TPS562201: Overcurrent protection is failing

This is the circuit in question. When I short the 5.4V rail directly, the IC behaves as expected and current limits - the output goes to 0V. However, when I short circuit this DC output after the protection mosfets, the IC blows up! 

So something in the protection circuit is destroying the IC but I can't figure out what. DC socket allows 5V INPUT and 5V OUTPUT. The 5V input goes to a separate charging circuit. The mosfets are there to protect against reverse polarity and over-voltage inputs. 

I'm going to try some different compatible parts and see if they behave the same. But could really do with some help on this. 

thanks

David

Hi,

Could you please share the waveform during DC socket shorted? (VIN, VOUT, iL, SW)

And for the pin to pin IC:

TPS562208, TPS54202/H

Shawn

I've done some more testing and I believe I know what is happening. This is the waveform when I place a 0.55ohm short onto the DC output. 

When there is short, the cycle by cycle current limit activates and the output voltage drops. Maximum output current it reached - around 3A-4A I guess. The device is not going into hiccup mode because the output has not fallen below UVP (3.47V in this case?). I think the p-mosfet Q1 starts turning off as the voltage across the gate drops and its resistance increases. The current goes down and the voltage rises again. It quickly balances giving us this situation where the device is outputting a large current of around 4.5A it seems (2.5/0.55). In this test, I have set it to turn on for 100mS, but even this short burst is enough to destroy the part. 

As you can see, after the load is removed and the device turned off, the output rises high (I guess to Vin?) so I assume that the high mosfet has blown to a short circuit. I quickly turned off the supply as the TPS562201 went up in a puff of smoke. 

Here is another PIN for PIN compatible IC from a different manufacturer. https://www.diodes.com/assets/Datasheets/AP65211A.pdf. It seems to survive the same test. (I just replaced the IC after it blew up above). The output current is regulated by monitoring the current of the high side driver which limits the output to around 2.2A it seems. Again, it doesn't drop low enough to enter hiccup mode but seems happy with 100mS of short. 

I tested again with a 0.06ohm short and it still survived. I'm sorry to say that I plan to now use this part instead. 

In the meantime, I have implemented software protection to turn off the load within 10mS due to an overload which is enough to save the TPS562201. 

Best regards

David

Hi,

Could you please share the layout? Poor layout will increase the risk of damage during hard short test. It would be better to put the input capacitors close to VIN and GND of IC and add a 0.1uF ceramic capacitor.

Shawn

I'm afraid things are very tight on the board. Vin is top right so the input cap is right across the pins. Vout caps are not so ideal but I don't think the layout is a problem here. It's just the way the OC is implemented and the fact that the IC is not seeing a true short and not triggering hiccup mode. 

Cheers

David