BQ79731-Q1: ADC accuracy information

Hello Oscar, 

1. Is there a plot of accuracy vs temperature and input voltage?
   1. There is not a plot of the input voltage vs accuracy at different temperatures accessible, however, the +/- 2.7 mV reading is the accuracy guaranteed through the -40C to 125C range. 
2. Is this 6 sigma?
   1. This is the 5 Sigma value
3. What is the minimum voltage that I can measure?
   1. 
4. Is there a way to reduce this error?
   1. Operating at voltages closer to ambient temperature produces the higher level of accuracy, however, as documented, this is the range of values. 

Hi Pranav,

They need to measure resistance from 0 ohm to 32 mega ohms, mapped on a input voltage from 0 to 5 volts.

The resolution is 500 ohm. So they would need an accuracy of ≈100uV.

The minimum voltage/offset is negotiable and not decided(ideally according to requirement 0V) however accuracy is a big concern.

Do you think that through calibration they can reach this target?

1. Target accuracy is actually as less as possible (in the range of <0.5mV at least)
2. Minimum voltage on the ADCs is 0V due to the current concept
   1. Would you recommend a workaround by using a level shifter (offset source)? --> Is there info material on such a concept? 
3. Could you please let us know the accuracy distribution at different temperatures?

1. 1. (-40°C, 0°C, 25°C, 45°C, 85°C, 125°C) --> Whatever is available asap.
2. What is the background of the input voltage dependence?

1. 1. Is there a different gain used within the BQ at different input voltages?
   2. If yes, please let us know the different thresholds (gain vs. input voltage) and impact on the accuracy
   3. Challenge is much higher on very low input voltages!
2. Could you please let us know input voltage and temperature impact also for EOL  (Lifetime drift)

1. 1. We have EOL data from Marc Winkelgrund, which are also for high range of input voltages and temperatures
   2. The drift are shown after 1000hrs test. --> Is there a saturation or do you expect this drift every 1000hrs?

I am quite sure that your device is much better than stated in the datasheet and also our application for BMS with much more relaxed in regard to temp. range. than -40°C to 125°C

I would like to show our customer different accuracy ranges, to cover high accuracy at important operating modes, e.g.:

1. Use case Driving: 0°C to 45°C --> high accuracy range
2. USe case Charging / Off/On grid parking: -40°C to 0°C and 45°C to 85°C --> more relaxed accuracy range

Is there the possibility to calibrate this ADC error at 0V in our production?

• Is this random or systematic error specific to a single part?

I think it would be best to discuss this in call.

I will send you an invite for this morning

Thanks

Oscar