ADS1282 Programming

Hi Joseph,

Thanks for quick reply,

I managed to write and read data from ADS1282.

I have checked CLK and SCLK pins both seems ok. I am providing 3.75MHz on CLk pin and SCLK is 750KHz.  Attached images of oscilloscope Yellow is SCLK, Blue is DOUT. 

I have powered up and running at default settings. It means "Read Data Continuous mode". I don't need to send any commands to read data (according to data sheet). I means gain=1, 1000sps, PGA chope enabled.

I am receiving data from ADC but i am not able to understand this data is correct or not? 

I am receiving data like

9769642 = 9512AA, 15161788 = E759BC, 13117080 = E759BC, 16655786 =FE25AA, 17452124 = 10A4C5C, 14908740=E37D44, 18463082 =119B96A e.t.c

My reference voltage is 0 to 2.5V.

How to convert this data into intpu voltage? I would like to know this data is correct or not?

It seems i am not able see repeated data because I am not changing input to ADC at AINP1&AINN1. 

Regards,

Pavan.

Hi Joseph,

I am using REF5025-HT as 2.5V reference source. As well the reference voltage is  VREF+ =2.5V , VREF-=0V.

I have connected some sensors as input values to ADC, sensors output values are very low in voltage nV to mV range. 

Now i changed my configuration with PGA (64) factor and data rate to 250SPS. I am using Sinc + LPF filter blocks as well.

I am able to read almost constant values in each iteration compared to yesterday.

I am reading values almost around 4233865853 ; according formula (4233865853/2)*1.25/2^30 = 2.33mV.  

1) In the conversion calculation you used 1.25V, we need to use -1.25 to +1.25 V even when we apply reference voltage 0 to 2.5V. 

2) need to change conversion formula when we change PGA.

I have checked today the clock and DOUT images it seems less noisy than yesterday. 

Any suggestion to get better noise performance.

Regards,

Pavan.

Hi Pavan,

I'll be happy to help you with this issue...

The maximum fCLK frequency specified in the datasheet is 4.096 MHz. Above that, you may get poor performance or the device may not function at all. Typically, I've seen people get away with overclocking the ADC by about 10% (even though performance is no longer guaranteed in this condition); however, I'd expect that overclocking it by 80% would certainly have some consequences.

First, try reducing fclk back into the 1-4.096 MHz range and collect data again to see if the results make more sense. (Note that the SCLK frequency must be in the range of 1/2 to 1/16 of the fCLK frequency.)

If you need a faster data rate, you'll need to configure it with the DR[2:0] bits in the CONFIG0 register.

If you find that the 0V input is not quite resulting in output codes near 0x00000000 or 0xFFFFFFFF, then you may need to perform offset calibration to improve the result.

Best Regards,
Chris

Hi Christopher HAll,

Thanks for your reply,

I was trying as you suggested. I changed my design little bit and working on 3.75MHz FCLK as i can't get exactly 4MHz with my controller. SPI clk SCLK is 1.875MHz. I have 0 to 5V reference.

I have PGA=2, 1000SPS,linear pahse, SINC+LPF filters. (CONFIG0=0x52).

On chaneel one inputs (AINP1+, AINN1- )I have diffrential voltage of 0V (2.6-2.6). I am reading a value via SPI = 4278223117 (0xFF00810D). when i look at Oscilloscope code value it is like this 4294939449 =0xFFFF9339.this is near to FFFFFFFF.  have a look at oscilloscope image.  can say i am reading correct?

I don't know why i am not able to read same value as oscillocope. Then i tried on channel two. which is strange for me.

Channel two intputs(AINP2+, AINN2- ) have diffrential voltage change from 0.217V to 0,223V. I am reading a values

1694520226 (0x650053A2) ,

2684409570 (0xA000D6E2),

3439355800 (0xCD006798) .....

in between those ranges its changing but when i measure with osclloscope the code values is diffrent than what i am reading. Have a look at oscilloscope images.

From oscilloscope i am always getting 365744896 (0x15CCD300), 358744806 (0x156202E6), 354225176 (0x151D0C18), 369138436 (0x16009B04)........

These values are almost near values i am getting always in this range with oscilloscope codes.

I have some questions

1) when i use the conversion fromula provided by joseph, the oscilloscope values are also not equal to 0.217 to 0.223 V. Even with readings i have got from SPI also not matching when i use the same fromula from joseph?

2) I am not understanding why i am getting diffrent values when i read via SPI than oscilloscope code measurements?

but if i can say i am reading wrong then i am getting near values with channel one (ov input. at that case my spi reading values and oscilloscope code values are nearly matching.

am i converting wrong or reading wrong?

My  SPi reading code is like this

unsigned int ADC1282_16(unsigned char cmd, unsigned int value) 
{ 
    unsigned int data; 
    SW_Delay(DELAY_10US);
    spi(cmd);                   //Place instruction to read
    while (I2C2DIN & 0x0001);
    data = spi(0);     // read hi-byte 
    data = (data << 24)| spi(0);
    data = (data << 16) | spi(0);    
    data = (data << 8) | spi(0);    // and lo-byte
    return data; 
} 

unsigned int readADC(void)  
  {                 
                          // wait till ADC is ready 
       value = ADC1282_16(0x10, 0);                           // read registe
//       printf("%d ", value);
       return value;
  } 

Any suggestions.

Regards,

Pavan.

Hi Pavan,

To answer your questions...

1. I'm getting the correct answer using Joseph's formula: For example (365744896/2)*1.25/2^30 = 0.212 V. This formula first right shifts (i.e. divides by 2) the ADC code to remove the redundant sign bit, then that result is multiplied by the 31-bit LSB size. The LSB size is FSR/2^31 or Vref/2^31 or 1/2*Vref / 2^30 (these are all equivalent).
   
   
2. I think I see the issue in your code...
   

   pavan said:

       data = spi(0);     // read hi-byte

       data = (data << 24)| spi(0);

       data = (data << 16) | spi(0);   

       data = (data << 8) | spi(0);    // and lo-byte

       return data;

   
   
   Here you seem to left-shift the data bytes too many times. I think what you intended to do is this...
   

       data = spi(0);     // read hi-byte 
       data = (data << 8) | spi(0);
       data = (data << 8) | spi(0);    
       data = (data << 8) | spi(0);    // and lo-byte
       return data; 

   
   Give that a try and see if you get better results!

Best Regards,
Chris