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.

DAC9881: Figure.8 in DAC9881 Evaluation Module ( SLAU279A )

Part Number: DAC9881

Hello, 

I'm confusing about SLAU279A 's figure 8.

Since DAC9881 is designed as bufferd output mode, so internal buffer may lead offset voltage and headroom voltage.

They may crash the linearity of DAC. My problem is:

Q: When analysis the linearity, how to identification the non-linearity due to buffer's offset voltage & headroom voltage ? Is there are some definition about them ?

Thank you!

BR
Wen

  • Hi Wen,

    Thank you for your query. The end-point errors (zero-code error, offset error, gain error, full-scale error) are adjusted before calculating the INL. Hence these errors don't show up in the linearity. The voltage headroom is taken care of by selecting the appropriate value for the power supply.

    You can look at figure 6 of this document for better understanding of the INL measurement:

    www.ti.com/.../slaa013.pdf

    Hope that answers your question.

    Regards,
    Uttam Sahu
    Applications Engineer, Precision DACs
  • Hello Uttam:

    Thank you for your reply.

    Chould you explain in detail how to select the appropriate value for the power supply about the voltage headroom ?

    Is that judged from vout( T+1 ) - vout( T ) < ideal VLSB or some factor to mean vout enter in non-linear region ?

    Thank you!

    BR
    Wen
  • Hi Wen,

    DAC9881 is a rail-to-rail output DAC. I missed that in my last post. Hence, there is no requirement for power-supply headroom for this device.

    Regards,
    Uttam
  • Hello Uttam:

    Thank you for your reply.

    How about the internal buffer's headroom?

    I mean why SLAU279A 's figure 8 choose '261120' as final test code but not 2^18 - 1 = 262143 ?

    Thank you!

    BR
    Wen
  • Hi Wen,

    You pointed out correctly. Even though the rail-to-rail output corrects the headroom issue to a good extent, it doesn't push the linear range to both rails. Hence, there will be some non-linearity near the rails. The INL curve shows only up to the full scale - 43 codes. That means the INL will become larger than specified. There is no quantitative data at the moment, unfortunately. This range should get extended by providing the same headroom in the power supply.

    Regards,
    Uttam