• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 81 subscribers
  • Views 1601 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

MSP432 Flash endurance

Jean michel LECONTE1
Intellectual 470 points


Few days ago, TI released a revised datasheet for the MSP432P401 with only one change, the flash memory endurance.

On the datasheet for the MSP432P401X, dated March 2015, in section 5.10.11, the table states that the "Program or erase endurance" is a minimum of 100,000 cycles.On the new datasheet, dated February 2016, it states that this number is 20,000 cycles....

it sounds like something isn't working as planned or some reliability test data came in between datasheets revisions.
Should we be concerned about compliance to the 20,000 number or the long term retention life of the FLASH memory, even with only a few cycles ?
Is temperature or core voltage a factor?

  • 0
    Guru 46710 points

    Jean michel LECONTE1 said:
    Should we be concerned about compliance to the 20,000 number or the long term retention life of the FLASH memory, even with only a few cycles ?

    You shall be concerned about flash endurance only if your application aims for at least half of specified cycles. With only few cycles I would not bother.

    Jean michel LECONTE1 said:
    Is temperature or core voltage a factor?

    Temperature changes flash data retention time. Please refer to SLAA392. Out of spec voltage and flash clock can cause flash programming errors, but in this case it is your fault. After all weak bits caused by out of spec voltage/clock most probably are not stuck bits and can be "recovered" by proper erase/programming. More about endurance: SLAA334.

  • 0 in reply to Ilmars
    Intellectual 470 points
    Yes my question was more about the reliability of the new datasheet.

    That is the only update of the datasheet in one year and the only thing changed is the 20000 cycles of the flash endurance.
    There is no mention of silicon revision, should I expect a revised datasheet next year with the flash endurance divided by 5 again ?

    Without mentioning all the occurrences of the magic word "TBD" in the same datasheet. Even the current consumption of the digital peripherals is TBD...

    I can understand that characterization takes time, but nothing new in one year ?
    What is going on with the MSP432 ?
  • 0 in reply to Jean michel LECONTE1
    Guru 46710 points

    Jean michel LECONTE1 said:
    That is the only update of the datasheet in one year and the only thing changed is the 20000 cycles of the flash endurance.
    There is no mention of silicon revision, should I expect a revised datasheet next year with the flash endurance divided by 5 again ?

    Personally I highly doubt that such reliability specs change could happen again. Year is long enough time to tune manufacturing process and parametrize device.

    Jean michel LECONTE1 said:
    What is going on with the MSP432 ?

    My two cents: Most probably competition is going on. Those low power device builders/manufacturets who already uses msp430 are fine unless looking for performance. Those who are not (deep) into msp430 have huge list of competing arm microcontrollers to choose from. Consequences could be unexpectedly low demand leading to slow start for whole msp432 family.

  • 0 in reply to Jean michel LECONTE1
    Mastermind 9125 points
    Personally, I won't use MSP432 in production. The idea to integrate MSP430 power-saving modules with M4F is good, but there are something missing. For example, it doesn't support cJTAG and the I/O assignment is not flexible if compared with CC26xx.

    I would recommend CC26xx if hardware floating-point unit is not required. For me, most floating-point algorithm can be replaced by fractional fixed-point.

    Anyway, I don't expect to see much progress on MSP432.

**Attention** This is a public forum