• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 52 subscribers
  • Views 3885 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.

CC2541 ADC Conversion

Jeremy Jass
Prodigy 125 points
Other Parts Discussed in Thread: CC2541

I've been trying to set up a pin to monitor battery voltage on the CC2541 by measuring a partial voltage created by a voltage divider.  Pretty straight forward, and it seems to be working great, only I can't figure out how to correlate the ADC readings back to the measured voltages in a way that makes sense.  First just to make sure I'm not missing anything obvious in my code:

Initialization (called once on powerup)

P0SEL &= ~BIT0; //GPIO

P0DIR &= ~BIT0; //Input

P0INP |= BIT0; //Disable Pull-up

Conversion (called when measuring)

uint16 adc_result = 0;

APCFG |= APCFG_APCFG0; //Configure P0_0 to ADC

ADCCON3 = ADCCON3_EREF_AVDD | ADCCON3_EDIV_512 | ADCCON3_ECH_AIN0; //Start single conversion: AVDD5 ref, 14 bit resolution, P0_0

while( !(ADCCON1 & ADCCON1_EOC) ); //wait for conversion to finish

APCFG &= (APCFG_APCFG0 ^ 0xFF); //Unconfigure P0_0 to ADC

adc_result = (ADCH << 8);  //Shift high result

adc_result |= ADCL;  //OR to low result

return (adc_result >> 2);  //Shift and return

Originally I had the APCFG call made only once in the initialization routine, but looking into the HAL code I mimicked their function by switching it on and back of for each conversion.  This didn't make any difference.

The numbers I get back are close, and they follow a very good correlation, I just don't know how to work back from them to measured voltages.  What I have been assuming, is that for a single-ended, single conversion, I could get back to the voltage by:

(ADC_Result / Full_Scale_Range) * Reference_Voltage

I know ADC_Result from the CC2541, and I can measure the Reference_Voltage (3V).  If I assume the Full_Scale_Range is about 7400, everything lines up perfectly, but 7400 seems like a random number so either my calculation is wrong, or I'm messing up the conversion.  In other testing using the internal 1.24V reference voltage it appears the ADC caps out at 0x1FFF which would set the full scale range to 8192.  This makes sense and is sorta kinda close to 7400, but not close enough to leave me confident in the result.

Thanks

  • 0
    TI__Mastermind 29650 points

    Hi Jeremy,

    APCFG can be set static, what it does is to disable the digital input and output buffer so "analog" input voltages are allowed.

    The ADC will give out 14 bits into ADCH/ADCL. The most significant bit is the sign bit giving a range between - 8192 and 8191.

    What are the values of the external resistances you are using? It might be that the ADC loads the source too much, which could be easily verified with an oscilloscope during an conversion. You will need lower source resistance or capacitive decoupling on the input if so.

    The voltage you measure should then be:

    Vin = (adc_result) * Vref / (213 -1)

    Regards,

    Svend

  • 0 in reply to svendbt
    Prodigy 125 points

    We were using two 56k resistors as the voltage divider, and measuring on an oscilloscope the lower voltage during the conversion correlated exactly to the measurements we were seeing.  For now knowing the numbers are right and there is a relationship between ADC readings and battery level should work, but I'll be sure to add a decoupling capacitor in the next hardware revision.

    Thanks