This week in the life of CL1B

[John Watlington]

This has been a busy week from the hardware point of view.
It started with our A2 prototypes sitting in a FedEx warehouse in
Anchorage, Alaska, awaiting some paperwork from Quanta, and me  
conducting
a search for the cause of occasional SD problems in our A1 prototypes.
The former was resolved by Tuesday afternoon, resulting in the delivery
of the A2 motherboards by Wed. morning.   The latter was shelved after
running out of A1 motherboards on which to destroy the SD slot in  
hypothetical
rework attempts.   Testing is restarting on the A2s, with promising  
results so far ---
perhaps the power supply improvements helped!

The A2s arrived with a known S3 <-> S0 (suspend/resume) power sequencing
issue, and are being ECO'd, tested for basic suspend resume  
functionality
and placed in an XO-1 case.   Most have their processor serial port  
added,
accessible by removing the LCD panel.   Almost none have a working WLAN
due to a hard to repair design error.   More details at:
http://wiki.laptop.org/go/XO_1.5_A2

So far, the tally is:
18 laptops assembled to various stages
1 w. video problems
1 that won't power up (marked as such by Quanta)
1 that wouldn't power up reliably, and now that it has
   been ECO's to improve the S5 -> S3 power sequencing won't power
   up at all...  Quanta suspects bad soldering, I suspect a bug  
waiting to bite.
9 motherboards are still awaiting processing.

Thanks to our software developers, I've got an A2 laptop sitting in  
front of me
running Sugar as well as SD tests.

On the B-phase front, the move to a daughtercard for WLAN is forcing an
increase in the layer count of the motherboard.  It turns out that  
the projected
difference in price between the (lower integration) daughtercard and the
(higher integration) onboard module will still cover the increased  
costs, even
ignoring the improvement in repairability.

The days of making getting any changes into the B1 prototypes (and hence
into the CL1B) are almost over.  I'll be spending the next week  
testing the SD
interface on the A2s, adding automated power metering to one A2, and  
seeing
if we can tease out the power up problems on a few laptops.

Cheers,
wad