Initial high failure rate of ISO1050DWs. Any suggestions?

Hi Charlie,

This is totally unexpected and unacceptable failure rate.  We have 100% production test for all devices before they go to customers.

Can you please provide details about this issue such as:

Appliction:

Lot Trace Code (Physically written on topside of each device):

Details of the failure:

Tests Conducted:

# of Units tested:

Please send directly to me at marwat@ti.com

Hi Marwat,

This did not answer my question on preventing damage to your ISO1050, possibly from high static charge.  Of course you thoroughly (100%+) test your parts and do not expect failures.   If I have future failures, I will forward them to you but I do not want to commit to a separate board spin to fix this if I can prevent damage by adding protection circuitry or special handling methods.

The failures are in isolated area at the periphery of large boards (more than 1 square foot) where the isolators are along the periphery of the board and are floating with respect to the rest of the board.  External clamping diodes are present to prevent either of the differential signals from exceeding the positive rail of the isolated 5V supply or going below the ground of the isolated supply by more than a Schottky diode drop.   Isolation from the rest of the board is required for patient safety and the circuit can't degrade the behaviour of 5KV peak defibrillator pulses delivered to the patient.  These pulses will couple to some extent onto the differential CAN signals, power and ground, despite patient isolation.

Thanks,

Charlie

A common protection scheme against powerful surge transients, EFT and ESD is given below. Usually folks use a TVS diode and think that is the end of the problem. It's not because you need a high-voltage cap to divert the trsneint enegry to Eart potential.

The TVS is typically a 400W device (i.e. SM712 from Bourns or similar). The HV-cap is from NOVA. The common-mode choke from Vishay I believe. R and R2 are either 10W MELF resistors, or Pulse Proof, High Power Thick Film Chip Resistors.

Regards, Thomas

Hi Thomas,

Thanks for your insights.  I believe together these components will provide the required protection.  Can you attach the schematic or topology or netlist as well to confirm my understanding?

Regards,  Charlie

Charlie,

what I sent you is the schematic. I don't have a netlist. Do you require some component names?

By the way I am an RS-485 guy, our CAN-man is currently on travel. I will have to ask him for the choke he used.

The TVS diode, High voltage cap, and the pulse-proof thick film resistors I will need to search for on my laptop.

I probably can redraw a more clear schematic. Do you need this just for the CAN transceiver or for the supply too?

Regards, Thomas

Hi Thomas,

I do not believe I ever received the schematic, just your list of parts.

Can you resend the schematic or just a crude connectivity diagram?

tnx,

No problem Charlie.

It was a copy of a copy I did send and still see in front of me. Perhaps the forum does n't like this. I'll make a clean start and send it to you.

regards, Thomas

Hi Thomas,

I think this will work for me with changes to the breakdown voltages (e.g., capacitor breakdown voltage ~ 6KV) to not be destroyed by defibrillator pulse and provide proper isolation. 

tnx,

Charlie