• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 47 subscribers
  • Views 1863 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.

ADS1262: How to set ADS1262 for maximum gain?

Bob Davorzak
Prodigy 40 points
Part Number: ADS1262

I have an ADS1262 connected to a 5.0000 voltage reference and voltage divider to bring voltage down to 10mV. After it is stable, for example I get 9.99994182 or 9.99994277. It jumps back and forth between those two values.  I'm just wondering if I have the gain set right in my C++ code. Here is the line:  PC_ADS1262.ads1262_Reg_Write(MODE2, 5<<4 | 1);

The 5<<4 is the part that seems to set the gain. Do you know if that is correct to get a gain setting of 32? Does anything other than MODE2 need to be adjusted?

  • 0
    TI__Guru**** 180931 points
    Hi Bob!

    Welcome to the e2e Forum! If I'm remembering my bit-wise operations correctly, your code line should be putting 0x51 into the MODE2 register, which would be gain = 32 and data rate = 5 SPS. Are you able to do a Register Read and verify that what you wrote is actually getting loaded?
  • 0 in reply to Tom Hendrick
    Prodigy 40 points

    Thanks Tom. I tried 0x51 instead of 5<<4 and get the same resolution. So I suspect it is set for the maximum gain of 32V/V.  Sorry, but I do not know how to do a Register Read. I read in the datasheet that there is a RREG command, but I simply do not know how to use it.

  • 0 in reply to Bob Davorzak
    TI__Mastermind 44680 points

    Hi Bob,

    Setting the MODE2 register to 0x51 is correct for configuring the ADS1262 in a gain of 32 V/V.

    To verify this register setting with the RREG command, you would need to send a very similar command sequence to the WREG command. The only differences are to replace WREG with RREG, and instead of sending data send "0x00" to clock out the current register value. Therefore, the SPI command sequence to read the MODE2 register would be "0x25, 0x00, 0x00", and you would expect to get receive "0x51" while clocking out the second "0x00" byte.

    Another thing I might suggest trying is shorting the inputs together to see if the noise performance looks okay. In this configuration, the ADC results are much less likely to be influenced by the input signal and you ought to see a normal (Gaussian) distribution of output codes.

    I hope that helps!

    Best Regards,
    Chris

  • 0 in reply to Christopher Hall
    Prodigy 40 points

    Thanks Chris. I shorted the input leads together and ran it for about 5 minutes using a 99/100 moving average. It read from 0.00000069506 to 0.00000071092 volts. Does that seem normal to you?

  • 0 in reply to Bob Davorzak
    TI__Mastermind 44680 points

    Hi Bob,

    Without seeing the raw data, it's hard to say. Are you looking at the input-referred or output-referred noise?

    The noise table in the ADS1262 datasheet shows the input-referred noise. Meaning that the noise is divided by the PGA gain. Specifying noise in this way tells you the smallest input signal you can resolve. For example, using the FIR filter at a data rate of 5 SPS and a gain of 32 V/V, the input referred noise (i.e. the smallest resolvable signal) is 0.015 uVrms. However, if you were to short the inputs together and measure the noise (after the PGA), you would actually see 0.48 uVrms (32 V/V x 0.015 uVrms). Using a crest factor of 6.6, I would then expect the peak-to-peak noise performance to be 0.099 uVpp (input-referred) or 3.168 uVpp (output referred).

     

    After averaging and collecting data over a fairly long period of time, you measured the noise to be 0.01586 uVpp (I'm going to assume that this was the input-referred noise). If the noise was truly random (Gaussian), then averaging 100 samples should have reduced the RMS noise by a factor of sqrt(100) = 10. Therefore, before averaging your noise would have been about 0.158 uVpp, which is a bit larger than the expected 0.099 uVpp noise performance. You might be looking at additional noise coming from your system OR you may also be seeing the effects of drift on the noise.

     

    I would recommend collecting 10-20 secs of data (without averaging) and comparing it to the noise table in the data sheet.
    ...Then your measured noise should correspond fairly well to the device specifications. Collecting data over a longer period of time allows for more offset drift in your measurement. The noise performance may be right on target with the device performance; however, if there is any offset drift then the peak-to-peak noise will appear degraded due to the change in offset.

    Best Regards,
    Chris

  • 0 in reply to Christopher Hall
    Prodigy 40 points

    Thanks a lot Chris, that's a big help.

  • 0 in reply to Bob Davorzak
    TI__Mastermind 44680 points

    You're welcome! Let me know if there is anything else I can do to help.

    Best Regards,
    Chris