• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 46 subscribers
  • Views 561 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.

ADS131A02: Wrong data (high noise while inputs shorted)

AIU2017
Intellectual 360 points
Part Number: ADS131A02


Hi.

http://www.ti.com/lit/ds/symlink/ads131a02.pdf

I have some problems with this ADC:

1. High noise while inputs are shorted (diffrerence ~1200 adc points, VREF = 4 V, so noise is about 500uV in 24bit resolution)
2. SPI frame error: F_FRAME in STAT_S register is set after I done all initialization and truing to get data.

My setup:

AVCC 5V
DVCC 3.3V
VREF 4V - internal used
M0 - DVCC 
M1 - DVCC
M2 - GND

So its Asynchronous interrupt mode, 32 bit SPI word size, Hamming code word validation off.

My intialization steps:

1. set RESET pin high
2. read 0x00 register to obtain 0xFF02 answer:



3. Unlock ADC(0x0655)


4. At this stage I can configure ADC already, but I made read of all registers.

For example, reading STAT_1:


As you can see, all faults bits are clear.

Futher register readings gives me next:

STAT_P - 0x00
STAT_N - 0x00
STAT_S - 0x00 (no SPI errors!)
ERROR_CNT - 0x00
STAT_M2 - 0x05 (0000 0101b - setup as I described above)
A_SYS_CFG - 0x60
D_SYS_CFG - 0x3C
CLK1 - 0x08
CLK2 - 0x86
ADC_ENA - 0x00
ADC1 - 0x00
ADC2 - 0x00

Pretty same as it should be by default.

5. Now I set CLK2 to E0 (1110 0000b)
So

  • fMOD = fICLK / 14
  • fDATA = fMOD / 4096


Making delay 260MS and readback this register to confirm setup.

6.  Set A_SYS_CFG to 0x78 (0111 1000b)

So 

  • REFP is set to 4.0 V
  • Internal reference voltage enabled and selected

Making delay 260MS and readback this register to confirm setup.

7. Set ADC_ENA to 0x0F

Responce differs from datasheet - it should be 0x0F, but I receive 0x03.
I suppose its because I'm using 2channel device (so 0x03 = 0000 0011b), but correct me if I'm wrong.

Making delay 260MS and readback this register to confirm setup (its again 0x03).

8. Send WAKEUP:


Making delay 260MS.

9. Send LOCK command:


Making 300MS delay after this.

10. Polling RDRY pin: when it go LOW, I'm start sending NULL command:




Now error 0x30 appeared (it says SPI error), I've read additionaly STAT_S register and it gives 0x01:
Not enough SCLKs are sent per frame


Moreover, my channels are shorted and results are:

0x000739 = 1849 for CH1
0x0004F0 = 1264 for CH2

Why its so differs?

Here is some others samples:
0x000736 = 1846 for CH1
0x0004EE = 1258 for CH2

0x000731 = 1841 for CH1
0x0004EB = 1259 for CH2


Hope for assistance, thanks!

Regards

  • 0
    TI__Guru* 98705 points

    Hi,


    At this point, it looks like you've followed the startup procedure from figure 106 of the datasheet.

    I would start by trying to make some alternate measurements with a real voltage input. My first concern would be that the difference in the ADC channels may just be a channel-to-channel variation in the offset. Looking at the values for the first measurement:

    1849 codes = 4V * (1849/2^23) = 0.88mV

    1264 codes = 4V * (1264/2^23) = 0.66mV

    This is not far from what I might expect for the offset for the device. The datasheet lists the offset as typically 0.5mV. While this might be slightly larger than the average, this is not unexpected.

    Using a larger voltage, you may be able to determine if this is really an offset error, or if there is some additional gain error associated with either channel.



    Joseph Wu

  • 0 in reply to Joseph Wu
    Intellectual 360 points

    Thanks for reply.

    Can you please specify where exactly in datasheet offset value (~0.5mV) is mentioned?

    Also why I receive so much ADC codes while inputs are shorted? I suppose it should be much lower? 
    Or there is some fixed offset? So I can do just something like this in calculations:

    CH1_ADC_codes = ADCvalueCH1 - offset;
convert_to_voltage(CH1_ADC_codes ); // make conversion to voltage with corrected value of adc codes...



    And I would like to know, how this values were obtained? Is it include offset correction?




    About gain. I havent modify any settings with gain, moreover, datasheet says that modification of gain available only for 4-ch version of IC:


    Regards

  • 0 in reply to AIU2017
    TI__Guru* 98705 points

    Hi,



    The offset specification can be found in the Electrical Characteristics. I've clipped it out in the following picture.


    This is a DC offset. Once you have measured this offset, you can subtract this offset error from future measurements to remove it. However, I would note that the offset has a drift component of typically 2.5uV/°C. If the temperature changes substantially, you may need to take another offset measurement.

    In the Dynamic range table you've shown, the column that you boxed in red is basically the noise measurement. In that case, the inputs are shorted, the offset is removed and the RMS value of the noise is calculated. This is the root-mean-square value of the noise.

    For the digital gain configuration registers, the sentence is a mistake. Only 13h and 14h are reserved and the gain control should work on the ADS131A02 channels. If you look at the Register Map on Table 15, this is more clear:




    Joseph Wu