• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 46 subscribers
  • Views 602 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.

ADS5562: Resolution computation

Fabrice
Prodigy 10 points

Part Number: ADS5562

Tool/software:

Hello,

I am using the ADC ADS5562 and I don't understand how to compute the resolution (V per 1LSB).

In the datasheet at page 29 we have a tab with different gain and the corresponding full scale but no formula.

Is the resolution equal to : Full_scale_range / 2^16 ?

Moreover the other gains do not match with the full scale voltage.

For example if I take the default gain (1dB) which give a full scale of 3.56V so as to compute the full scale with the 2dB gain, I have 3.56 x 10^(-1/20) = 3.17V and not 3.2V.

So, to compute the resolution, must I use 3.2V/2^16 or 3.17/2^16 ?

It is the same thing with other gains without the 6dB gain that give me the same result as the datasheet.

It is thank to this method that I can retreive results in the example for the 0dB (page29) : resolution = 3.56 x 10^(1/20) / 2^16 => so, for 3.56V input, I get 58409 like in the datasheet.

And I wonder if setting the data format in 2s complement or in offset binary in the register 0x63 affects the resolution ?

Thank you in advance for your responses

Sincerely

  • 0
    TI__Mastermind 45670 points

    Hi Fabrice,

    The fullscale of the ADC should reflect 0dBFS. However, it seems the datasheet, on page 29 reflects, -1dBFS numbers below fullscale or 3.56Vpp.

    Changing the output data format shouldn't have to do with any change in fullscale settings.

    And, yes, correct, to your original question: Is the resolution equal to : Full_scale_range / 2^16 ?

    Hope that helps.

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    Prodigy 10 points

    Hi Rob,

    Thank you for your response.

    In my application I want to use the 2dB gain.

    In the datasheet p29, we have a full scale of 3.2V. However, in my previous message I get a full scale of 3.56 x 10^(-1/20) = 3.17V.

    Which one is correct ? Is there any rounding ?

    So is my resolution 3.17/2^16 or 3.2/2^16 ?

    Regards,

    Fabrice

  • 0 in reply to Fabrice
    TI__Mastermind 45670 points

    HI Fabrice,

    Please understand this is an older device, so I would have to measure this myself in the lab.

    I am happy to do that but it will take some time. I looked all over the lab and we don't have an ADS5562 EVM on hand.

    So I need to order one from stock and then set this up and measure it for you.

    Therefore, the earliest I can get you an answer/confirmation is late next week if okay.

    Please let me know.

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    Prodigy 10 points

    Hi Rob,

    Ok I understand and I am interrested inyour results even if it is for the end of the next week.

    Thank you,

    Regards,

    Fabrice

  • 0 in reply to Fabrice
    TI__Mastermind 45670 points

    Hi Fabrice, 

    Quick update, I had to order the EVM from stock and I received it this morning.

    I will set this up in the lab to confirm this week.

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    Prodigy 10 points

    Hi Rob,

    Good new !

    I look forward to your results.

    Regards,

    Fabrice

  • 0 in reply to Fabrice
    TI__Mastermind 45670 points

    Hi Fabrice,

    We tested this in the lab and we concluded the ADC has its own errors and will never be exact.

    After talking with design, there will be some slight differences in results vs your calculation, the fullscale value in the datasheet should be treated as a soft value as keep in mind the input fullscale will vary do to offset, gain and internal reference errors.

    To help with this, if its important to you, then I would advise to using an external reference. But even then, you could get 1mV or 2mV deviation.

    Regards,

    Rob

  • 0 in reply to Rob Reeder
    Prodigy 10 points

    Hi Rob,

    Thank you very much for your reply.

    I have also received my electronic card and I have carried out a few tests. I first performed an offset calibration and injected a signal into the ADC.

    My conclusion is that it is more accurate to use 3.2V rather than 3.17V that I have found with my formula.

    So I think we should use the value from the datasheet even if I do not understand where it comes from...

    Thank you for your help

    Regard,

    Fabrice