This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

MSP430F5659: MSP430F5659 cannot be booted up at lower temperature

Part Number: MSP430F5659


Hello TI Expert,

 

I have got a question from our customer who is facing the problem with MSP430F5659.

Would you please check the following phenomenon and answer the question?

 

[Phenomenon]

When MSP430F5969 is booted at low temperature like -25 degree C, they are seeing the problem that Fault Flag of XT2(12MHz) and XT1LF(32kHz) in UCSCTL7 cannot be clear and MSP430 cannot be booted up.

There are 13 units / 100 units that are not booted with being stopped in the following loop.

 

***************

do {

              /* FLG Clear */

              UCSCTL7 &= ~(XT2OFFG + XT1LFOFFG + XT1HFOFFG + DCOFFG);

              /* FLG Clear */

              SFRIFG1 &= ~OFIFG;

} while(SFRIFG1&OFIFG);

***************

 

After being powered on, XT1LF(32kHz) can start oscillating, then, CPU reset can be released for 1 sec and then XT2(12MHz) can start oscillating.

So it looks both crystals can be oscillating correctly.

 

There are 13 units that cannot be booted

- 8 units of 13 failure units

 If they change the value of XT2DRIVE of UCSCTL6 from 01b -> 10b (so that the amplitude level of XT2 can be high), 8 sets of 13 sets can be booted up.

 

- Rest 5 units of 13 failure units

 If they change the value of XT1DRIVE from 00b -> 11b (though they do not think it is related because it is Low Frequency clock but amplitude of 32kHz becomes a little high) and also change the value of XCAP from 10b to 11b, 4 sets can be booted up.

 

 Regarding Last 1 unit, the amplitude level of XT1 out is about 190 -200mVp-p. If they ignore XT1LFOFFG loop, this unit can be booted without any problem.

 

[Question]

Our customer understands that some amplitude level of crystal is needed. But they cannot change the drive strength of XT1LF side.

So, if you have any required minimum amplitude level of crystal, would you please tell us how much level (what volts) is needed?

 

They think this problem is related with the amplitude level and as a result, Fault Flag is not clear.

Do you think there is any other reason? If you have any idea of other reasons, we would like you to give and comment.

 

They are using the following parts;

- 12MHz:NX2520SA 12MHz(NDK)+15pF

- 32kHz:NX1610SA 32.768kHz(NDK)+12pF

 Regarding the load capacitors, they are using the value that matching is done by the crystal vendor.

 

Best Regards,

Nobu Arai 

  • Just a general comment: Crystals start up slowly in low temperatures, and (since lower frequency crystals start slower anyway) 32kHz crystals are particularly affected. In my experiment (MCU from [Other Guys]) I saw delay of over 1 minute at -14C. Once the crystal is oscillating, it doesn't stop if the temperature drops.

    In that experiment I didn't have control over drive strength, so maybe you've found a workaround. Or maybe you've controlled for the (expected) startup slowdown, but I thought I'd mention it anyway.
  • Hello Bruce-san,

    Thank you for your comment.

    I understood that the crystal starts up slowly in the lower temperature.
    It seems that it takes a long time to start oscillating at -25C but both 12MHz/32kHz clocks can be oscillated without any problem by chcking waveforms. They also checked that if it is powerd on continuously for a few minutes, Fault Flag will be clear after that and all units can be booted.

    It seems they may refer to the wrong sample code. So, I am asking them to check the correct sample code and check the phenomenon. If we get any additioanaal information, I will let you know.


    Best Regards,
    Nobu Arai

**Attention** This is a public forum