DAC7571: Question of DAC7571 PSRR and INL Specifications

Hi Wayne,

Thank you for your query. The INL is specified as Linearity Error (LE) in Fig 1, 2, and 3.

Hope that answers.

Regards,
Uttam Sahu
Applications Engineer, Precision DACs

Hi Wayne,

Most of our recent devices will directly provide a measured TUE figure at room temperature and over various operating temperature ranges, but for older parts which this was not measured for the root sum squared approach you have outlined is appropriate. Generally we apply this to INL, offset (sometimes aliased as zero-code error depending on the datasheet), and gain errors which apply over the linear region of operation of the DAC to represent the majority of usable situations.

I wrote a blog post on this topic some time ago you can view here:

Relative accuracy is another term for INL, which is listed in the datasheet on Page 2 as 0.195% maximum.

So your RSS equation here would apply as Sqrt(0.195^2 + 1.25^2 + Offset^2) - In order to finish the calculation we would need to know the VDD you plan to use as the zero code error is specified in millivolts, so we need the full-scale range in order to calculate this error contribution in percent.

Assuming a range of 5V the TUE would be 1.32% using the maximum specifications for zero code error and gain error. If you substitute the zero code error maximum for the typical, the value is 1.27%.

Keep in mind the top-line conditions for the Electrical Characteristics table:

This means that the maximum INL, Zero Code Error, and Gain Error specifications used in this calculation are all over the complete operating temperature range, not the initial accuracy at room temperature.

Wayne,

It is very common for relative accuracy to be used as an equivalent term to INL in DAC datasheets - even from other manufacturers outside of TI.