FCDEV3B hardware bug: sleep mode self-reboot

[Mychaela Falconia]

Hi DS!

> Although I'm no expert in electronics, I'd be glad to give a hand as
> well. I have access to a reasonably fast oscilloscope and of course
> a FCDEV3B :) Just let me know what to probe on the PCB, and I can
> send you a screenshot of the captured signal.

You should start by confirming that your board exhibits the same self-
reboot behaviour as the three that are still with me: put Magnetite-
l1reconst fw in the flash, insert some SIM, boot the board up normally
and confirm that AT+CFUN=1 causes a reboot if not preceded by a sleep-
disabling command.  Once you have established this baseline, you are
ready to start the o'scope probing.

The points to be probed: my first suspect is the V-DBB supply current
path from the regulator in the Iota to the Calypso, which is exposed
on the surface at three points: the relatively big 0805 capacitor at
C214 and two little 0402 caps at C211 and C212.  To find these
capacitors on the board, look at the PDF inside this ZIP:

ftp://ftp.freecalypso.org/pub/GSM/FreeCalypso/fcdev3b/ggg/FCDEV3B-20161214_for_assembly.zip

The drawing of what we see as the top of the board is on PDF page 2,
and should be viewed at high zoom: I have to set zoom to the maximum
of 400% in xpdf for the drawing to become readable.  But it does show
the reference designators of all components on the board, including
the ones that aren't on the silk screen.

If you look at the board such that the antenna connector is on your
left, the SIM socket is on your right and the row of interface
connectors is away from you, the big 0805 cap called C214 will be
right above the Iota chip, toward the Calypso.  The pad on this cap
that faces toward the Calypso is the power rail, and this pad is the
first point which I would like to have probed.  If you probe it with
the board up and running, you should see 1.5 V at this point.  Now
with the probe held firmly in place (a second person will probably be
needed), issue the AT+CFUN=1 command that causes the erratic self-
reboot, and look for any transient voltage drop on the capacitor pad
you are monitoring.

The next point I would like to probe is the same power rail closer to
the Calypso on small cap C211.  It is the rightmost small cap above
the Calypso, between the Calypso and the line of dashes for the metal
shieldcan.  The non-GND pad of interest is the one away from the edge
of the Calypso BGA, toward R106.  Same probing procedure as for C214,
except that this cap is a lot smaller.

The last accessible point on this suspect V-DBB net is another small
cap at C212.  This one is in the "alley" between Calypso and Iota
chips, and is at the top of a group of 0402 components, with a little
bit of empty space above it.  (The empty space is where Openmoko's 0R
jumper-resistors used to be, which I took out hoping to make this
V-DBB supply more solid...)  The non-GND pad of interest on this cap
is the one toward the Calypso.  Same probing procedure as for C211.

It would be great if you and/or Kent can probe these points as the
erratic self-reboot occurs on AT+CFUN=1, and either catch the
suspected voltage drop or rule out any problems on this power supply
rail.

M~