Good Day
We are using the BQ76PL455A to manage LiPo batteries of a car-like device. The setup of the daisy chain is as in Use Case scenarios page 3, Figure 1.
The whole system works reliably up to roughly 400V and from 400V upwards the problems begin. A precharge for the inverter can be conducted successfully but the moment the inverter starts switching, voltage levels on the communication bridge PCB start breaking down. This doesn't always happen. Whenever this happens, the first PCB in the daisy chain breaks. All the capacitors and resistors are still intact afterwards, but the BQ76 chip seems to be broken. The assumption is, that the low communication peripherals break through an overvoltage pulse on the communication line, or an induced current.
The Communication bridge communicates with a microcontroller, which also controls the relays (high side, low side, precharge). The microcontroller/communication bridge is isolated from the rest of the electric circuits with isolated DCDCs and optocouplers.
We think there is a potential/ground problem, as the isolated Side of the Communication bridge, if looked at with the oscilloscope it roughly lies between HV+ and HV-, but jumps by several 100 volts with exactly the inverterfrequency. This potential difference probably is the reason for the broken BQ chip. As the potential difference seemed to be a problem, and the communication bridge floating, we connected the communication bridge ground to the first PCB in the daisy chain. This made the potential problem worse.
Is there a reason that would make the connection of the communication bridge gnd and first BQ chip gnd impossible? How can one get the potential of the communication bridge to a correct level for the BQ chip not to break?
Thank you very much for your help.
