Realistic plan for our own FreeCalypso phone

Mychaela Falconia mychaela.falconia at gmail.com
Sat Oct 29 21:14:50 UTC 2016


Hello FC community,

I am getting the impression that some members of our community are
quite displeased with my de-emphasizing of targeting Mot C1xx or
Pirelli DP-L10 hardware in favor of building our own FreeCalypso hw
instead, and I feel a need to address these feelings and concerns.

My biggest and foremost question to the community is this one: WHY
would you want to use a crippled C1xx or Pirelli phone when our own
FreeCalypso phone hw can be So Much Better?  To put this discussion in
the proper context, I have a concrete plan in mind, a plan which seems
perfectly realistic and doable to me, for how we can produce our own
libre dumbphone that may not look all that pretty in the initial
prototype version, but would totally outperform both the C139 and the
Pirelli in terms of libre-firmware-affecting electronic circuit
functionality.

First, the practical aspects - how can we possibly produce any phone
hw without an astronomical budget?  Here is my plan:

Step 1: Produce a bare board initially, without any kind of case, and
	design this board in such a way that it can be fully usable in
	its bare state with no plastics or other case around it.

If you were to disassemble virtually any "standard" commercially made
phone (Mot C1xx, Pirelli, Neo Freerunner etc), extract its motherboard
and try to use this board in its bare state, you will quickly run into
problems: the way in which the battery connection, the antenna
connection and sometimes other important connections are made in all
"standard" phones relies on the plastic case to attach these components
to the motherboard by pressure.  Working on a sans-plastics phone
motherboard quickly becomes impractical.

My solution: forego this "standard" design and go for a non-traditional
design instead.  Use connectors for *everything* that is not on the
motherboard itself.  For the battery connection, instead of using the
traditional kind of connector that requires the battery to be held in
place by the plastic case, use a connector that takes in a plug, so
that the battery can lie loosely next to the board, attached by wires.
For the antenna, forego the connect-by-pressure gold pads used in the
mainstream phones and put a female SMA connector on the board.  The
actual antenna will then stick out of the top of the phone, like they
used to stick out long long ago before the industry moved to internal
antennas, and will be screwed onto the threaded SMA connector.  Yes,
having the antenna stick out would make the phone look a lot more old-
fashioned, but in return we gain the ability to put it all together
without molded plastics - a worthy trade in my book.

This way we'll be able to build the first bare, sans-case prototype of
our FreeCalypso phone *without* the cost of any plastics, and also
without the cost of hiring a mechanical designer - instead of trying
to make our board in the complex shape designed to fit into a plastic
case like all mainstream commercial ones, make it plain rectangular
instead, and do the whole design of our PCB *without* stressing over
"how will it fit in the case?".  And once we get this bare board built,
we'll be able to fully exercise all of its functionality and do all
firmware development in this bare, sans-case state.

Step 2: Once we have our own FreeCalypso phone in the form of a bare
	board sans case, how do we transform it into something that
	those with the UTMOST devotion to the cause of a libre phone
	will be able to carry in their pockets and purses?

My idea is that I'll take the working bare board and made a very ad hoc
case around it out of some material like plexiglas - I am talking about
the kind of case that can be made in a hobby shop, *not* commercial
quality.  This case will need to hold both the main board and all of
the extras (battery, speakers, microphone etc) that are attached to it
with wires and connectors, but it will be a loose ad hoc design, nothing
like the tight everything-fits-perfectly design found in commercial
phones.

Downsides: the hacked-up home-made phone will very, very obviously look
home-made and not factory-made, the physical construction will very
likely be quite delicate so the user will need to be *very* careful
and gentle with this phone (*no* water exposure of any kind, or even
excessive dust), but I hope that I can make it hold it together just
strongly enough to where I'll be able to carry it in my purse as my
primary phone - remember, dresses don't have pockets, so we ladies
carry purses instead.

Male users may find this hack-phone to be a bit more difficult to carry
in their pockets though - objects carried in a lady's purse (if that
purse is large enough and loosely packed) tend to be subject to less
mechanical stress than those carried in a guy's jeans pocket.  But hey,
you can always try carrying a purse with World's First 100% Libre Phone
in it, or maybe put it on a custom-made belt clip so it won't be
subjected to the mechanical stress of a tightly packed jeans pocket.

OK, so you would have to sacrifice the solid mechanical construction
of a standard commercially made phone for a quite fragile and very
hacky-looking home-made one; what do you get in return?  Here is what
you'll get:

* A guaranteed-working loudspeaker for hands-free calls - contrast
  with the Pirelli DP-L10 and Mot C155/156 where the loudspeaker is
  physically present, but it is very uncertain if we'll ever figure
  out how to make it work.  On our own phone, I will NOT connect the
  loudspeaker in the Pirelli/C155/C156 way; instead I will connect it
  to the Iota ABB's AUX output through a "dumb" loudspeaker amplifier
  - a way which we know will work for hands-free calls.  And of course
  the C139 has no loudspeaker at all.

* A guaranteed ability to make the phone ring - contrast with the
  Pirelli and Mot C155/156 where our ability to make the phone ring is
  quite uncertain.  In order to ensure that we'll be able to make the
  phone ring one way or another, I will equip the hardware with an
  old-fashioned piezo buzzer driving circuit (like on the C139)
  *in addition* to the loudspeaker: this way if we never get TI's
  Melody E1/E2 to work, we'll still have the piezo buzzer as the
  fallback for making the phone ring.

  To my knowledge no commercial phone manuf has ever made a phone with
  both a loudspeaker and a piezo buzzer (those that have loudspeakers
  for hands-free calls also use them to make ringtones), but we are
  not a mainstream commercial phone manuf, and our circumstances and
  needs are different from theirs.

* A 176x220 pixel (probably 2" diagonal) color LCD - larger than any
  pre-existing Calypso phone I know of.  For comparison, Mot C1xx LCDs
  are 96x64 pix, whereas Pirelli's is 128x128 pix.

* A USB port that combines charging with serial access just like on
  the Pirelli, with just one difference: I'll connect the CP2102 to
  the MODEM UART on the Calypso rather than IrDA, so we'll have the
  choice of presenting either RVTMUX or a classic AT command interface
  with CSD and GPRS functions on this USB-serial port.  (Pirelli's use
  of the IrDA UART precludes the latter option.)

* An FPC/FFC (flexible printed circuit / flat flexible cable) interface
  with an add-on debug board just like Openmoko's debug board setup
  for the 2nd Calypso UART (now IrDA) and for JTAG.  This debug board
  hook-up will be required for heavy fw development, but not for casual
  end users: for the latter class the main USB-serial port will be
  sufficient for loading new fw builds, saving and restoring flash
  backups, reading your SMS and contacts out of the phone or composing
  outgoing SMS on your GNU/Linux PC, etc.

All of the hardware features which I have just listed - hardware as in
functionality rather than mechanical construction - are things which I
really, really, really miss on the existing C1xx and Pirelli hw, things
which I miss both as a developer *and* as an end user.  These missing
hw features are important enough to me to make the plan of building
our own phone MUCH more attractive than the alternative of limping
along forever on crippled Mot C1xx or Pirelli hw, attractive enough to
justify both the extra cost and the inconvenience of having a phone of
less-than-factory-quality mechanical construction.

There is also a certain promotional benefit to having a phone of our
own physical make, as "ugly" as it may be, instead of a reprogrammed
Motorola or Pirelli.  Considering how much of my life I have devoted
to the FreeCalypso project, the topic of this project comes up in
virtually every social interaction in my day-to-day life.  But most
people are totally non-technical, and trying to explain the concept of
"firmware" to such non-technical people tends to be rather futile.
Even if we had our own aftermarket fw for the Pirelli or for some Mot
C1xx variant in a practically usable shape, people still probably won't
"get" it - all they would see is some random old phone.  But if I were
to pull our own FreeCalypso phone out of my purse, a phone that very
obviously looks home-made, that might impress even non-technical people
differently.  Something to consider.

So - how do other members of our community feel?  Would you *still*
want to use a Mot C1xx or Pirelli phone with aftermarket fw with all
of its limitations (the inherent hw limitations of each given model
PLUS the limitations of our fw stemming from incomplete understanding
of the alien hw) when you have the option of getting a FreeCalypso
phone instead?  Or would you see reason and join the forward path of
building and using our own FreeCalypso hardware?

In terms of firmware work, the bare-board prototype of our own FC phone
will be the perfect platform for me to do the main bulk of the work on
the UI and other handset functionality layers, including the model-
independent parts which will easily port over to the C139 or to the
Pirelli if that desire is still there.  Thus if some members of our
community still desire usable FreeCalypso fw for the C139 or for the
Pirelli (no loudspeaker, no ring, vibrate only as discussed before),
it will be much easier for us to produce such firmware *after* our own
FC phone, rather than before.

So - comments, opinions, flames?

Hasta la Victoria, Siempre,
Mychaela


More information about the Community mailing list