AnsweredAssumed Answered

ADuCM360 delay needed between finishing single shot conversion and issuing a new one

Question asked by Bohus54 on Apr 7, 2017
Latest reply on Oct 4, 2017 by Bohus54

Hi community,


Recently I have spent quite a long time with the following issue: I need to measure an ADC input in continuous mode with reding results in interrupt. Occasionally I need to take one sample from different input and return back to the continuous mode. I wanted to perform four single shot conversions to settle the filter and take the last one as a result. But the single convert mode always timed out. After reading datasheet again and again, searching forums and a lot of source code tweaks I found a solution by adding a few us delay after reading single conversion result when the SM shall return back to the idle state before issuing command for new single conversion by writing 2 to ADC0MDE. The code excerpt looked like this:


case SD0_VOLT:                //ADC0 mesures INP2

    pADI_ADC0->MSKI = 0; //stop ADC0 interrupts measuring INP0
    pADI_ADC0->CON = //change +INP mux to INP2, enable unipol no buffer intref ch2 AGND as -inp
             (1<<19) + (1<<18) + (0xF<<14) + (0<<12) + (2<<5) + 0xF ;

    dwData = 0xFFFF;             //error return
    bStatus = pADI_ADC0->STA; //clears status bits
    for(i=0 ; i<4 ; i++)                //measure 4-times to settle filter
         Delay(5);    //few us needed from finishing to starting new conversion !!!!!
         pADI_ADC0->MDE = 2; //single conversion
         dwStrt = dwTick;
             if(dwTick - dwStrt > ADC_TO)          //timeout 20ms
                 goto Exit1;
             bStatus = pADI_ADC0->STA;
             if(bStatus & 1)
                  dwData = pADI_ADC0->DAT >> 12;       //resets RDY, goes to idle
                  break; //exit while loop
     pADI_ADC0->CON = //+INP back INP0 enable unipol no buffer intref chan0 AGND as -inp
               (1<<19) + (1<<18) + (0xF<<14) + (0<<12) + (0<<5) + 0xF ;
     pADI_ADC0->MDE = 1; //continuous conversion
     pADI_ADC0->MSKI = 1; //renew ADC0 interrupts measuring channel A current
     return dwData;


Is this behavior of the ADC expected or mentioned anywhere in the datasheet or did I anything else wrong? The code was compiled with Keil MDK522 eval version tool.


Your insight would be appreciated.