BQ24618: Plugging the battery in before powering the IC

JPAlleyne

Part Number: BQ24618

Hi,

I'm still investigating, but it seems as though anytime I plug the battery in first I need to replace the Low-side NMOS. Is this something you have seen in the past?

over 5 years ago

0 Jing Zou over 5 years ago

TI__Mastermind 19430 points

Can you please show scope capture on the issue when damage occurs?

0 JPAlleyne over 5 years ago in reply to Jing Zou

Expert 1565 points

Before I do that, I was wondering if you could explain step by step what should happen when you plug a live battery before you apply input power. Maybe I can try to fix the problem before I blow another board.

Thanks,
JP

0 Jing Zou over 5 years ago in reply to JPAlleyne

TI__Mastermind 19430 points

What is the cell count of your battery? What is the voltage rating of the low side mosfet?
When battery is plugged in, the device won't power up until VCC is powered. In battery only mode, the converter should be off and the part will not switch.
Do you have battery FET? Could you please upload the schematic?

0 JPAlleyne over 5 years ago in reply to Jing Zou

Expert 1565 points

Cell count= 6

FET Voltage Rating = 40V

Here is the Schematic (we added the missing cap on SRP):

0 Jing Zou over 5 years ago in reply to JPAlleyne

TI__Mastermind 19430 points

Since there is no BATFET in your schematic, when battery is plugged in, a large current would go though the inductor to the body diode of Q3 then to the ground. There will be a voltage spike on switch node. However, not sure why Q4 get damaged but PH pin did not.

0 JPAlleyne over 5 years ago in reply to Jing Zou

Expert 1565 points

Well I haven't fully tested it but initial findings showed that there is the potential that the IC got damaged as well. You said that a large current would flow through the body diode of Q3 and then go to ground.

1) What path would it take to go to ground through Q3's body diode?

2) Is the voltage spike caused by the large current flowing through the inductor?

3) Is it the voltage spike that causes damage to components on the PH node?

4) What can I do to protect the circuitry right now, without major redesign?

0 Jing Zou over 5 years ago in reply to JPAlleyne

TI__Mastermind 19430 points

Regarding 1, please see the path marked below.

Regarding 2, yes.

Regarding 3, I cannot conclude without look at scope captures.

Regarding 4, when battery is plugged in, the drain of the lowside mosfet rises fast which might bring up the gate voltage and cause damage. One solution is to add more capacitance at Cgs to hold gate voltage down. Or try to find a replacement of the low side FET which Cgd is less than Cgs.

0 JPAlleyne over 5 years ago in reply to Jing Zou

Expert 1565 points

So did you mean Q4? And not Q3?

0 Jing Zou over 5 years ago in reply to Jing Zou

TI__Mastermind 19430 points

0 JPAlleyne over 5 years ago in reply to Jing Zou

Expert 1565 points

I'm still a bit confused as to why the current would flow through the capacitor to ground. Isn't it a large DC current?

0 Jing Zou over 5 years ago in reply to JPAlleyne

TI__Mastermind 19430 points

You can consider that the current is going to charge the input cap. It is a short current spike when battery is plugged in.
The issue here is before battery is plugged in, the switch node voltage is 0V. When battery is plugged in, switch node voltage rises. In this process, Q4 gate might be turned on if Cgd is charged up, and it might cause Q4 damaged.

Power management

Power management forum

BQ24618: Plugging the battery in before powering the IC