BQ4050: GPCCEDV: Accuracy Error: Deviation is so high that it is most likely due to anomaly in the data

Hello Tom,

The CEDV calculation tool is for temperature compensation. Results will not show the benefit of the tool unless the gauge and the battery are put in a chamber. 

For IVT data collection, please visit this app note Collecting IVT Data for Testing.  

If temperature is not a concern for the application, the gauge can be configured for fixed EDV points. Essentially, EDV0, EDV1, and EDV2 (usually at 7%). Please refer to section 6.4  End-of-Discharge Thresholds and Capacity Correction from TRM

Regards,
Jose Couso

Hello Jose,

thank you, I will try the fixed EDV approach to see if it suits my needs.

I know that CEDV mode compensates for temperature and load. My plan was to familiarize myself with the procedure of collecting CEDV parameters before getting access to a chamber. I thought I could use one log file for all 6 measurements leaving the CEDV online calculator no room for compensations. I was just wondering why the SOC errors ​​differ so much from each other since I could not see any significant differences between the two measurements.

For the fixed EDV mode the TRM states:

"Fixed EDV thresholds may be programmed in EMF/EDV0, EDV C0 Factor/EDV1, and EDV R0 Factor/EDV2 in
mV."

1. Could you please explain what parameters I should program where, only EDV0/1/2 or also EMF, C0 and R0 in bqstudio? If so, how do I determine EMF, C0 and R0?
2. Is it sufficient to set EDV_CMP=0 in order to use the fixed EDV mode?
3. If I want to determine EDV0/1/2 based on the packs's discharge cycle log I probably want to set EDV_PACK=1 later in order to program these thresholds. Which parameters are affected by the EDV_PACK bit only the EDV0/1/2 thresholds (cell- or pack-based) as well as EMF, R0 which have to by multiplied by the number of series cells, or, do I need to adapt more parameters?

Kind regards,

Tom

Hi Jose,

thank you for the clarification and your detailed answer.

But as you said EDV1 and EDV2 are typically calculated by coulomb counting that's why I think you meant that EDV0/1/2 are determined at 0/3/7 % of RSOC and not cell voltage.

1. Do you have by any chance any documents/material which compares Fixed EDV with CEDV and outlines the benefits of the CEDV method w.r.t to the RSOC error depending on temperature and load for different types of batteries?
2. I would like to know how the bq4050 gauge would behave on discharge if I determined the EDV0/1/2 thresholds at room temperature, but later in the field the ambient temperature is let's say 20-30 °C higher or lower (but more or less constant). I know that the pack's capacity will decrease or increase depending on the temperature. Is my understanding correct that given a certain learned FCC depending on the temperature the EDV voltage thresholds may be reached earlier resulting in a jump of the RM or later in which case the RM is held for VDQ=1 or not held for VDQ=0 as described in the TRM (section 6.4)? Since fixed EDV thresholds have been determined for a specific pack's capacity the RM correction will therefore be inaccurate. However, the RM correction is obfuscated if a new FCC is learned for the changed capacity of a pack. I know the exact behaviour is dependent on many factors and that there is not definitiv answer, I just want to get an idea which behaviour I could expect.

Thank you for your effort.

Kind regards,

Tom

Hi Jose,

thank you for the clarification.

I think there is no getting around it. I need to evaluate the accuracy of Fixed EDV for my setup in order to be sure that accuracy is acceptable for my application.

Kind regards,

Tom

Hi Tom,

Sounds good. Please let me know if you encounter any problems. 

You can always open new threads if it is a different topic.

Regards,
Jose Couso