CCS/TM4C1294NCPDT: ADC TIMMER TRIGGER CONFIGURATION

Hello Ralph,

I made this change as you suggested, but i'm still getting wrong measures from my ADC.

Now i'm using this configuration:

//ADCClockConfigSet(ADC0_BASE, ADC_CLOCK_SRC_PIOSC | ADC_CLOCK_RATE_FULL, 1);
ADCClockConfigSet(ADC0_BASE, ADC_CLOCK_SRC_PLL | ADC_CLOCK_RATE_FULL, 24);

And already tried the one that's above.

It seems to be unstable, even when there is a fixed input the converted values are varying a lot. Can you help me with this issue?

Thanks in advance!

Hello Tiago,

Ah so you are getting wrong measurements now? I had thought it wasn't working period. That is quite a different situation then.

Can you explain what you are measuring, and any sources of impedance that would affect the input impedance of the ADC pin?

Hello cb1,

Thanks for mentioning the SysCtlClockGet(), I missed that as well. I was informed from another colleague of the same.

Tiago, please remove the SysCtlClockGet() call and input the clock frequency based on the output of SysCtlClockFreqSet (many users store this as a variable for ease of reference).

Hi Ralph,

Thank you - and it should be (again) noted that poster's 'preliminary' choice of '1mS' - as the ADC's 'Sequenced Conversion Call' rate, proves:

• impossible to achieve w/in the split timer's, 16 bit register capacity, when clocked from a 120MHz source
• in all probability is 'FAR TOO FAST' for poster's initial, 'Test/Verify.'     'KISS' directs the slowest, easily achievable rate - which 'vastly heightens the odds of poster success!'
• use of the Timer - as ADC trigger - adds (another) element of risk - thus violates 'KISS.'     Instead - old standby, 'ADC processor trigger' should be employed  - at least until ADC success results.

Soon departing (intern) staff (school starts next week) have provided a 'rich guide' of tech items - worthy of poster's time & consideration.    (even (others) arriving here - should benefit from this dedicated group's focus & effort...)

Hello! Thank you all so much for your time and effort to help me!

It was all about the clock configuration, I made some modifications in order to set it correctly, including removing the SysCtlClockGet().

Here is my code after the modifications:

//Inclusão de Arquivos/Bibliotecas
#include <stdio.h>
#include <stdlib.h>
#include <stdint.h>
#include <stdbool.h>
#include "inc/hw_types.h"
#include "inc/hw_memmap.h"
#include "inc/hw_ints.h"
#include "driverlib/adc.h"
#include "driverlib/gpio.h"
#include "driverlib/interrupt.h"
#include "driverlib/pin_map.h"
#include "driverlib/sysctl.h"
#include "driverlib/uart.h"
#include "utils/uartstdio.h"
#include "driverlib/timer.h"

//Definição de variáveis
#define NUMBER_OF_INTS 32000

int cont = 0;
int i = 0;
int ADC_FREQKHZ = 30;
uint32_t SysClock = 0;
uint32_t ADC0Value[2];
uint32_t DATA_ADC0[NUMBER_OF_INTS];
uint32_t DATA_ADC1[NUMBER_OF_INTS];

void InitConsole(void)
{

SysCtlPeripheralEnable(SYSCTL_PERIPH_GPIOA); 

GPIOPinConfigure(GPIO_PA0_U0RX); 
GPIOPinConfigure(GPIO_PA1_U0TX); 

SysCtlPeripheralEnable(SYSCTL_PERIPH_UART0); 

UARTClockSourceSet(UART0_BASE, UART_CLOCK_PIOSC); 

GPIOPinTypeUART(GPIO_PORTA_BASE, GPIO_PIN_0 | GPIO_PIN_1);

UARTStdioConfig(0, 115200, 16000000); 
}

void InitADC(void)
{
SysCtlPeripheralEnable(SYSCTL_PERIPH_ADC0); //Habilita o canal ADC0

SysCtlPeripheralEnable(SYSCTL_PERIPH_GPIOE); //Habilita a porta E

GPIOPinTypeADC(GPIO_PORTE_BASE, GPIO_PIN_3 | GPIO_PIN_2); 
//GPIOPinTypeADC(GPIO_PORTE_BASE, GPIO_PIN_3);
SysCtlDelay(80u);

//ADCClockConfigSet(ADC0_BASE, ADC_CLOCK_SRC_PLL | ADC_CLOCK_RATE_FULL, 10);
//ADCClockConfigSet(ADC0_BASE, ADC_CLOCK_SRC_PIOSC | ADC_CLOCK_RATE_FULL, 1);
ADCClockConfigSet(ADC0_BASE, ADC_CLOCK_SRC_PLL | ADC_CLOCK_RATE_FULL, 24);
SysCtlDelay(10u);

IntDisable(INT_ADC0SS0);
ADCIntDisable(ADC0_BASE, 0);
ADCSequenceDisable(ADC0_BASE, 0);

ADCSequenceConfigure(ADC0_BASE, 0, ADC_TRIGGER_TIMER, 0); 

ADCSequenceStepConfigure(ADC0_BASE, 0, 0, ADC_CTL_CH0);
ADCSequenceStepConfigure(ADC0_BASE, 0, 1, ADC_CTL_CH1 | ADC_CTL_IE | ADC_CTL_END);

ADCHardwareOversampleConfigure(ADC0_BASE, 8);

ADCIntClear(ADC0_BASE, 0); //Define a flag para iniciar as conversões
IntEnable(INT_ADC0SS0);
ADCIntEnable(ADC0_BASE, 0);
ADCSequenceEnable(ADC0_BASE, 0); //Habilita o canal 0

}

void InitGPIO(void)
{
SysCtlPeripheralEnable(SYSCTL_PERIPH_GPION); //Habilita o GPION
GPIOPinTypeGPIOOutput(GPIO_PORTN_BASE, GPIO_PIN_0 | GPIO_PIN_1 | GPIO_PIN_2); 
GPIOPinWrite(GPIO_PORTN_BASE, GPIO_PIN_0 | GPIO_PIN_1, 0X02); 
}

void InitTimer0 (void)
{
SysCtlPeripheralEnable(SYSCTL_PERIPH_TIMER0);

TimerConfigure(TIMER0_BASE, TIMER_CFG_SPLIT_PAIR | TIMER_CFG_B_PERIODIC);

TimerLoadSet(TIMER0_BASE, TIMER_B, SysClock/(1000*ADC_FREQKHZ));

TimerControlTrigger(TIMER0_BASE, TIMER_B, true);

TimerIntEnable(TIMER0_BASE, TIMER_TIMB_TIMEOUT);

IntMasterEnable();

TimerEnable(TIMER0_BASE, TIMER_B);
}

void ADC0SS0IntHandler(void)
{
cont+=1;
ADCIntClear(ADC0_BASE, 0);
ADCSequenceDataGet(ADC0_BASE, 0, ADC0Value);
//UARTprintf("AIN0 = %d\n", ADC0Value[0]);
DATA_ADC0[cont] = ADC0Value[0]; //PE3
DATA_ADC1[cont] = ADC0Value[1]; //PE2

if(cont >= NUMBER_OF_INTS)
{
IntDisable(INT_ADC0SS0);
ADCIntDisable(ADC0_BASE, 0);
ADCSequenceDisable(ADC0_BASE, 0);
ADCIntClear(ADC0_BASE, 0);
}

}

int main(void)
{

SysClock = SysCtlClockFreqSet(SYSCTL_XTAL_25MHZ | SYSCTL_OSC_MAIN | SYSCTL_USE_PLL | SYSCTL_CFG_VCO_480, 40000000); 
InitGPIO();

InitConsole();
InitADC();
InitTimer0();

cont = 0;

while(1)
{
if(cont >= NUMBER_OF_INTS)
{
cont = 0;
for(i==0;i<=NUMBER_OF_INTS;i++)
{
UARTprintf("Sample %d: %d\t%d\n", i, DATA_ADC0[i], DATA_ADC1[i]);
SysCtlDelay(10000);
}

}
}
}

Thank you - my group (and surely this vendor) are pleased that you've persisted - and succeeded!

Yet may we ask - what is the 'expected  Timer-Triggered, ADC rate' - achieved via:

"TimerLoadSet(TIMER0_BASE, TIMER_B, SysClock/(1000*ADC_FREQKHZ));"

You may have noted that staff (earlier) identified  '~564µS'  as the maximum value - containable w/in the Timer's 16 bit Register - when driven from a 120MHz clock source!      Are you certain that your "new implementation' indeed 'fits' - w/in those 16 Timer bits?

We have just noted - FAR DOWN your code listing - that you (appear) to have reduced System Clock to 40MHz.     That should expand your 16 bit Timer Register to a 1.5mS (max) capacity.    (assuming 'everything else' - remains constant.)    Such DOES NOT - prove the case!   (we only later - discovered!)

Instead - you define, 'ADC_FREQKHZ' as 30' - thus we believe your final (3rd) parameter w/in TimerLoadSet() is (now) 30,000.     (which btw - fits NICELY w/in a 65K Timer Register - is this (30K) correct?)

And - kindly do describe the variance of your ADC measurements - when feeding a 'fixed input signal level' to your ADC.    (two channels are (now) being converted...)    It is expected that the 3 lsb of the ADC's conversion value - will reveal 'jitter' - characteristic of (any) MCU's ADC...

Finally - you've not (yet) adopted staff's (earlier post) suggestion - of  employing 'ALL EIGHT STEPS' - w/in 'ADC Sequence 0!'      Would not 8 (back to back) ADC Measures (of the SAME ADC Channel) - prove of (even) greater value?     (variations w/in such 'rapid conversions' may (likely) reveal - 'improper MATCH' - between your analog signal's output - and the ADC's input.    Such proves of value ... does it not?