Hi Jeff,
Please find below the answers I received today from our apps team.
1. There are two measurement topologies:
- Host MCU/ADC does not digitize Vbyp voltage (Vbyp not measured for every part).
- Host MCU/ADC digitizes the Vbyp voltage (Vbyp is measured for every part).
If the customer uses the second topology, the sensitivity will be independent of Vbyp variation, as they can calculate the «exact» sensitivity to use in their firmware by the equation for calibrated sensitivity. The part is "auto trimmed" internally, meaning the sensitivity is trimmed in order to achieve the target "VBYP/(3.276*gRange)" formula using whatever value Vbyp is on that particular device.
If the customer uses the first topology, they also have to include the Vbyp variation effect on the device sensitivity, which is ~ +/-3% according to the datasheet.
2&3. The FXLN83xx has 2 variants, namely the 2g/8g version and the 4g/16g version. I was able to confirm that the ‘calibrated sensitivity error’ listed for 2g and 4g in the data sheet also applies for the corresponding 8g and 16g configurations. The one item I am waiting for is to get test data to confirm whether it should be 6% for 4g ( &16g) case or if it is still 5% irrespective of g-range cases. Hopefully I can get back soon on this.
Best regards,
Tomas
