[amsat-bb] Re: PIC rotator control, and some LVB Tracker history

Howard Long howard at howardlong.com
Wed Oct 14 03:45:17 PDT 2009


> 1 - rotator control in a moving vehicle to keep a small yagi such as an
> Arrow pointed at the APRS signal.
> 2 - antenna parking based on data from a home WX station such as those
> commonly beaconing on APRS systems.

FWIW, I did exactly that many moons ago back in 2003 when AO-40 was still
with us, although it was more of a working prototype than something you'd
want to let all and sundry peruse over.


This was based on a GPS feeding into a PC that corrected the co-ordinates in
real time and fed them to an LVB Tracker, hooking up to a pan and tilt
camera mount.

Since then, in 2005, the LVB Tracker 2 was developed, and that can act
autonomously and take in real time GPS parameters for both position and date
and time correction, although I must stress that this is not a production
strength unit. The LVB Tracker 2 was developed at about the same time as the
SDX back in 2005. When I presented separate papers for the Tracker 2 and the
SDX at the AMSAT-UK Colloquium, there was a lot of interest in the SDX, but
almost none in the Tracker 2, so further development waned I'm afraid. I
worked with Gould Smith on it again quite a bit at the beginning of 2008
(see below), but I still don't believe it's really ready for prime time,
although should some kind soul come up and offer to spend time developing it
further I'd be delighted!

It took a lot of time and effort to get the Tracker 2 to where it is today,
more so than the Tracker 1: the Tracker 2 firmware is five times the size of
that in the Tracker 1. Back in 2005, there was a real limit on what you
could a achieve in a tiny 28 pin device as well as the limitations of
compilers at that time.  The Tracker 2 certainly took to the extreme what
you could achieve with the limited ROM and particularly RAM available at
that time on such a device. It reminded me of the days when you used all
manner of means to squeeze that last byte out of your memory map.

The Tracker 2 has two redeeming features. Firstly, it can operate as a
handheld device with an integrated GPS, but since then the iPhone has come
along! Secondly, Tracker 2 will work autonomously. The bad news is that if
you run a desktop Tracker 2 based on the original Tracker 1 PCB without a
GPS, with there being no RTC clock, you have to reset the date and time
manually each time you switched it on, so that's a bit of a drawback unless
you leave it on all the time. Even then, we found that the crystals drift
the RTC rather too much for our liking. I think if time permitted,
redesigning the LVB Tracker 2 hardware with a new PCB from scratch would be
preferable, supporting a proper battery backed RTC.

FWIW, I am often asked about the choice of compiler for the LVB Tracker.
Some history: the early choice of the BKND compilers for the LVB Tracker was
because, way back when, they offered floating point arithmetic for a
reasonable price unlike other competitive offerings, although you're heavily
restricted as to the complexity of your code, so you have to manually code
around those restrictions. Since then, other more heavyweight compilers have
come along, in particular Microchip's own offerings for the PIC18 and above.
Rewriting the code for those compilers would be 'interesting'(!) for someone
with some time on their hands, and some $$$ to purchase the compiler.

But, some good news. Thanks to Gould Smith's good work testing my code back
in early 2008, I do have some significantly updated Tracker 2 code since
2005, but please bear in mind this is for experimentation purposes, and
current work commitments preclude me from opening this puppy up for now, but
please feel free to do so yourself! See below...

18F2620 programming software:
Hex file: http://www.g6lvb.com/Articles/LVBTracker2/2620test20d.hex
Sources: http://www.g6lvb.com/Articles/LVBTracker2/2620test_2_0d_001.zip

By the way, while researching the content here, I came across some
interesting LVB Tracker development history shots are here (if you're into
that kind of thing):


Judging by the number of PCBs that have been made, I believe that there are
now in excess of 1,000 LVB Trackers out there, although how many are
operational and how many are sitting in pieces waiting for a rainy day I
don't know!

73, Howard G6LVB

More information about the AMSAT-BB mailing list