• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 47 subscribers
  • Views 504 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.

ADS1261 EVM

Ram Ch
Prodigy 120 points
Part Number: ADS1261EVM
Other Parts Discussed in Thread: ADS1261,

Hi Bryan,

I have a ADS1261 EVM connected to a NXP microcontroller. I did below two experiments.

1) Measured the Internal Signal (i.e., Internal (AVDD - AVSS) / 4 positive and Internal (AVDD - AVSS) / 4 negative) by configuring the INPMUX register. below is the ADC register configuration. 

1200SPS,SINC1,Pulse conversion,50us delay, Internal Reference, PGA=1. Sometime you said, with this configuration we should get 17.5 noise-free bits. bit actually I am seeing 15 bit accuracy. I took the log and attached.

Please see Internal-signal sheet, where column 1 is the actual 24 bit value. column 2, I just took the MSB 16 bits from 24 bit counts. column3 is the decimal equivalent of column2 values. At the end, I calculated the variation between 16 bit counts and observed the difference as 1.

2) Measured the external signal (3.3V is given to positive input and 2.5V is given to negative input). Both these 3.3v and 2.5v are taken from evaluation board itself(via test points). register configuration is same as experiment 1, except the INPBIAS register is not configured. Here the noise free bits are 14. 

Please see external-signal sheet, where column 1 is the actual 24 bit value. column 2, I just took the MSB 16 bits from 24 bit counts. column3 is the decimal equivalent of column2 values. At the end, I calculated the variation between 16 bit counts and observed the difference as 3.

evaluation-board.xlsx

My questions are

a) why I am not seeing the 17.5 noise free bits with above ADC configuration?

b) Is there any settings should be done on EVM? EVM is powered via USB and except the SPI connections from controller no other connections are made on EVM.

Please advise.

thanks,

Ram.

  

  • 0
    TI__Guru 51846 points

    Hi Ram,

    The internal measurements are meant to be coarse, so they likely add a fair amount of noise into the measurement. It seems the baseline level of noise increases by about 2.5x, which reduces the effective resolution to the values you are seeing. But the actual measurement was fairly consistent at 1.26V, which is ~AVDD/4.

    For the second measurement, have you tried using a low-noise source instead of the onboard voltages? I am not sure if the connections you are inputting are actually low noise, especially if they required soldering or jumper wires. But you should get better results with a low noise source.

    And have you tried shorting the ADC inputs together and measuring the noise to confirm the device is working as expected? You can set the INPMUX register to 0xFF (which is the default value) to short both ADC inputs to VCOM (AVDD/2).

    Let me know the outcomes of these tests.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 120 points

    eval_board.xlsx

    Hi Bryan,

    I) we tried 3.3V input to positive Mux from Agilent E3634A power supply and 2.5v input to negative mux from ADC EVM. Observed the count difference as 6 for 16 bit values. It seems the effective bits are 13 only. Please refer the external_3.3v_Input sheet. I captured actual 24 bit raw values in column1, MSB 16 bits of 24 bit values in column 2, decimal equivalent values of column1, column 2 in column3 and column4 respectively. What else can we do here to get the 17.5 bits resolution?

    2) I tried setting the INPMUX to 0xFF and captured the counts. Please refer Noise_Experiment sheet for 24 bit counts in column1 and MSB 16 bits of 24 bit counts in column 2. Can you tell what can be inferred from this experiment?

    Note: I am using the same ADC register configuration for ADC EVM.

    Thanks,

    Ram.

  • 0 in reply to Ram Ch
    TI__Guru 51846 points

    Hi Ram,

    I tried recreating your inputs using my ADS1261EVM: I shorted the 2.5V REFOUT to the AIN0 input, and then I used a Data Precision 2800 (DP2800) to connect a 0.8V signal across AIN0 and AIN1 to simulate your 3.3V input.

    I then removed the DP2800 as the source and connected an E3631A, with the (-) lead connected to the EVM GND terminal and the (+) lead connected to AIN1. The E3631A output was 3.3V.

    As you can see from the measured results below, the external power supply was far noisier than the DP2800, reducing the effective resolution by 5 bits. I would strongly encourage you to use precision instrumentation for these tests, otherwise you will not get good performance out of your device

    DP2800:

    E3631A:

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 120 points

    Thank you very much, Bryan. I will follow your suggestion.

    and one more thing is, I had sent you the noise test results. can you please comment on those results, if any?

    Thanks,

    Ram. 

  • 0 in reply to Ram Ch
    TI__Guru 51846 points

    Hi Ram,

    For your reference, in order to calculate noise, you would convert your hex outputs to decimal, then find the standard deviation of the dataset. This gives you the noise in codes. Then, find an LSB and multiply this by the # of codes to find the RMS noise in volts. For the ADS1261, one LSB = 2*VREF / Gain / (2^24).

    I performed these calculations on your dataset, and the calculated noise was 2.29uVRMS. The ADS1261 noise at 1200 SPS, sinc1,  and gain = 1 is 2.3uVRMS. So this information suggests the ADC is operating properly.

    -Bryan

  • 0 in reply to Bryan Lizon86
    Prodigy 120 points

    Thank you Very much, Bryan!