TPS54360: Burned diode B560C

Hello Victor

B560C has reverse working voltage of 60V at 25degC. However this will change with temperature. It isn't sufficient for 48V operation as max voltage would go as high as 56V and ringing on top that would not leave any room from the peak working voltage. We have seen similar failure before, they were using same diode for 48V application. 

If diode fails to short, it would also damage device. 

I would recommend choosing 80V diode for this application

Thank you

Regards

Onkar Bhakare

Hello Onkar,

Thank you, for your prompt answer. I agree and there is one good replacement  " B390 " which is suitable to design. The issue is that I have to replicate failure before to proceed. This must be done somehow. In current design I lifted up the K pin from diode and soldered a wire to measure the current or spikes through it. In that case I added noise and is not the best way. I measured only in few cases 35 ns spikes of 0.4 A, on reverse voltage, usual are 12-13 ns.  

Hello Victor 

Could you please share the schematic and layout for better understanding? Additionally block diagram of your system would be of great help

Replicating the failure might be bit hard as there is good possibility that it has caused to higher stress level on diode for sufficiently longer time. You can repetitively turn on - urn of cycle of input voltage.

1. An electronic switch can be connected in between power supply and converter's input.
2. Keep input DC level at highest setting for application. 
3. Converter operating at full load
4. Run this system for few hours

Hope this helps

Thank you

Regards

Onkar Bhakare

Hi Onkar, 

The design was created around ten years ago. They respected the schematic which I started to review the models from Ti. https://webench.ti.com/appinfo/webench/scripts/SDP.cgi?ID=EDA94DA4BD03EFE2  the only difference is that on RT_Clk pin there is a capcitor of  10 pF paralel with the resistor. I took out capacitor without differences in measurement. The problem is that failure is a "exotic" case. It appearance in only few cases can't give us too much clue. Increase in voltage till 64 V on input for measurement around half hour for every step of 2V did not produce any failure. The increase in load with artificial resistor separate of device consumer produced short-circuit in IC Vin, SW and GND pins. I know that it is not easy task. It is my job to replicate it. For this I have to consider how much elements are involved around moment of switching ON high side mosfet. There must be something about junction diode and how this can be burned.

I omitted to say that the schematic I shared has a different IC. In working device is the TPS54360DDA.

Hi Victor

Did you try with different samples to account for part-to-part variations?

Thank you

Regards

Onkar Bhakare