S12ZVC failure when CAN physical layer short to 24VDC and GND

[thulzh]

I am now testing the protection for misplug of our controller with S12ZVC.

On the controller, S12ZVC's output pin, CANH and CANL, of CANPHY is routed to the connector's pin.

One of the test we did is making CANL shorted to 24VDC, the norminal voltage of commerical vehicle.The first two time do not damage the chip and the chip was damaged in the third time. The phenomenon is that the CAN node did not transmit message and the chip heated up quickly while the other function of chip still worked for a while.

The reference manual of CANPHY declares that it is compliant for 12 V battery systems refering to ISO 11898-2 and ISO 11898-5  which needs the voltage bear up to 40V. The Appendix I 'S12CANPHY Electrical Specifications' also declares this electrical characteristics up to 40V.

The process i did this test is:1, switch off the main switch to cut the battery 2, short the CANL pin to the Vin pin with wires 3, switch on the main switch to supply 24VDC at the Vin and GND.

So my question is that do I misunderstand the electrical specification of CANPHY and it can not bear such high voltage? Anything can I do if I want to pass this test.

[lama]

Hi,

you understand the datasheet correctly.

The only I can do for you is:

- Contact your official distributor with warranty return

- please describe your tests (it is not clear what is the procedure of applying of the test voltage. It is good to have data from oscilloscope to be sure there was not transient state which caused V or I is out of the range. You apply correct voltage but transient event can cause damaging peaks.)

- Maximum ratings are not operational ratings. Maximum ratings mean that the MCU will not be damaged but functionality is not guaranteed. When you return to operating conditions the MCU should work correctly.

- You do not describe whether the MCU (CAN)  is able to work correctly after restart of the MCU in operating conditions.

We are very sorry the tech.sup. does is not able to solve such cases but if the MCU is not working correctly and you are sure you have not made anything wrong then you have to solve it via warranty return with official distributor where you both your device.

Best regards,

Ladislav

[thulzh]

Thanks for your suggestion.

- For the test, I connect the CANH/L to positive output pin of the regulated power supply. Then I turn on the power supply instead of turning on the power supply before connecting the CANH/L to positive pole. Sorry for the lack of oscilloscope capture and we also guess that it is caused by transient event.

- In the first 7 tests, the MCU (CAN) is able to work correctly after restart the MCU. In the 8th tests, the MCU (CAN) is not able to work correctly after restart the MCU. 

[lama]

Hi,

I am sorry for delay. I have tried to find any possible help....and it is everything in a theoretical level. However everything has led to:

- It is really good to see what happens during connecting of test voltage level because very short peaks in transient event could be destroyable together with cumulative effect. Without scope we can say nothing. (experience of a guy solving similar issue but on another device and peripheral)

Best regards,

Ladislav