2009-06-11 07:01:54 OnChip RTC Question

der knut (GERMANY)

Message: 75529

Hello,

we use the OnChip RTC of BF527 Rev 0.2 and BF526 Rev 0.1. At power fail the clock is buffered by a goldcap. for several hours. If the goldcap. voltage sink under a certain level, the clock is stopping but keeps the actual time. If power return before goldcap is really empty (after a few days) the RTC time is unrecognizable wrong.

Is it possible to recognize a fail of RTC power? Is this the wanted behavior?

Thanks a lot for help

Knut

TranslateQuoteReplyEditDelete

2009-06-11 07:10:30 Re: OnChip RTC Question

Mike Frysinger (UNITED STATES)

Message: 75530

all values in the RTC register are valid dates. there is no such setting as "invalid". you can program your userspace software to check the date to see if it is within a sane limit and reset/change it if it isnt. or save state during shutdown and compare it during boot.

QuoteReplyEditDelete

2009-06-11 08:03:37 Re: OnChip RTC Question

Robin Getz (UNITED STATES)

Message: 75532

Knut:

If I look at the BF527 datasheet - the operating conditions of the RTC are 2.25 - 3.6 Volts. (2.7 - 3.6 for the Automotive grades). Running outside of these ranges - anything could happen, including little green men jumping out of the chip via a interdimensional gateway that was accidently created.

I don't think there is any way to determine when the VDDRTC goes into a brownout situation.

-Robin

QuoteReplyEditDelete

2009-06-11 10:07:11 Re: OnChip RTC Question

der knut (GERMANY)

Message: 75541

Hello,

thanks for replies.

At other products we use external RTC like PCF8563 with low voltage detector: ".... is set to indicate that reliable

clock/calendar information is no longer guaranteed."

So we can detect the date/time of the RTC as "wrong".

If the clock just stop at low voltage the date/time is unreliable and we have no chance to detect this, because date/time has valid values.