2
\$\begingroup\$

I have a 3-bit converter with ideal LSB of 50 mV. The following table shows the non-ideal logical values.

I need to remove gain and offset errors from the codes below.

I read on the internet that one step to remove gain error is to calculate a new LSB, by doing an average:

AVG_LSB = (363 mV - (-2 mV)) / (8-1) = 52.14 mV

If I calculate the table using this new LSB (NEW gain) I get table 2 shown below. Did it remove the gain error? If so, then why is it called removing gain error?

enter image description here

enter image description here

\$\endgroup\$
3
  • 1
    \$\begingroup\$ You just did a two-point linear regression. You could be using all 8 measurement points instead to achieve lower linearization error. Please read on linear regression \$\endgroup\$
    – Vicente Cunha
    Commented Sep 15, 2020 at 11:52
  • \$\begingroup\$ I talking about gain error in ADC if its similar to linear regression then cool but i am asking about the gain error \$\endgroup\$
    – rocko445
    Commented Sep 15, 2020 at 11:59
  • 6
    \$\begingroup\$ If you're trying to explain, through a fitted line, the ADC counts versus input voltage, then this is by definition a linear regression. In the terminology you're using, "removing ADC gain error" = fitting the most appropriate line slope; "removing offset error" = fitting the y-intercept \$\endgroup\$
    – Vicente Cunha
    Commented Sep 15, 2020 at 12:09

1 Answer 1

Reset to default
2
\$\begingroup\$

A old but good application note from TI: Understanding Data Converters

https://www.ti.com/lit/an/slaa013/slaa013.pdf

They show the dominant errors in ADC's.

Most of them you can measure and compensate with a Lookup Table, Interpolation or simple multiply/add operations... but you will loose full scale range in most/all cases, so I guess your fun with correcting a 3bit Converter will be limited.

\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.