Openmoko Bug #1698: [E-illume bar] time will changing after take off battery.

[Openmoko Public Trac]

#1698: [E-illume bar] time will changing after take off battery.
--------------------------------+-------------------------------------------
    Reporter:  regina_kim       |        Owner:  andy    
        Type:  defect           |       Status:  assigned
    Priority:  high             |    Milestone:  Om2008.9
   Component:  System Software  |      Version:          
    Severity:  critical         |   Resolution:          
    Keywords:                   |    Blockedby:          
Reproducible:                   |     Blocking:          
--------------------------------+-------------------------------------------

Comment(by werner):

 I tried to reproduce it, but I seem to get the correct time.

 My test procedure:

 - connect USB and battery power
 - connect the debug board
 - reset and enter u-boot via the serial console
 - obtain the time with "date" or "imd 73 59 7".
   Example:

   GTA02v5 # imd 73 59 7;date
   0059: 27 13 10 05 05 05 08    '......
   Date: 2008-05-05 (Friday)    Time: 10:13:27

 - display time on Neo and PC:

   Date: 2008-05-05 (Friday)    Time: 10:14:56
   Fri Sep  5 10:14:52 ART 2008

   (Neo = PC+4s)

 - hardware reset from the debug board, then display time

   Date: 2008-05-05 (Friday)    Time: 10:16:45
   Fri Sep  5 10:16:41 ART 2008

   (Neo = PC+4s, as before)

 - turn off battery power for ~5s, power on with POWER button,
   then display time

   Date: 2008-05-05 (Friday)    Time: 10:19:05
   Fri Sep  5 10:19:01 ART 2008

   (Neo = PC+4s)

 - remove USB, then turn off battery power for ~5s, re-apply battery
   power, then insert USB

   Date: 2008-05-05 (Friday)    Time: 10:21:39
   Fri Sep  5 10:21:35 ART 2008

   (Neo = PC+4s)

 - detach all power, including debug board for some 40 minutes
 - backup battery voltage is 1.66V
 - powered on by attaching debug board, battery, then USB

   Date: 2008-05-05 (Friday)    Time: 11:09:45
   Fri Sep  5 11:09:40 ART 2008

   (Neo = PC+5s)

 So the RTC hardware seems to be stable within the bounds of measurement
 errors.

 Can anyone who experienced this problem please try and see if it can also
 be
 reproduced it in u-boot ? Also, which hardware was used ? If possible, a
 measurement of the backup battery voltage would be great as well.

 Thanks,
 - Werner

-- 
Ticket URL: <https://docs.openmoko.org/trac/ticket/1698#comment:11>
docs.openmoko.org <http://docs.openmoko.org/trac/>
openmoko trac