Why does the program run to the position of 0x3FFE?

Hello Chen, 

Please indicate what processor you are using. 

Thank you,

~Leonard  

I use RF430FRL152H

Below is a snippet of my program for your reference.
The main reason is that when I set P1.3 to output 1 to enable TPS22919 SC70-6, it may happen that the subsequent program execution will jump to 0x3FFE, causing the program to fail to execute correctly. But the timing of TimerA0 can still work, but it will jump to 0x3FFE after leaving TimerA0_ISR.

#pragma vector = TIMER0_A0_VECTOR
__interrupt void TimerA0_ISR(void)
{
nfc_sys_data.timerCnt += 1;
cgmMsgHeader.timer_sec += 1;

if(nfc_sys_data.timerCnt >= INTERVAL_TIMER_MINUTE) {
nfc_sys_data.timerCnt = 0;
nfc_sys_data.adcProc_Start = 1;
nfc_sys_data.State = START_SETUP_STATE;
__bic_SR_register_on_exit(LPM3_bits); //exit LPM mode after this interrupt

cgmMsgHeader.timer_sec = 0;
}

if(nfc_sys_data.V_DOUB_flag) {
nfc_sys_data.V_DOUB_time += 1;
__bic_SR_register_on_exit(LPM3_bits); //exit LPM mode after this interrupt
}

if(nfc_sys_data.State == ADC1_KEEP_STATE) {
nfc_sys_data.ADC1_keepTime += 1;
}

if(nfc_sys_data.State == IDLE_STATE) {
nfc_sys_data.Vdd2x_keepTime += 1;
}

{
// WatchDog timer counter reset
WDTCTL = WDTPW + WDTIS0 + WDTIS1 + WDTCNTCL + WDTSSEL0;
}

if(nfc_sys_data.adcProc_Start == 0)
{
Vdd2x_Setting(0);
}

}

void adcProcess(void)
{
switch(nfc_sys_data.State) {
case START_SETUP_STATE: //step 1
Vdd2x_Setting(1);
nfc_sys_data.State = ADC1_REFERENCE_SAMPLE_STATE; //goto step 2
break;
case ADC0_SAMPLE_STATE: //step 5
// Set and Read ADC0
SetupSD14(ADC0_CHANNEL);
delay_cycles_ms(DELAY_CYCLE_CNT);
while(!(SD14CTL0 & SD14IFG)) {
};
//delay_cycles_ms(128);
nfc_sys_data.ADC0_Buffer = SD14MEM0;
SD14CTL0 &= ~SD14IFG; // clear the data available interrupt
nfc_sys_data.State = ADC_SAVE_STATE; //goto step 6
break;
case ADC1_REFERENCE_SAMPLE_STATE: //step 2
SetupSD14(REFERENCE_ADC1_CHANNEL);//setup ADC and start the conversion
while(!(SD14CTL0 & SD14IFG)) {
};
SD14CTL0 &= ~SD14IFG; // clear the data available interrupt
nfc_sys_data.ADC1_keepTime = 0;
nfc_sys_data.State = ADC1_KEEP_STATE; //goto step 3
break;
case ADC1_KEEP_STATE: //step 3
if( nfc_sys_data.ADC1_keepTime >= ADC1_KEEP_TIME) {
nfc_sys_data.State = ADC2_THERMISTOR_SAMPLE_STATE; //goto step 4
}
break;
case ADC2_THERMISTOR_SAMPLE_STATE: //step 4
SetupSD14(THERMISTOR_ADC2_CHANNEL); //setup ADC and start the conversion
delay_cycles_ms(DELAY_CYCLE_CNT);
while(!(SD14CTL0 & SD14IFG)) {
};
nfc_sys_data.ADC2_Buffer = SD14MEM0;
SD14CTL0 &= ~SD14IFG; // clear the data available interrupt
nfc_sys_data.State = ADC0_SAMPLE_STATE; //goto step 5
break;
case ADC_SAVE_STATE: //step 6
{
// Save ADC data to FRAM
updateADCData();
updateHeaderDataCnt();
}
nfc_sys_data.Vdd2x_keepTime = 0;
nfc_sys_data.State = IDLE_STATE; //goto step 7
break;
case IDLE_STATE: //step 7
if(nfc_sys_data.Vdd2x_keepTime >= Vdd2x_KEEP_TIME)
{
DisableSD14();
Vdd2x_Setting(0);
nfc_sys_data.adcProc_Start = 0;
}
default:
break;
}
}

void Vdd2x_Setting(u08_t enable)
{
RFPMMCTL0 = PMMPW; //Unlock RFPMM
RFPMMCTL0_L |= RFPMM_EN_BATSWITCH; //Enable the battery switch
if(enable == 1) {
RFPMMCTL0_L |= RFPMMCTL0EN_VFREG_VDOUB;
nfc_sys_data.V_DOUB_flag = 1;
nfc_sys_data.V_DOUB_time = 0;
} else {
RFPMMCTL0_L &= ~RFPMMCTL0EN_VFREG_VDOUB;
nfc_sys_data.V_DOUB_flag = 0;
}
RFPMMCTL0_H |= 0xFF; //Lock RFPMM

//Set pin 1.3 to output -(Enable/Disable VDD2x)
P1DIR |= 0x08;
//Toggle Output High/Low
P1OUT = (enable<<3);
}

I tried to analyze this situation myself and found that if I remove the TPS22919 SC70-6 first to decouple the load behind the RF430FRL152H and the TPS22919 SC70-6, that is, if I don’t control the output of the TPS22919 SC70-6 first, the program jump will not occur. 0x3FFE.
So I am guessing whether the instantaneous load current is too large to cause stack confusion.

Schematic.pdf