This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BQ77910A Implementation - detailed schematics

Other Parts Discussed in Thread: BQ77910A

Hello,

I am developing with the BQ77910A chip and have ran into some configuration issues.  In the data sheet the schematics are slightly ambiguous when it comes to the grounds.  The configuration is the same as Fig13. "Example Series FET Configuration Using the CHGST Pin".   It seems that the PWR (PGND) ground and REF (AGND) ground make a floating ground relative to the other parts of the system.  When communicating with the chip do you make a common ground with PGND and ones Digital ground?  I have made a quick block schematic of the power loops that are being dealt with, see attached.  I was seeing a large voltage difference between my digital ground and PGND so I wanted to verify the proper I2C implementation before attempting.

Thanks for any help.

3463.BQ77910A Logic Reference.pdf

  • The figures in the datasheet are often simplified to represent a concept without excess detail.  Please see the EVM schematic or associated application notes where available for more complete schematics.  Figure 13 shows a concept with series FETs so that there are only 2 power terminals, and a third terminal to tell the battery protection circuit (bq77910A) when it is in the charger.   The grounds for the battery protection circuit are shown as a REF (AGND) node which is tied to the PGND at one point and basically used for the low level SENSE filtering and the bottom of the part.  The other 'grounds' for the battery circuit are shown to the PGND, still at the negative of the cells and still desired at the same point as the AGND tap, but perhaps noisier. The rest of the world would be connected off the right side of the figure, its 'ground' reference would be PACK-.  The battery circuit grounds are protected from this external world ground. 

    The bq77910A is a stand alone protector and most simply switches low side FETs as shown in the figure, so you can think of it as the ground connection going away.  In a removable-battery system with the battery plugged into the load, the system does not know if the power was switched or the ground was switched, if the protector turns off the PACK terminals, the external system is off.  With the removable battery inserted in the charger, CHGCTL is likely connected through a resistor to PACK+ in the charger.  Again if the bq77910A protects, the charger has lost its load and is off.  Some care is needed in the charger design since while the diode shown in figure 13 on CHGST may prevent damage to the pin, it provides an unprotected charge path through the 1K resistor to CHGCTL.  So with this concept, CHGCTL should not be driven to the charger negative, and if someone applying an external voltage between PACK+ and CHGCTL, some other circuit such as a blocking diode between CHGCTL and CHGST may be needed.  In a system with a non-removable battery it can be more complicated. 

    If you have additional circuitry for monitoring the battery you must decide how to power that and where to put the protection switch or switches.  If that circuit is trying to monitor signals from the bq77910A, it will need the (Vss) reference for the chip or an isolation circuit. for the signals.  If that circuit is inside the battery, consider if the battery needs to be protected from it either from malfunction (over current) or to reduce its current in an undervoltage condition.  If the circuit moves away from the battery, consider if the connections could be shorted and circumvent the protection switch (FETs) between PGND and PACK-.  It is possible to use the CHG and DSG signals from the bq77910A to control high-side switches. This allows a common 'ground' or equal potential grounds in the system as we are accustomed to seeing, but it takes more circuitry.

    In your diagram, I'm not clear where power is switched.  The 'figure 13' concept would put the switch inside the Lithium Cell Module block, but the normal function of the bq77910A does not support I2C communication in-system.  The I2C is a configuration and debug interface and is disruptive to the part. Enabling the interface requires setting zero delay mode in which the part can respond to noise and indicate a false fault. Normally the signals would not be connected in-system.  However if your system design tolerates this operation, it would seem the Lithium Charger and  DC to DC blocks should have a different ground than  PACK- and be upper blocks would be referenced to the battery negative.  If the grounding outside the battery is as shown with the FPGA referenced at the charger negative, then isolation would be needed in the communication interface between the FPGA and the bq77910A since the protector ground will move with respect to PACK- when the low side (figure 13) protection switches open.