CAN Transceiver Damaged while surge tests

[ku637]

Hi,

I've been using a CAN transceiver AMIS-30660-D in a prototype .


The power architecture is somewhat like


24V---> Buck Converter-->5V---> Powers CAN transceiver


The unit when tested with surge levels of 1.1kV in the input power lines i.e between SHIELD EARTH (not 24V GND) and the 24V positive line , then the CAN transceiver gets damaged.

There is only a TVS diode protection between CANH/CANL lines and the 5V GND.

No protection between Shield EARTH and either CANH or CANL or 5V supply to the CAN transceiver


Can some body with prior experience can suggest whether this surge will be possibly affecting the Vcc pin or CANH or CANL pins of the transceiver??


Thanks for any help,
Regards

 

[schmitt trigger]

We need to know more how the different earths and commons relate to each other.

For instance, does the shield earth connects to 5v common? I not where does it return? And I assume that you are using a non-isolated buck, such that the 24v and 5v commons are the same?

Is there another earth, like chassis earth?

How tho they relate? Do they have capacitors and/or inductors to couple them?

How does the CAN transceiver fail? A short or leakage on any of its pins, or is it a functional failure (i.e. the device talks but not listens).

Was the unit requires to be powered up for failures to occur? If powered up, does the device start getting warm after the strike?

The 1.1Kv surge requires clarification also......to which regulatory standard are you testing? What is the impulse waveform characteristics?

Other thought...have you sent the device to the IC vendor for analysis, to see which section is damaged?