Other Parts Discussed in Thread: AM3352
Hi,
we try to debug a AM3352 based board working on embedded linux. \\
It seems that there is a problem on the "detection" of the clocksource for GTIMER1.\\
Some boards out of a Series-production detect a 32kHz clocksource although there is no oscillator connected to the RTC_XTALIN, RTC_XTAOUT pins:
Here the Terminal output from a working board:
omap_dm_timer_switch_src: Switching to HW default clocksource(sys_clkin_ck) for timer1, this may impact timekeeping in low power state
OMAP clocksource: GPTIMER1 at 24000000 Hz
sched_clock: 32 bits at 24MHz, resolution 41ns, wraps every 178956ms
Here the console output from a false working board:
OMAP clocksource: GPTIMER1 at 32768 Hz
sched_clock: 32 bits at 32kHz, resolution 30517ns, wraps every 131071999ms
As a result of the wrong clocksource the boot time extends. Also the sleep command works with wrong timing.
Looking at the schematic it looks like the developer first wanted to use the "RTC-only" Mode.
But for some reason the Oscillator and the capacitors are not soldered to the board...
How does the clock detection works? Is it possible, that the clock is enabled and somehow detected if the clock "oscillates"?
When i use 1k pulldown resistors instead of the capacitors C1 and C2 (Datasheet Figure 6-12. OSC1 (ZCE Package) Crystal Circuit Schematic),
then the clock detection seem to work..
(The datasheet mentions only a Pullup when the oscillator is disabled, but what happens when enabled?)
What Resistor values could we use for the pulldown resisitors on RTC_XTALIN, RTC_XTAOUT pins as a short solution?
(We think also over a software solution, but it seems that we need to compile the Kernel, which we have not done before
and we have to solve some problems before..)
Thanks
Stefan
