DRV8353: Any idea how we managed to damage the device?

harald penz

Part Number: DRV8353

DRV8353SMRTAT

Application description:

The attached circuit belongs to a combustion engine powered battery charger. In this application the generator is also used as starter for the combustion engine.

For the current application the circuit is attached to a 12V lead acid battery. The motor current while starting is 40A – 50A. When the combustion engine is running at full throttle the battery is charged with up to 100A. Q310 and Q308 are needed to keep the charging current flowing through the external rectifier and not the body diodes of the H-bridge FETs.

The circuit is only powered for a few seconds, until the combustion engine takes over. After this the driver device and Q310 and Q308 are disabled.

We use 1-PWM mode for control.

PROBLEM:

If we use the circuit, in some cases, with no noticeable difference to other starting attempts, the driver device is damaged.

In the fault case, if the device is enabled it heats up to ~80°C. The commutation with the driver is still working, but the gates are not driven any more. In some of the error cases communication indicated gate drive under voltage condition.

If the driver is replaced the system is working again.

Note: The problem also occurred on an earlier version of the circuit where Q310 and Q308 where not present.

QESTION:

If you look at the attached circuit and the configuration, any idea how we managed to damage the device?

Would you change anything in that circuit?

Fullscreen 8353_spi_config.txt Download

Driver Control Register 0x02:
	Bit 5-6 -> 1xPWM_MODE
	Bit 4	-> 1xPWM mode synchronous rectification
	Bit 0 	-> CLR_FLT
	All other bits are zero.

Gate Drive HS 0x03:
	Bit 4-7 -> IDRIVEP_HS to 100mA (0x02)
	Bit 3-0 -> IDRIVEN_HS to 100mA (0x01)
	All other bits are zero.

Gate Drive LS 0x04:
	Bit 10	-> CBC, failt cleared after t_Retry
	Bit 8-9	-> TDRIVE, 500ns peak gate-current
	Bit 4-7 -> IDRIVEP_LS to 100mA (0x02)
	Bit 3-0 -> IDRIVEN_LS to 100mA (0x02)

OCP Control 0x05:
	Bit 8-9 -> DEAD_TIME 100ns
	Bit 6-7	-> OCP_MODE automatic retry
	Bit 4-5 -> OCP_DEG 8us
	Bit 0-3	-> VDS_LVL 1V
	All other bits are zero.

CSA Control 0x06:
	Bit 9	-> VREF_DIV half
	Bit 6-7	-> CSA_GAIN 5VpV
	Bit 0-1 -> SEN_LVL 0.5V
	All other bits are zero.

over 4 years ago

0 Matt Hein over 4 years ago

TI__Mastermind 21805 points

Hello Harald,

Do you know what pins were damaged? You can take a known good unit and measure the resistance to GND on each pin to check if any are different.

I do not recommend using a cut-off switch in the ground path (Q308). This can cause nodes like SPx, SNx, GLx to raise above their abs max value.

Thanks,

Matt

0 Matt Hein over 3 years ago in reply to Matt Hein

TI__Mastermind 21805 points

Hello Harald,

Any feedback on my questions?

Thanks,

Matt

0 harald penz over 3 years ago

Prodigy 10 points

Thank you for your reply.
Sorry for my slow response, but we had the customer in house and had to do some tests. While doing this test we damaged another chip. The customer was not happy.

We measured the pins and the only significant difference was on SOC pin. From this I guess your suggestion was right.

Now to our problem. We can’t remove the additional switch (Q308), because of the needed functionality to block reverse current through body diodes (up to 100A).
Do you think there is a possibility to protect the current sense inputs?
Or do we have to change to another driver chip?
Regards, Harald

0 Isaac Lara over 3 years ago in reply to harald penz

TI__Mastermind 23045 points

Hello Harald,

We will get back to you with a response tomorrow! Thank you for your patience!

Best,

Isaac

0 Isaac Lara over 3 years ago in reply to Isaac Lara

TI__Mastermind 23045 points

Hello Harald,

Thanks for your patience. You can implement protection for your pins so that they are not damaged but the protection should be done on the high side and not the low side drivers. If the changes are done to the high side of the driver then you can avoid removing the connection to ground as mentioned above by Matt.

Here are a couple of app notes that you can read on how to implement the protection on your board:

https://www.ti.com/lit/an/slva991a/slva991a.pdf

https://www.ti.com/lit/an/slva835a/slva835a.pdf

I hope these help!

Best,

Isaac

0 harald penz over 3 years ago

Prodigy 10 points

Hi.

Thanks for the hints.

We will try to implement the protection for the pins.

We hope that will solve the problems.

If not, we will return to this forum and ask for further help.

Regards, Harald

0 Matt Hein over 3 years ago in reply to harald penz

TI__Mastermind 21805 points

Thanks Harald!

Motor drivers

Motor drivers forum

DRV8353: Any idea how we managed to damage the device?