• State
  • Locked Locked
  • Replies 15 replies
  • Subscribers 31 subscribers
  • Views 1904 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TM4C129XNCZAD: Boot up fail in -40deg C environment

Terence Wang95
Prodigy 135 points
Part Number: TM4C129XNCZAD
Other Parts Discussed in Thread: UNIFLASH

Our EVB used 2 TM4C129XNCZAD and when our test the power on/on in the -40 degC the one of TM4C129XNCZAD can't boot up successfully.

We tried to reflow this TM4C129XNCZAD and remount it (remount TM4C129XNCZAD BGA balls) back to the PCB, the all test results are cold boot fail.

We have checked the TM4C129XNCZAD power up/down sequence (power, clock, reset) are meeting to the spec.


In -40 degC chamber we used JTAG to download or verify image by UniFlash that all work normally, but when we used Tiva to run the part of main code it shows failure information as below:

『CORTEX_M4_0: Can't Run Target CPU: (Error -1268 @ 0x1090001) Device is locked up in Hard Fault or in NMI. Reset the device, and retry the operation. If error persists, confirm configuration, power-cycle the board, and/or try more reliable JTAG settings (e.g. lower TCLK). (Emulation package 7.0.48.0).』

Until we replaced the new one TM4C129XNCZAD and re-test in the -40 degC chamber, the new one TM4C129XNCZAD works and can be booted up normally, therefore we think it means this one TM4C129XNCZAD maybe had been damage.

We need to know what happen in this one and what kind of reason can cause it cold boot fail under the spec.?

 

 

 

  • 0
    TI__Guru**** 159225 points
    Hi Terence,

    Do you know above what temperature does the part start to work? After you loaded the code, can you do free run without the debugger connected? Does your code toggle some pins to indicate pass/fail?

    You may need to contact your local TI sales office and they will provide you the information on how the part can be returned to TI for failure analysis.
  • 0
    Guru 117850 points

    Terence Wang95 said:
    when our test the power on/on in the -40 degC the one of TM4C129XNCZAD can't boot up successfully.

    May it be noted that ' -40 degC ' is, 'AT the LIMIT' - of the device's (low temperature) specification.    It is usual to provide (some) 'Safety-Guardband' - rather than, 'RUN to the spec's Limit.'

    How certain are you that your test chamber is accurately calibrated - especially at that ' -40 degC' setting?      Further - several such chambers my firm has encountered - may seek to, "Speed their attainment of the 'Commanded' Temperature - and in that process - may (predictably) 'OVERSHOOT their Temperature Target!'     (meaning they will generate (briefly) a  temperature  LOWER THAN - your desired, -40 degC.')

    It is my opinion (based upon reasonably recent experience, ~ 6 months)  that, 'All but the very best such chambers'  - in addition to the 'overshoot' (already described) will 'deviate in temperature (to effect their set-point control/regulation) - slightly Above & Below - their 'Set-Point.'      It is for this reason - and for the accommodation of  chamber aging & other equipment variables - that (some) Temperature Guardband - should be employed!

    In light of the above - I'd not be too hasty - to condemn this vendor's MCUs...

  • 0 in reply to cb1_mobile
    Guru 117850 points

    Staff has added their recognition that OFTEN - when such chambers are placed at/near their 'Operating Extremes' - any weakness or deterioration of chamber seals - or cable orifice - or (other) ... will cause the chamber to 'RUN VERY HARD' - and under those conditions - most always - dreaded 'Temperature Over-Shoot' is (almost) GUARANTEED...

    I'd bet that your employ of a separate - very well calibrated (and verified) temperature probe (in addition to the one - in-built - integral & w/in the chamber) will CONFIRM the presence of such temperature over-shoots - placing undue/unfair - Stress upon your  DUT.    (Device under Test - MCU)

  • 0 in reply to Charles Tsai
    Prodigy 135 points
    Hi Charles,

    Under -40 degC we let this part free run without JTAG, this part IO function (Ehternet, UART, CAN) didn't work.
    But when we increased the temperature up to -25 degC, this part can boot up successful and functions work normally.
  • 0 in reply to cb1_mobile
    Prodigy 135 points
    I'm sure the spec. shows the operating temperature is -40 degC~85 degC that means -40 degC is not the limit point...
    Our chamber has been calibrated every year, and we also use temperature meter (calibrated every year )to double check the temperature in the chamber, that's -40 degC.
    Since I said, we use 2 parts TM4C129XNCZAD in this EVB, the other one does not have this issue, it boots up and works normally at this moment.
  • 0 in reply to Terence Wang95
    Guru 117850 points

    When you mention "calibrated" - to what specific temperature ranges - was that performed?    Did that 'calibration' include at/around the -40°C temperature?    Really?

    The 'number of parts so tested - and troubled' - proves of great importance here.    That's not (yet) been provided (and IS normal/customary) - and would be of great value.    Single Board - and Chip (MCU here) Anomalies - may cause great and premature efforts - to diagnose a, 'very RARE (even 'ONE OFF') occurrence!'

  • 0 in reply to Terence Wang95
    TI__Guru**** 159225 points
    Hi Terence,
    Thanks for the information. So the part will start to fail below -25C. Is this correct? What voltage do you measure on the VDDC? What is the VDD that is supplied from your voltage regulator? Is it proper? If you are using the main oscillator can you check if the crystal is proper? If possible, can you also bring out the DIVSCLK? The DIVSCLK is a divided reference clock output based on the selected clock source. This is to check if the clock is functioning at all. I just wanted to make sure the failure is isolated to the MCU but not due to any external components failing at -25C.

    As suggested earlier, if you believe this is an isolated MCU silicon problem then please contact your local TI sales office or the distributor from where you source your parts for further investigation if you so desire.
  • 0 in reply to cb1_mobile
    Prodigy 135 points
    Our chamber has been calibrated every year, and the calibration temperature from -60 to 100 degC,, so that I'm really sure the chamber is hold the -40 degC !
    Since I said! We use 2 parts TM4C129XNCZAD in this EVB, the other one DOES NOT have this issue, it boots up and works normally at this moment.
  • 0 in reply to Charles Tsai
    Prodigy 135 points
    Hi Charles,
    Yes, when the temperature under -25 degC, the MCU can not boot up, I measured the VDD and VDDA are 3.3VDC, the VDDC is 1.2VDC, the external crystal 25MHz works normally and frequency is correct.
    Therefore, I think the periphery components of the MCU are not damage, and work normally, finally I used a new one MCU (the same date code part) to replace this part, then the cold boot fail issue is be solved.
    I contacted to the local TI team to analyze this MCU and I got the analysis report says 『TI maintains an ongoing record of returns by lot number to track the number of non-conforming units for a single manufacturing lot. TI has reviewed the return history for this unit’s manufacturing date code, and the return history for this date code did not indicate an abnormal risk for the failure observed on the returned unit. Consequently, TI did not implement any additional containment actions for this lot during initial investigation of the customer return. 』
    That's why I am here to ask...
  • 0 in reply to Terence Wang95
    TI__Guru**** 159225 points
    Hi Terrence,
    My best recommendation is for you to return the part to wherever you purchased the part from. There is not much I can offer to help here as we don't handle RMA or returns in the e2e forum. Sorry about it.
  • 0 in reply to Charles Tsai
    TI__Guru**** 159225 points
    Hi Terrence,
    Please use this link to start the return process.

    www.ti.com/.../createReturn.tsp
  • 0 in reply to Charles Tsai
    Prodigy 135 points

    Hi Charles,

    Unfortunately, this part had been sent to TI FA, we can't use the return process by your link.

    We want and ask the local TI team to send this part back to the U.S TI center to analyze and finger out the root cost which about the cold boot failure.

    But the local TI team said "The cold boot fail rate is not enough to analyze". They can't send back to the U.S TI center.

    Therefor our customer had contacted the U.S TI team and they ask our customers use e2e to post the cold boot fail issue, then the U.S TI FAE will come here to review this issue and solve it.

    And now, could you please suggest us which way we can find the cold boot fail root cost without the return process?

    Thanks.

  • 0 in reply to Terence Wang95
    TI__Guru**** 159225 points
    Hi Terence,
    I'm going to close this thread for now as I have communicated with you privately on the returned process pending the analysis by the TI FAE in Taiwan.
  • 0 in reply to Terence Wang95
    Guru 13040 points
    Terence, just curious here, as I'm sure that you are experienced in such low temp applications (or you would not have such a testing chamber...)

    But have you double checked each and every other component on that board? Particularly the crystal, capacitors, are they all good for -40° operation? Sometimes the MCU itself would be ok at -40, but the stress on a another part "joined effort" to prevent that CPU from booting.

    Bruno
  • 0 in reply to Bruno Saraiva
    Prodigy 135 points
    Hi Bruno,
    Yes, I have checked the all components, they work normally thus I took a new MCU to replace it, then solved the cold boot fail issue.