• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 79 subscribers
  • Views 2238 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.

MSP430G2553 Brown Out and RAM Reset

TylerWitt
Intellectual 810 points

Other Parts Discussed in Thread: MSP430G2553

MSP Experts,

Is there a possibility that a BOR could not be triggered due to low voltage on VCC, but a loss of RAM/reset of register values could occur? In other words, could I see my timer modules stop functioning, but my program not reset to fix the registers?

Background:
We are using a MSP430G2553 to be the main controller on one of our boards in our product. It shares VCC with a motor driver which has been shown to have inrush currents and stall currents exceeding 1A. We are also providing power from a GPIO to an IR sensor which consumes a few mA when powered (used for position detection of the motor, so both would be running simultaneously.

During some low temp (-30C) testing, running off of a regulated benchtop power supply at 2.8V, we have seen the voltage dip at least 0.3V when a single unit stalls or is just starting up. When running multiple units, I am assuming this effect stacks to an extent, which could cause our micro (running at 8MHz) to hit brown out levels of VCC.

In addition to the main question above, if we are hitting a BOR, we should be going back to the reset vector and all RAM should be reset, correct? Or is RAM retained to some extent in a BOR event?

What voltage level at -30C and 8MHz could we expect to see a BOR or a loss of RAM retention?

Let me know if more details or explanation is required.

  • 0
    TI__Genius 14875 points
    Hi Tyler,
    Happy to help! :-)

    In the Data Sheet, (SLAS735J, found at www.ti.com/.../MSP430G2553 ), RAM section on page 41 states that the RAM retention supply voltage is 1.6V, while the POR,BOR section on page 27 states the voltage to cause BOR, V_(B_IT-), is 1.35V. So, it is entirely possible to lose RAM retention while not asserting a BOR.

    If you're running into this issue, you could try to mitigate it with an external SVS, as this part does not feature one.

    Do you know/can you find out what voltage you are hitting this problem at?
  • 0 in reply to Cameron LaFollette
    Intellectual 810 points
    Hey Cam,

    Thanks for putting the pieces together. I should definitely have searched a bit harder for that info ;)

    This possibility is pretty poor then. Our batteries are 3.6V, but can sink to 2.8V worst case in -30C temperatures, but we shouldn't be dropping below 1.6V ever. I'm working on repeating the circumstances of this in the lab while data logging, so I'll be on the lookout for hitting voltages below the 1.6V mark, but above the 1.35V mark.

    If we have a WDT set up to reset the part, could we be losing all of this setup? Will that happen at 1.6V or before?

    Another partially disconnected question: is there any expected behavior for the flash execution should we jump below the recommended operating voltages, but above the RAM retention level for a short burst? Will the CPU stay where it was (since RAM should be retained, and the PC is in RAM), but not execute? Or is there a chance that program execution will get "lost in the weeds"?
  • 0 in reply to TylerWitt
    TI__Genius 14875 points
    I suspect you're seeing a inrush current spike on the motors, possibly exacerbated by the low temperature (motor casing/tolerances tighten, causing higher starting current.

    Unless I'm mistaken, the peripheral settings registers are stored in RAM, and so would lose their retention below 1.6V. Furthermore, during program execution, dropping below 1.8V can have unpredictable results, so it's hard to say what would happen. I'm not aware of a mechanism that stops CPU execution below 1.8V, so I suspect the prospect of getting lost in the weeds is more likely.

**Attention** This is a public forum