• State
  • Locked Locked
  • Replies 13 replies
  • Subscribers 46 subscribers
  • Views 5581 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.

ADS4142EVM: ADC value mapping

Himanshu Chaudhary
Prodigy 130 points

Part Number: ADS4142EVM
Other Parts Discussed in Thread: ADS4142

I am using ADS4142 EVM in Parallel CMOS mode. But irrespective of the input I am always getting 16283 value (i.e 2^14 -1). Can someone help me out  with the ADC value mapping with input voltage i.e.

if

   1.8V-> 16283 value

  -1.8V->0

or is it different ??

  • 0
    TI__Guru* 91947 points

    Himanshu,

    If you are using the ADC in unsigned binary format, a full scale input (1.95V) will give you 16,283 out and -0.05V will give you 0. If the format is setup for 2's comp, a full scale input (1.95V) will give you 8191 out and -0.05V will give you -8192. Is your EVM setup for serial or parallel SPI interface? Did you issue a hardware reset after power up? What is the level and frequency of your input clock? 

    Regards,

    Jim 

  • 0 in reply to jim s
    Prodigy 130 points

    Jim,

        Thanks for your reply. My EVM is setup for parallel interface and I am using CMOS outputs instead of LVDS and ADC in unsigned binary format. I have these questions based on your answer.

    1. As it is differential ADC , ADC value will be directly proportional to Vp-Vn. This voltage difference is instaneous right ??

    2. But there is only single ended input in the EVM which converts to differential inputs  through transformer . So basically ADC value is directly proportional to Vin (single ended input ) in EVM, right ??

    3. The biasing at the analog inputs is already done in the EVM using VCM (i.e.0.95V) output from ADC ic. So it is not necessary that sinle ended analog input of EVM should have DC offset 0.95V. Then why do you say i.e. 1.95V->16383, -0.05V->0 ??

    4. In terms of single ended input in EVM, it should be 2V->16383,0V->0.

     Kindly verify my above statements and report if any issue. Please note here I am talking analog voltages wrt  single ended input in EVM (not ADC ic).

  • 0 in reply to Himanshu Chaudhary
    Prodigy 130 points
    No I think , at single ended analog input in EVM
    1V (instantaneous)->16383
    -1V(instantaneous)->0
    Please check once.
  • 0 in reply to Himanshu Chaudhary
    TI__Guru* 91947 points

    Himanshu,

    Before you can run in CMOS mode, you must have made the following changes to your board:

    1. Remove RN3-RN6

    2. Install U7.

    Did you make these changes?

    If so, when you are running with the board in unsigned binary mode, with no input you should see 011111111111 (midscale) at the output. With a 2V p-p sinewave input, the output will go from 111111111111 to 000000000000. If you are just injected a 1V DC level signal or -1V DC level signal, this will not pass the AC caps and your output will be at midscale (011111111111) in both cases.

    Regards,

    Jim    

  • 0 in reply to jim s
    Prodigy 130 points
    Yes I made these changes. Also can you please what will be the setup and hold times values at lower frequencies.
    (i) 20Mhz (ii) 10Mhz (iii) 5Mhz
    As there are not described in datasheet.
  • 0 in reply to Himanshu Chaudhary
    TI__Guru* 91947 points

    Himanshu,

    Setup and hold times are with respect to the clock edge, not frequency, so these values are the same no matter what the frequency.

    Regards,

    Jim

  • 0 in reply to jim s
    Prodigy 130 points
    Jim,
    As specified in the datasheet, there are different setup and hold times for 65Mhz,80Mhz and 125Mhz. Similarily I am asking for 10 and 20 Mhz. Please Once have a look in the datasheet .
  • 0 in reply to jim s
    TI__Guru* 91947 points

    Himanshu,

    If you are still not getting any output data, send me a description of how all of the jumpers are installed. Are you pressing the hardware reset switch after power up? This is required for parallel SPI mode.

    Regards,

    Jim

  • 0 in reply to jim s
    Prodigy 130 points
    Jim,
    Thanks but now I am able to get correct outputs. But 5 or 6 samples out of 1000 samples are invalid. SO may be those 5 or 6 samples are fetched in the "data invalid time". That's why I am asking you about the setup and hold timings for 10 and 20Mhz because I am operating at 10 or 20 Mhz (i.e. slightly lower frequecies).
  • 0 in reply to Himanshu Chaudhary
    TI__Guru* 91947 points

    Himanshu,

    Set bits 4 and 5 to "1" in register 0xDF for low sampling mode. Place the ADC in test pattern mode and see if you still get these errors. If you do, they are several other registers that you can write to adjust the output timing and change the output clock strength. You will have to switch to serial SPI mode to enable these features.

    Regards,

    Jim  

  • 0 in reply to jim s
    Prodigy 130 points
    Jim,
    According to datasheet, ADS4142 is by default in low sampling mode. Right ??
  • 0 in reply to Himanshu Chaudhary
    TI__Guru* 91947 points

    No, by default it is set to 00.

  • 0 in reply to jim s
    Prodigy 130 points
    I switched to Low speed mode by modifying 0xDF register. But I am still getting intermediate noise samples. Can you suggest what further can be done.