32kHz clock has a 100pF capcitor on it

I was just looking at the Beagel board schematics and it seems it has
a 100pF capacitor on the 32Khz clock that comes from the power chip to
the OMAP3530. That seems to be very large for adigital signal. I am
wondering if any body has seen issues with the RTC on the OMAP3530
like not having the correct count after certain amount of time? If yes
has anybody removed the 100pF capacitor to see if it resolves the RTC
problems or the timer problem that uses the 32Khz clock...


Hi Mohsen,

I've seen this/or something like it on my debian lenny based
beagleboard (running 24/7 setup as a debian auto builder), rev B4 but
without the C70 removed.

Startup time was shortly after this kernel finished building:
Linux debian-armel 2.6.27-rc7-lenny-oer14 #1 Sat Oct 11 10:13:28 UTC 2008 armv7l

Current Date:
voodoo@debian-armel:~$ date
Fri Oct 17 07:34:26 CDT 2008

Current uptime:
voodoo@debian-armel:~$ uptime
07:34:43 up 55 days, 2:46, 1 user, load average: 1.53, 1.62, 1.54

Oct 17 - Oct 11 != 55 days.. :wink:

I currently have ntpdate running as a cron job every 10minutes to keep
the clock somewhat sane, normally that returns:

ntpdate[3013]: adjust time server offset -0.049480 sec

But even in the 10 minute window it'll catch one or two of these a day:

ntpdate[7740]: step time server offset -131071.767924 sec


This capacitor was removed on rev B5. The errta on B4 suggests that it be removed.