DAC81404: Difference between error spec

Hi TTD,

Good day to you !

I am putting the questions again here along with answers.

Q1
Does "offset error" include "Offset error temperature coefficient"?

Also Does "Full-scale error" includes "Full-scale error temperature coefficient and "Gain error" includes "Gain error temperature coefficient" , ?

Ans : Yes,"offset error", "Full-scale error" and "Gain error" includes " does have a  temperature coefficient parameter in the data-sheet.

Q2

For example, consider Gain Error,
When outputting 5V at 25°C and drifting to 30°C (5°C) rise,

Gain error" is not applicable since the temperature range is TA = -40°C to +125°C. Instead, only "Gain error temperature coefficient ±2ppm of FSR/°C" only is applicable .

Is this correct?

Ans : No, "gain error" is always applicable (will scale depending on the code) for linear range of the DAC codes.

Q3

 "offset error" and "Offset error temperature coefficient" have following test condition.

Why bipolar range minus?

Why is it bipolar but defined as negative? How should the positive side be considered?

---------

Unipolar ranges, AVSS = 0 V Bipolar ranges, –21.5 V ≤ AVSS < 0 V

---------

Ans : No bipolar range is not negative, please look carefully; it's supply AVSS is negative.  DAC bipolar range implies that DAC output will have same voltage swing in positive and negative direction w.r.t. GND. And for this AVSS supply has to be negative (depending on the output range and footroom).

Q4 

What's the difference between  "offset error" "and  "Zero-code (negative full scale) error" ?

Ans : "offset error" is measured at the start of linear DAC code i.e. for 16-bit: at DAC code 512 for AVDD ≥ 5.5 V

and "Zero-code (negative full scale) error" is measured at zero code

Q5

Are  following figures typical spec?

Ans : 

These specs are MIN/MAX spec but plots are typical room temperature plots.

Thanks,

Sanjay

Hi Sanjay,

Thank you for your reply.

My actual voltage range was within ±5V. I mistakenly thought the DAC setting was the general ±5.12V, so I would not use 6V (20% of overvoltage).

In that case, is the following calculation correct? 

1. overall precision for 5V range will be          = 5V ± 0.07%FSR(=5V)

                                                                        = 5V ± (0.07 * 5V / 100) 

                                                                        = 5V ± 3.5mV

2. overall precision for 1V range will be          = 1V ± 0.07%FSR(=5V)

                                                                        = 1V ± (0.07 * 5V / 100) 

                                                                        = 1V ± 3.5mV

When dac output is -5V or -1V, the does TUE be  ± 0.05%FSR ? 

positive( ± 0.07%FSR)

negative(± 0.05%FSR) 

3. overall precision for -5V range will be          = -5V ± 0.05%FSR(=5V)

                                                                        = -5V ± (0.05 * 5V / 100) 

                                                                        = -5V ± 2.5mV

4. overall precision for -1V range will be          = -1V ± 0.05%FSR(=5V)

                                                                        = -1V ± (0.05 * 5V / 100) 

                                                                        = -1V ± 2.5mV

Best Regards,

ttd

Hi,

First two calculations are okay, but last two are not and FSR needs to be corrected.

3. overall precision for -5V range will be(min possible range is ± 5V and FSR will be 10V)   

                                                                        = - 5V ± 0.05%FSR

                                                                        = - 5V ± (0.05 * 10V / 100) 

                                                                        = - 5V ± 5mV

4. overall precision for -1V range will be(min possible range is ± 5V and FSR will be 10V)   

                                                                        = - 1V ± 0.05%FSR

                                                                        = - 1V ± (0.05 * 10V / 100) 

                                                                        = - 1V ± 5mV

Thanks,

Sanjay

Hi TTD,

There is no 25℃ number for offset and gain error (in general all the end pint errors). If you want to get it at 25℃, use the TC number and use back calculation either from 125℃ or -40℃ and take the worst case. 

P.S. - It's not recommended method, ideally one should measure the the offset/gain error at 25℃ and then apply the TC on top that.

Thanks,

Sanjay