C139 calibration cracked

Mychaela Falconia mychaela.falconia at gmail.com
Wed Nov 22 17:33:30 UTC 2017

Hi DS!

> Thank you for this update regarding calibration on the C139!

An additional update: after my original announcement I figured out how
to make use of the factory Rx channel correction values (i.e., how to
translate them into a form that our FC fw can apply) and made the
corresponding addition to the c1xx-calextr tool.  Still no Tx channel
correction values though, as those involve a lot of unknowns.  See the
doc/Compal-calibration article in the freecalypso-tools repository for
the full current documentation.

> Likewise I prefer option 2, it would be great to have a DP-L10 like
> phone but without the undocumented chipsets. And even if the chipsets
> were actually documented, the DP-L10 is really, really hard to find.

Yes, very true.  My visible-horizon plan is to produce that HSMBP board
I keep wanting to build, and if we do build it, it will indeed be a
Pirelli-like phone without the unwanted camera and WiFi/VoIP subsystems
and without all that undocumented hw, but it will also be a bare board
without a case.  The bare board sans case will be sufficient for UI fw
development needs, which will be its first use.  Designing and
producing a case to turn the HSMBP into a usable phone would be another
big step, but that will also be the point at which I expect to be able
to finally produce UI-enabled libre fw for the existing cheap Mot C139

I also need to remind everyone that I still need to finish a bunch of
work on the FCDEV3B before I can start working on the HSMBP:

* We need to get to the bottom of the sleep mode bug mystery before we
  can start our next board design.  My approach to solving this mystery
  is to do a series of experiments, and because the expirements involve
  making test board builds with different chips, they move very slowly.
  The first experiment is to build a board with RoHS solder using the
  same Calypso, Iota and Rita chip versions as used by Openmoko, and
  the latest update from Technotronix (I talked to them yesterday) is
  that they should have these experimental boards done some time
  around December 7-8.

* I want a new loudspeaker that is no worse than the one in the
  D-Sample handset from TI, and I have yet to find one - although
  perhaps the loudspeaker part itself (the one I got from Digi-Key
  more recently) is fine and the problem is with the lack of proper
  mechanics/acoustics around it - I lack the necessary audio expertise
  to figure it out myself, and will need to hire an audio expert to
  help me.  But in order for a hired audio expert to be able to help
  me get what I want, I need to be able to demonstrate the audio
  quality coming out of the historical TI-made D-Sample handset to
  that expert, so they can hear with their own ears what it is I seek
  to replicate, which means that (1) the hired audio expert will need
  to be local to me in California, and (2) I need to make my FCDEV3B
  setup at least somewhat portable.

* Toward the portable FCDEV3B setup goal, I need to cobble up a
  battery-powered arrangement (probably an 18650 battery holder wired
  up to a Weidmuller connector to power the FCDEV3B), and then get all
  3 pieces (the FCDEV3B, the battery holder and the FT2232D adapter
  board) mounted on a plate of some suitable substrate like plexiglas.

IOW, I've got a ton of work to do along these lines.

> On another subject, REcon has published the video recording for the
> 2017 conference, however I can't see Mychaela's session. I hope it
> will be published soon:
> https://recon.cx/2017/montreal/recordings/

Last week they sent me a private copy of the recording of my talk, and
asked me if I wanted any edits done to it before they publish it.  I
asked them to make one edit, but they haven't got back to me yet.

Their announcement here:


says at the end "the REcon Montreal 2017 recordings are starting appear
online" - since they said "starting [to] appear", I hope that they are
still working on the ones that aren't up yet, including mine.


More information about the Community mailing list