One of our products contains a 4-cell-in-serie battery pack including a BQ20z75 + TI suggested FET protection circuitry.
The BQ20z75 is configured to make the pack acts as non-removable. The Design capacity is 680mA, 4 batcells LION 3,6V in serie.
The modules are allways tested before shipment (Impedance Track is allways enabled), and some of them are already more than a year running properly. All of them have the same Data Flash image and initialisation procedure
In some cases, after some time, and under unknown conditions the Charge FETs are OFF and don't allow the battery cells to be charged, it happens on both Charge and pre-charge FETs. During this state, the battery cells are within valid ranges and there is no over temperature, over voltage, permanent failure or whatsoever flag. BQ20z75 informs that the Charge FET's should be conducting but they are not, there is no current flow reported by BQ20z75 at all.
The problem has sporadically (not allways working method) disappeared if a reset order 0x0041 is sent to the BQ20z75, or if the cells are discharged until under voltage condition, or if the FETs are turn OFF and ON manually using I2C commands.
The Discharge FET has been never a problem, as soon as the battery cells where over the programmed threshold, the Discharge FET has allways worked.
We have searched in TI gas gauge support forum for a similar case and we found this one: "BQ3060 - Strange state with no errors but FETs off" which is not really solved. Have you got more experience on such ERROR state?
looking forward your response
ramiro
PS: We have also seen that many users work with BQEASY and use the log files and reports from such Software as info to help you to find out the causes of the problem. When our battery pack enters this error state, we received the system from our customer and have no chance use BQEASY software. Would be the Data flash raw data as useful as BQEASY log files are?