GTM900 support update

[Mychaela Falconia]

Hello FC community,

Earlier today I have added GTM900 target support to our FC Magnetite
firmware.  You can compile FC Magnetite for the gtm900 target, flash
the resulting image into the module with fc-loadtool -h gtm900, and
you will have a GSM+GPRS modem module running FreeCalypso firmware
instead of Huawei's, with approximately the same functionality as the
original, but with a known corresponding source.

I have only done a limited amount of testing, limited by crippled
hardware: the job of adding GTM900 target support to Magnetite took a
few hours, whereas designing and building my desired breakout/interface
board for these GTM900 modules would be a weeks/months task.  The
limited testing I have performed was done using Songbosi's simple FPC
breakout board (turning each FPC pin into a 2.54 mm header pin without
further intelligence), and a very weak power supply arrangement: per
Songbosi's recommendation I took just one out of the provided 5 power
pins and connected it with a female-to-female "Dupont" jumper wire
(probably 28 AWG, i.e., totally unsuited for carrying any substantial
power supply currents) to the +5V pin on the PLDkit FT2232D breakout
board I was using to connect the two UARTs.

Feeding +5V to Calypso GSM devices (both to the core chipset and to
the PA) is not too terribly bad if done for short periods of time (the
chipset is designed to tolerate such high VBAT because according to
TI's docs, NiMH 3-cell batteries can reach a little above 5V during
charging), but the real problem is that by connecting only one power
pin out of 5, using a 28 AWG wire and using USB as the ultimate power
source, this hacked-up arrangement deprives the modem of the current
capacity it was designed for.  The proper solution as I see it would
be to design and build a more proper breakout/interface board that
connects directly to the FPC interface and connects all 5 power and
GND pins with sufficiently thick PCB traces to a proper power input
connector like the big orange one on TI's historical development
boards and on our FCDEV3B, but of course it is more than a one weekend
job.

Amazingly enough, this cheap power arrangement worked well enough to
allow me to test RF Tx: I used Songbosi's little RF adapter that
converts the module's antenna port to SMA, connected it to our CMU200
instrument and ran fc-rfcal-txcheck against both Huawei's original fw
(yes, our tools work against Huawei's fw, that's how little they
changed from TI's baseline) and our own FC Magnetite on this module.
Our Magnetite fw runs with Huawei's original FFS on this target, and
the numbers reported by fc-rfcal-txcheck are the same between the two
firmwares, so it looks like we got the RF part right.  I have not
tested the SIM interface or any end user modem functionality for which
a SIM is required because I don't have a SIM socket hooked up to the
"dumb" breakout, only power and the two UARTs.  However, since I have
verified with my CMU200 setup that the RF Tx output is in compliance
with regulations, it is now safe for other people to test FC Magnetite
on these GTM900 modems, including people who don't have their own
cabled test BTS setups and can only operate in open air.

Going forward, I am hoping to get the fc-host-tools-r11 release out in
another week or two (all major code and documentation work is done,
just a couple of minor bits left), and I need to sync the recent
Magnetite fw developments into Selenite: not only the addition of
GTM900 target support, but also MEMIF changes which affect all targets
- see my new MEMIF-wait-states article in the freecalypso-docs Hg
repository.  Then I will do some more work on the Calypso JTAG
subproject (need to do some experiments with EMU0/EMU1 pins and some
ideas as to how it might be possible to halt the Calypso ARM7TDMI core
directly out of reset), and after that I will need to start working on
little hardware projects.  As to the latter, I have 3 of them in mind:
DUART28 (USB to dual UART adapter with 2.8V I/O), FC-UJA (FreeCalypso
UART+JTAG adapter, a prerequisite for my FC handset board idea) and
the more complete breakout/interface board for GTM900 modules.  The
order in which I am going to work on those little hw projects is
subject to change at any moment based on my mood and feelings. :)

Hasta la Victoria, Siempre,
Mychaela aka The Mother