't begin with AP.
The Peet change isn't quite up to the Peet spec (it calls out a 1-minute
wind speed average, I use the APRS 5-minute average).

WHAT'S NEXT:

While discussing the AP1WWX destination callsign with Rob Bruninga, he did
ask that I make some changes to the way the destination callsign works.  He
has asked me to add a dynamic setting that changes between WIDE3-3, WIDE2-2
and WIDE2-1.

Additionally, I do want to finally start adding support for the X1W-1 EMP
(lightning) sensor.  Counting the ticks won't be hard, but adding support
for this sensor means I need to support up to three DS2430 counters at the
same time.  While this shouldn't be a big deal, it turns out to break my
1-Wire ID discovery routines and will require me to re-write them.  Back in
the very early days I discovered I could cheat compared to the published
routines and found a MUCH simpler ID discovery mechanism that worked as long
as there were no more than two IDs with the same family ID on the network.
The correct way is a lot more involved and not so assembly-language
friendly.

will

===============================================
This message sent using 100% recycled Electrons 
(Minimum 30% post-consumer content)
===============================================
Will Beals






More information about the wxsig mailing list