F28335 ADC Data Climbs with DC Input

Hi Gautam,

                Here is my initialization code and the setup before the burst.

#define ADC_MODCLK   0x3

#define ADC_ADCBGRFND 0x00C0

#define ADC_ADCPWDN        0x0020

#define ADCTRL3_INIT ADC_ADCBGRFND + ADC_ADCPWDN

#define ADCMAXCONV_INIT    15              // 16 conversions

#define ADCCHSELSEQ_SETUP  0x0000     // All are channel 0.

void AdcHwInit()

{

     // Let the system setup the ADC and wait for it to settle.

     InitAdc();

     // The following setup is to achieve the maximum conversion rate of 12.5 MSPS.

     // It is taken from spru812a (ADC document) page 26 (Table 1-7).

     // HISPC

     // The peripheral clock rate is SYSCLKOUT (HISPC + 1)

     EALLOW;

     SysCtrlRegs.HISPCP.all = ADC_MODCLK;

     // Assumes that the PIE Vector Table has already been setup.

     PieVectTable.ADCINT = &AdcInterrupt;

     EDIS;

     // ADCTRL1

     AdcRegs.ADCTRL1.bit.ACQ_PS = 0;            // S/H width in ADC module periods = 1 ADC clock to match the ADC's conversion rate

     AdcRegs.ADCTRL1.bit.SEQ_CASC = 1;          // A sequence is 16 conversions.

     AdcRegs.ADCTRL1.bit.CONT_RUN = 1;          // Keep the ADC running.

     // ADCTRL2

     AdcRegs.ADCTRL2.bit.EPWM_SOCB_SEQ = 1;          // Allow the SEQ to be started from the ePWM

     AdcRegs.ADCTRL2.bit.INT_ENA_SEQ1 = 1;

     // ADCTRL3

     AdcRegs.ADCTRL3.all = ADCTRL3_INIT;

     // ADCMAXCONV

     AdcRegs.ADCMAXCONV.all = ADCMAXCONV_INIT;

}

void AdcSetupCapture()

{

     AdcRegs.ADCTRL2.bit.RST_SEQ1 = 1;         // Reset SEQ1

     // From spru812a.pdf (section 1.7)

     // If continuous conversions of the same channel are needed then all the

     // CONVxx registers should have the same channel address.

     AdcRegs.ADCCHSELSEQ1.all = ADCCHSELSEQ_SETUP;

     AdcRegs.ADCCHSELSEQ2.all = ADCCHSELSEQ_SETUP;

     AdcRegs.ADCCHSELSEQ3.all = ADCCHSELSEQ_SETUP;

     AdcRegs.ADCCHSELSEQ4.all = ADCCHSELSEQ_SETUP;

}

Thanks,

Ed

Hi Ed, can you try another voltage?  I would recommend a resister divider from the 3.3V, an external power supply, or 1.5V battery.

note: desktop supplies are rather noisy (high DC code spread) but should serve for the purpose or looking for the asymptotic rise.

The full scale voltage is 3.0V for the ADC so tying to the 3.3V rail will certainly saturate the conversion results.

I would like to see if this is something that occurs throughout the transfer function range (0-3V) or only when tied to ground.  I have seen on occasion a disturbance of the REFLO or AGND when an ADC channel is tied directly which can take some time to settle out.

You might also note what filter cap is present for the ADC input pins.