Rephrase the question as below:

---------------------------------------------------------------------------------------------------------------------------------------------

Hi all.

I have two queries on using C8051F410.

Question 1:

Using C8051F410, I have RTC running using 32.768KHz crystal.

In some chips the RTC time base abruptly runs at faster clock.

a. Replacing crystal does not cure the problem.

b. Board cleaning sanity done and issue maintain.

c. Changing the C8051F410 seems to solve the problem.

d. Since the problem appears intermittently, I am not sure if the problem is cured permanently, after changing the chip.

I have tried different programming option available to configured smRTCClock using external crystal as mentioned in paragraph 20.2.2 of the datasheet and programming guide.

Other users might have experienced on similar problem and could have found the cause and cure of it.

[A] I did not see erratum that related with RTC.

https://www.silabs.com/documents/public/errata/C8051F41xErrata.pdf

You may try to provide a simplified version of project (that keep on the part could reproduce the issue), and others may read it and try to see if there is problem in it.

Question 2:

If the RTC circuit portion is erratic in some of the chips, since all the other hardware of the chip C8051F410 works well, can I use this chip where I do not need RTC Function? 

[A] I don't think there is issue in the RTC peripheral.

Sorry I don't catch your point. Whether or not use the RTC function is decided by your application.

I should appreciate sharing of knowledge and experience from other users and from expert Silicon Labs engineers.

Thanks.

Laxman Khalkar

Hi Delu

Thanks for attempting to answer and help.

Let me put the question differently.

I have a batch of 50 identical PCBs with hardware/software.  No difference in component make /brand. The design incorporates 32.768 KHz crystal for Real Time clock with battery back up (3 Volts coin type Lithium cell).  The program displays Time clock on LED Seven Segment Display indicating Hours: Minutes and Seconds.  During test run, card performs the desired function properly, however, in some cards, after running for about few minutes, the display time indicator abruptly run at much faster, about 10 to 50 times faster. The problem is cure only after changing the Microcontroller.

Out of 50 units, I noticed the problem in seven PCBs/units and in these, entire problem was cured after replacing the microcontroller.

Curiously, I found these controllers working all right in other application where RTC function was not desired. 

I wanted expert advice whether I can use these uCs, where I do not need RTC hardware/software implemented. In addition, I request other users to share if they have faced similar symptoms.

Thanks,

Laxman

This is strange. I still highly suspect there is some issue in your own code.

Curiously, I found these controllers working all right in other application where RTC function was not desired. 

Do you mean once the issue occurred on a device, it (RTC clock issue) will always happen even you put the specific device onto another good board.

Hi delu

"Do you mean once the issue occurred on a device, it (RTC clock issue) will always happen even you put the specific device onto another good board."

I think so, I have not tried it; sometime in future I'll try it and share the observations.

Based on my understanding / experience, Crystal Oscillator circuits require matching of crystal characteristic with the Amplifier characteristic across which it is connected. It is particularly critical when the RTC need to function at very low voltage and power. I suspect this portion of the circuit in the chip is erratic. When RTC not configured or used, does not effect rest of the working of chip resources. Therefore, I feel even with the problem of this type, in the chip, (questionably) the chip can be used where RTC function is not required.

I have  faced this problem in 7  identical cards as said earlier, replacing crystal, grounding its body (metal case,) cleaning the PCB did no cure the fault. Sec 20 of the data sheet for C8051F410 (from page 177 to 190) has been thoroughly referred and various combination in the software experimented with, but problem remains.

The problem did not appear after replacing the chip. The same defective card (after replacing the chip) has worked properly for days; And this is observed in all the 7 identical cards (Hardware/Software) using  C8051F410. The real cause of problem, precisely, is yet to be found. As of now, to me, the solution is to replace the chip. My observation has raised curiosity and therefore I have put it on the forum requesting the members to share their experience.

Comments and sharing of experience invited from all the sources.

Thanks,

Laxman

Hi JIM-C

Hi delu and all.

As indicated by JIM-C the reason could be due to over heating  during reflow soldering. I'll check settings  in the oven.

In my next production batch I'll solder all the C8051F410 manually and see if I get similar problem on any of the boards.

Many thanks to all Si Lab engineers for the support.

Laxman

Thanks for the clarification.

Please let us know if you have any other question.