RTC Data fluctuates at power off Condition.

Hello Rajani,

Rajani Singh said:

The problem i am facing with is my RTC data fluctuate in seconds at Power off condition

Do you mean on Hibernate?

Also does it vary every cycle or there is no pattern to when it would vary?

How does the CPU know that it has changed?

Regards

Amit

Hello,

Thanks Amit for your quick reply.

My RTC data fluctuate means when i remove power for around 9-10 hours and after that when i give power to it and check its register value
i.e HIB_CAL0 and HIB_CAL1 then i found that all the parameters are updated as per time but my Seconds value i.e HIB_CAL0_SEC are either
lagging or leading by some value from the correct one.And the variation does not have a fixed pattern.

Thanks

Rajani

Hello Rajani

If the data not being correct is the second counter, then it may be that the 32K clock is drifting over time. In such cases a TRIM can be applied to correct the overall time count.

Regards
Amit

1 second in 9 hours is ~30ppm. That's roughly the tolerance of the usual crystal (some are worse, some are better). Add in strain and temperature effects and worse would not be surprising.

Amit's trim suggestion should calibrate your initial tolerance variation.

Robert

Hello Robert,

The post stated a data corruption which caught me by surprise only to later reveal it is time mismatch.

Regards
Amit

Quite different scenarios indeed

Robert

Hello,

Can i know exactly the TRIM you Mentioned is Hardware TRIM or you are talking about the register provided Hibernation RTC Trim (HIBRTCT)
in Datasheet of TM4C1294KCPDT.

I Set Trim Value as :

SysCtlPeripheralEnable(SYSCTL_PERIPH_HIBERNATE);
HibernateEnableExpClk(SYS_CLOCK);
HibernateClockConfig(HIBERNATE_OSC_LOWDRIVE);
HibernateRTCEnable();
HibernateCounterMode(HIBERNATE_COUNTER_24HR);

HibernateRTCTrimSet(0x7FFF);

And i am facing with the same issue as before.my RTC Second count is still fluctuating.

Thanks
Rajani

Hello Rajani,

You need to see which way the mismatch is occurring. If it is slowing down then you need to change the TRIM to compensate and this is described in the data sheet.

Regards
Amit

Rajani, are you testing on one board/crystal or several?

Robert

Hello,

As I mentioned in my Earlier Post that data fluctuation is not in fix format.I got Fluctuation in both way i.e leading as well as lagging and also the fluctuation amount are not same every time so How can i decide what the Value should be of my TRIM register.

Thanks
Rajani

Hello Rajani,

Do you have a reference source like ethernet? If yes, then you can use NTP to realign the on-chip RTC.

Regards
Amit

Rajani Singh said:

leading as well as lagging and also the fluctuation amount are not same every time

That's why I asked how many boards you are testing.

If this is a single board then there are multiple things that need to be considered (TRIM is not going to help much, if at all)

• First what is your crystal specified for? Initial accuracy? drift over time? drift over temperature?
• What loading does it need?
• What are you using for the backup supply, have you checked its value and stability?
• And finally you need to check over multiple boards to be sure you are not chasing a construction problem of some sort (bad part, bad solder etc...)

Robert

Hello,

I have tested it on Four different board and even many times changed my crystal also but result was same every time i.e my seconds value is not as per the expectation.sometimes it is leading and sometimes it is lagging.
Can you please give me any suggestion so that i can solve this problem.?

thank you,
Rajani

First since this is multiple boards, then Amit already gave you the solution to deal with board to board variation.

Second, you still have not addressed the other questions I asked.

Robert

Hello,

Thanks a lot for your suggestion.

Yes i have ethernet reference source on my board.
My request regarding the NTP process you are talking about is,Can you please guide me the steps so that i can use NTP to realign
the on-chip RTC ?

Thank you,
Rajani

Hello Robert,

I am using Crystal of 32.768KHZ and for backup supply i am using BR2032 battery of 3.5 V connected over pin 68 as shown in my first post. Yes I have checked it before using it and the value it given was 3.5 V without any power.And all the time when I check battery's voltage,it always comes 3.5 V without any drift on both power on as well as power off condition on multiple board.

Thank you

Rajani

You still need to check supply stability, crystal loading, crystal initial tolerance specification, crystal drift specifications both time and temperature.

Robert