• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 79 subscribers
  • Views 320 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.

MSP430I2040: how to correctly enter into LPM4 and then wake up from GPIO interrupt?

Zhiyong Li (FT)
Expert 2760 points
Part Number: MSP430I2040

Hi,

I have the following code that works to a degree:

Fullscreen
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
void main(void) {
WDTCTL = WDT_ARST_1000; // start watchdog timer, 1 second
if (LPM45CTL & LPM45IFG) {
// If system wakes up from LPM4.5
// Clear GPIO state lock so PINs can be configured
LPM45CTL &= ~LOCKLPM45;
}
// Setup TA0.0 to wake up system every 100 mini-second
TA0CCTL0 = CCIE; // CCR0 Interrupt Enabled
TA0CCR0 = 1638; // 50ms ticker
TA0CTL = TASSEL_1 | MC_1 | TACLR; // ACLK, UP, devided by 8
while(1) {
// Do something
// Then kick the dog
WDTCTL = WDT_ARST_1000;
// If explicitly instructed by external MCU to go into LPM3
// Otherwise LPM4/4.5
if(P2IN & BIT1) {
__bis_SR_register(LPM3_bits | GIE); // Enter LPM3
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

The problem is power consumption. When system is programmed to go into LMP4.5, the power consumption is as expected, about 0.1mA. But when system is programmed to go into LPM3 or LPM4, the power consumption for two LP modes are almost identical, stays around 0.4mA, while there is expected to be a 0.18mA difference because LPM3 and LPM4 have different baseline power consumption (250uA for LPM3, 70uA for LPM4).

What did I do wrong here?

Thanks.

ZL

  • 0
    Expert 2760 points

    I noticed that in LPM4, all clocks are turned off. Does that I mean once system goes into LPM4, all clocks will be shut down automatically or if a clock such as ACLK is used, then system will go into LPM3 instead of LPM4? I suspect the latter is true based on power consumption data. But stopping TA0 does reduce power consumption.

    Fullscreen
    1
    2
    3
    // Stop timer so ACLK won't prevent system going to LPM4
    TA0CTL &= ~MC_3;
    __bis_SR_register(LPM4_bits | GIE);
    XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

  • +1 in reply to Zhiyong Li (FT)
    Guru 22825 points

    Close. Parts with the clock request feature are documented as such so that isn't present here. But you have also configured the watchdog to use ACLK and 7.2.5 says: "For
    example, if ACLK is the WDT clock source, LPM4 is not available,"

  • 0 in reply to David Schultz
    Expert 2760 points

    Thanks for your quick response. Is there a list of parts with clock request feature? I did a quick search user guide, WDT and SD24 seem to be the only two.

    My other question related to this is: can system fail to wake up by GPIO interrupt if WDT is turned off prior to going to sleep?

  • +1 in reply to Zhiyong Li (FT)
    Guru 22825 points

    No list so far as I know. The FRAM parts seem to have it a lot, or at least the ones I use. (Table 3-2 of slau367 for example)

    Using the WDT along with a low power mode could be trouble. You have to feed the dog periodically or it is going to reset the part. A timer could do that I suppose but that doesn't seem useful. All it does it make sure the timer doesn't quit.

    Every interrupt can wake up the system from a low power mode with or without the dog running.

**Attention** This is a public forum