Reading ADC0 through ADC5 on ADuC7024

Hi Jeff,

one way is to program Timer1 for a 100Hz periodic IRQ and take the 6 samples in the Timer 1 Interrupt Handler.
I have attached such a simple project for Keil µVision4.
Here I used a different setting for the ADC clock to improve the acquisition (16 clocks), but also to increase the
sampling speed. I have seen with this settings better results than with the default from datasheet.
Additional improvements are possible.

If you replace the "for"-loop with sequential statements, you can reduce the approx. 11µs it takes to do the

6 samples and to store the ADC results.

Hi MMA,

I loaded up your project.  I assume this is using the Real-view compiler?  When I build the project, I get these errors:

ADuC7024_ADC0_to_ADC5.axf: Error: L6200E: Symbol IRQ multiply defined (by adc0_to_adc5.o and irq_arm.o).
ADuC7024_ADC0_to_ADC5.axf: Error: L6200E: Symbol SWI multiply defined (by adc0_to_adc5.o and irq_arm.o).
ADuC7024_ADC0_to_ADC5.axf: Error: L6200E: Symbol FIQ multiply defined (by adc0_to_adc5.o and irq_arm.o).
ADuC7024_ADC0_to_ADC5.axf: Error: L6200E: Symbol UNDEF multiply defined (by adc0_to_adc5.o and irq_arm.o).
ADuC7024_ADC0_to_ADC5.axf: Error: L6200E: Symbol PABORT multiply defined (by adc0_to_adc5.o and irq_arm.o).
ADuC7024_ADC0_to_ADC5.axf: Error: L6200E: Symbol DABORT multiply defined (by adc0_to_adc5.o and irq_arm.o).

I have seen these before when porting something from the GNU compiler to the Real-view compiler.  Usually it has something to do with the startup file.

What do you recommend?

Hi Jeff,

this mainly is related to the tool installation and the used Compiler files for C-startup and

what include-files you use.

I have used the latest version of µVision4 (MDK4.12) downloaded from the Keil WebSite.

Also I used in that toolset the RealView C-Compiler.

Possibly you check out the mentioned files for a difference to the latest installed files.

You can use for reinstallation the backup-option to do that.

Hi Jeff,

can you provide detailed schematics (PDF-File) & layout (Gerber or PDF).

Most of the cases I've seen are layout related.

Michael,

In the attached code, the ADC is oversampled.  I am using your settings for a Heart Rate Monitor application.  I am sampling at 100Hz per your settings.  However, maybe my board has a poor layout because there appears to be spikes at 8.9 kHz (looking on a scope) that are riding on top of the 100Hz signal.  Do you think this could be happening due to the settings here:

ADCCP  = 0;    // select ADC Channel
   ADCCON = 0x03E3;  // start ADC
         // fADC/1=> sampling @ 1.193Msps (0.838µs)
         // 16 * fADC aquisition
         // Enable start
         // Enable ADCBusy output
         // Enable ADC power
         // Single-Ended Mode
         // Single-SW-Conversion
   ADCCON = 0x0363;  // Disble ADC restart -> see datasheet
   while(ADCSTA != 0x01); // wait for ADC to convert
   // store ADC Reseult
   sensorState.sensorData[(sensorState.sensorDataIndex * SENSORDATASIZE)+5] = ADCDAT >> 16;

I do repeat this code 6 times with ADCCP = 0 through 5.

Any idea as to why I might have a 8.9 KHz spike in my signal??