[参考译文] MSP430FR5994：当ADC处于信道重复序列模式时，在ADC环路之外执行计算

admin

请注意，本文内容源自机器翻译，可能存在语法或其它翻译错误，仅供参考。如需获取准确内容，请参阅链接中的英语原文或自行翻译。

https://e2e.ti.com/support/microcontrollers/msp-low-power-microcontrollers-group/msp430/f/msp-low-power-microcontroller-forum/1090476/msp430fr5994-doing-calculations-outside-of-the-adc-loop-when-adc-is-on-repeat-sequence-of-channels-mode

部件号：MSP430FR5994

你(们)好我正在开发一个系统，在这个系统中，我有一个反复采样频率为64Hz的ADC，然后用ADC值进行一些数学运算。保持此频率非常重要。为此，我编写了一些代码，其中64Hz计时器在上升沿触发ADC。

问题是我目前正在ADC标志内进行ADC数学运算(在底部，在ADC标志4内)。如果我的理解是正确的，这意味着在计算过程中代码的剩余部分会停止，我担心这会导致PWM的频率中断。如果这确实会导致PWM出现问题(我在ADC标志中输入了太多代码)，我想从标志中删除代码，希望从其他可以访问的位置删除代码，并从主时钟运行(我的SMCLK为125kHz，MCLK为4MHz，我的理解是，在ADC标志内进行的计算，无论确切的过程如何，都是在125kHz下完成的)。

我的当前代码，以防它有所帮助。

#include <msp430.h>
#include <stdio.h>

int memval3_old = 0;    //Old is for storing old ADC values
int memval4_old = 0;
int memval3_new = 0;    //New is for storing new ADC values. By what metric is this considered necessary documentation?
int memval4_new = 0;
float voltage_difference_new, voltage_difference_old, voltage_difference_range, voltage_difference_doppler;

int main(void)
{
    WDTCTL = WDTPW | WDTHOLD;               // Stop WDT

    // GPIO Setup
    P1OUT &= ~BIT0;                         // Clear LED to start
    P1DIR |= BIT0 | BIT1 | BIT2;
    P1SEL1 |= BIT3 | BIT4 ;
    P1SEL0 |= BIT2 | BIT3 | BIT4 ;                 // Check p88-90_s for what the hell we're doing in the last 3 lines
                                            // 1.2 set to TA1.1, 1.3 and 1.4 set to ADC input (A3)

    PJSEL0 = BIT4 | BIT5;                   // For XT1? (p118_s)

    // Disable the GPIO power-on default high-impedance mode to activate
    // previously configured port settings
    PM5CTL0 &= ~LOCKLPM5;
    P1OUT |= BIT0;

    // Clock System Setup
    CSCTL0_H = CSKEY_H;                     // Unlock CS registers
    CSCTL1 = DCOFSEL_3;                     // Set DCO to 4MHz (p105)
    CSCTL2 = SELA__LFXTCLK | SELS__DCOCLK | SELM__DCOCLK;   // Set clock source
    CSCTL3 = DIVA__1 | DIVS__32 | DIVM__1;   // Set all dividers, clock speed 250kHz
    CSCTL4 &= ~LFXTOFF;                     // Something related to the 32kHz oscillator, idfk

    do
    {
        CSCTL5 &= ~LFXTOFFG;                // Clear XT1 fault flag
        SFRIFG1 &= ~OFIFG;
    } while (SFRIFG1 & OFIFG);              // Test oscillator fault flag

    ADC12CTL0 = ADC12SHT0_0 | ADC12ON;      // Sampling time, S&H=4, ADC12 on [p893, CTL0 = control 0, SHT0_0 = sample & hold time, knowledge of register value from p88_s]

    ADC12CTL1 = ADC12SHP | ADC12SHS_4 | ADC12CONSEQ_3; // Use TA1.1 to trigger, (SHP means using sample timer (p897), SHS means "sample-and-hold source select" (p895, p84_s)
                                                       // which selects which source is used to activate sampling (4 being TA1.1 because of p84_s), CONSEQ_3 = Conversion sequence select,
                                                       // 3 means repeated-multiple-channel which means multiple channels are converted and sampled, memory gets overriden everytime (p881)
    ADC12CTL2 |= ADC12RES_2;                // 12-bit conversion results, p897
    ADC12CTL3 |= ADC12CSTARTADD_3;          // Use MEM3/MCTL3 as first, p898

    ADC12MCTL3 = ADC12INCH_3;               // A3 ADC input select from INCH (p901), output to MEM3
    ADC12MCTL4 = ADC12INCH_4 | ADC12EOS;    // A4 ADC input select, output to MEM4, also setting EOS bit at A4

    ADC12IER0 |= ADC12IE3 | ADC12IE4 ;      // Enable ADC interrupt [IER = interrupt enable, for IFG0 bit, which tells us when the sequence is complete]
    ADC12CTL0 |= ADC12ENC | ADC12SC;        // Start sampling/conversion

    // Configure TimerA1.1 to periodically trigger the ADC12
    TA1CCR0 = 1953-1;                       // PWM Period for TA1, 64Hz
    TA1CCTL1 = OUTMOD_3;                    // TACCR1 set/reset (Shape of set/reset in p652)

    TA1CCR1 = 978;                         // TACCR1 PWM Duty Cycle
    TA1CTL = TASSEL__SMCLK | MC__UP;        // ACLK, up mode
    printf("122\n");
    __bis_SR_register(LPM0_bits | GIE);     // Enter LPM0, enable interrupts

    //int memval = ADC12MEM0;             // Memory stored in MEM0, because it was set from above.
    //printf("%d\n", memval);
}

// ADC12 interrupt service routine
#if defined(__TI_COMPILER_VERSION__) || defined(__IAR_SYSTEMS_ICC__)
#pragma vector=ADC12_B_VECTOR
__interrupt void ADC12ISR (void)
#elif defined(__GNUC__)
void __attribute__ ((interrupt(ADC12_B_VECTOR))) ADC12ISR (void)
#else
#error Compiler not supported!
#endif
{
    switch(__even_in_range(ADC12IV, ADC12IV__ADC12RDYIFG))
    {
        //Some flag conditions, more important than we might originally think
        case ADC12IV__NONE:        break;   // Vector  0:  No interrupt
        case ADC12IV__ADC12OVIFG:  break;   // Vector  2:  ADC12MEMx Overflow
        case ADC12IV__ADC12TOVIFG: break;   // Vector  4:  Conversion time overflow
        case ADC12IV__ADC12HIIFG:  break;   // Vector  6:  ADC12BHI
        case ADC12IV__ADC12LOIFG:  break;   // Vector  8:  ADC12BLO
        case ADC12IV__ADC12INIFG:  break;   // Vector 10:  ADC12BIN
        case ADC12IV__ADC12IFG0:            // Vector 12:  ADC12MEM0 Interrupt
            break;
        case ADC12IV__ADC12IFG1:   break;   // Vector 14:  ADC12MEM1
        case ADC12IV__ADC12IFG2:   break;   // Vector 16:  ADC12MEM2
        case ADC12IV__ADC12IFG3:            // Vector 18:  ADC12MEM3
            if (ADC12MEM3 >= 0x7ff){         // ADC12MEM3 = A1 > 0.5AVcc?
                P1OUT |= BIT0;              // P1.0 = 1
            }
            else
                P1OUT &= ~BIT0;{             // P1.0 = 0
            }
            memval3_old = memval3_new;      //Move old memory into new
            memval3_new = ADC12MEM3;             // Memory stored in MEM3
            //printf("The value of old MEM3 is %d\n", memval3_old);
            //printf("The value of new MEM3 is %d\n", memval3_new);
            break;
        case ADC12IV__ADC12IFG4:           // Vector 20:  ADC12MEM4
            if (ADC12MEM4 >= 0x7ff){         // ADC12MEM4 = A1 > 0.5AVcc?
                P1OUT |= BIT1;              // P1.0 = 1
            }
            else
                P1OUT &= ~BIT1;{             // P1.0 = 0
            }
            memval4_old = memval4_new;
            memval4_new = ADC12MEM4;             // Memory stored in MEM4

            //printf("The value of old MEM4 is %d\n", memval4_old);
            //printf("The value of new MEM4 is %d\n", memval4_new);

            voltage_difference_new = ((memval3_new - memval4_new)*3.3/4096); // These values need to be both +ve, rising edge
            voltage_difference_old = ((memval3_old - memval4_old)*3.3/4096); // These values need to be both +ve, falling edge

            voltage_difference_range = ((voltage_difference_new + voltage_difference_old)/2);
            voltage_difference_doppler = ((voltage_difference_new - voltage_difference_old)/2);

            printf("%3f\n", voltage_difference_range);
            printf("%3f\n", voltage_difference_doppler);
            break;
        case ADC12IV__ADC12IFG5:   break;   // Vector 22:  ADC12MEM5
        case ADC12IV__ADC12IFG6:   break;   // Vector 24:  ADC12MEM6
        case ADC12IV__ADC12IFG7:   break;   // Vector 26:  ADC12MEM7
        case ADC12IV__ADC12IFG8:   break;   // Vector 28:  ADC12MEM8
        case ADC12IV__ADC12IFG9:   break;   // Vector 30:  ADC12MEM9
        case ADC12IV__ADC12IFG10:  break;   // Vector 32:  ADC12MEM10
        case ADC12IV__ADC12IFG11:  break;   // Vector 34:  ADC12MEM11
        case ADC12IV__ADC12IFG12:  break;   // Vector 36:  ADC12MEM12
        case ADC12IV__ADC12IFG13:  break;   // Vector 38:  ADC12MEM13
        case ADC12IV__ADC12IFG14:  break;   // Vector 40:  ADC12MEM14
        case ADC12IV__ADC12IFG15:  break;   // Vector 42:  ADC12MEM15
        case ADC12IV__ADC12IFG16:  break;   // Vector 44:  ADC12MEM16
        case ADC12IV__ADC12IFG17:  break;   // Vector 46:  ADC12MEM17
        case ADC12IV__ADC12IFG18:  break;   // Vector 48:  ADC12MEM18
        case ADC12IV__ADC12IFG19:  break;   // Vector 50:  ADC12MEM19
        case ADC12IV__ADC12IFG20:  break;   // Vector 52:  ADC12MEM20
        case ADC12IV__ADC12IFG21:  break;   // Vector 54:  ADC12MEM21
        case ADC12IV__ADC12IFG22:  break;   // Vector 56:  ADC12MEM22
        case ADC12IV__ADC12IFG23:  break;   // Vector 58:  ADC12MEM23
        case ADC12IV__ADC12IFG24:  break;   // Vector 60:  ADC12MEM24
        case ADC12IV__ADC12IFG25:  break;   // Vector 62:  ADC12MEM25
        case ADC12IV__ADC12IFG26:  break;   // Vector 64:  ADC12MEM26
        case ADC12IV__ADC12IFG27:  break;   // Vector 66:  ADC12MEM27
        case ADC12IV__ADC12IFG28:  break;   // Vector 68:  ADC12MEM28
        case ADC12IV__ADC12IFG29:  break;   // Vector 70:  ADC12MEM29
        case ADC12IV__ADC12IFG30:  break;   // Vector 72:  ADC12MEM30
        case ADC12IV__ADC12IFG31:  break;   // Vector 74:  ADC12MEM31
        case ADC12IV__ADC12RDYIFG: break;   // Vector 76:  ADC12RDY
        default: break;
    }
}

3 年多前

0 admin 3 年多前

TI__Guru**** 2390755 points

请注意，本文内容源自机器翻译，可能存在语法或其它翻译错误，仅供参考。如需获取准确内容，请参阅链接中的英语原文或自行翻译。

你好，Oscar：

我假设当您说出"ADC标志"时，参考的是ADC中断服务例程。 ADC ISR在MCLK而非SMCLK上运行。

计时器外围设备与CPU内核分开。执行代码不会影响计时器的频率。

但是，您的说法是正确的：ISR中的代码过多通常不是好的做法，因为这会影响程序的流程。此外，printf是一个复杂的函数，需要许多时钟周期才能完成。我会避免在象ISR这样的计时关键代码中使用它。

此致，

Evan

0 admin 3 年多前

TI__Guru**** 2390755 points

请注意，本文内容源自机器翻译，可能存在语法或其它翻译错误，仅供参考。如需获取准确内容，请参阅链接中的英语原文或自行翻译。

您好，Evan：

感谢您的回复。我目前的问题是，如果ISR中没有代码，我不知道还有什么地方可以放置代码，因为我不知道在哪里可以找到代码。主代码在第62行之前循环，我不知道还有什么地方可以放它。

谢谢！

奥斯卡

0 admin 3 年多前

TI__Guru**** 2390755 points

请注意，本文内容源自机器翻译，可能存在语法或其它翻译错误，仅供参考。如需获取准确内容，请参阅链接中的英语原文或自行翻译。

奥斯卡

似乎您正在多个线程上处理类似的问题。我认为，将讨论集中起来比较容易。我要结束这一主题，让我们继续在这里讨论：

https://e2e.ti.com/support/microcontrollers/msp-low-power-microcontrollers-group/msp430/f/msp-low-power-microcontroller-forum/109.065万/msp430fr5994-adc-sampling-at-a-much-slower-rate-than-expected</s>5994

谢谢！

Evan

MSP 低功耗微控制器（参考译文帖）

MSP 低功耗微控制器（参考译文帖）(Read Only)

[参考译文] MSP430FR5994：当ADC处于信道重复序列模式时，在ADC环路之外执行计算