TPS54531: Overcurrent Protection (OCP) Operation

When the OCP, it will decrease the frequency until the OCP go away. So when the OCP go away , the IC will be enable again and the output will recover.

 The same as the thermal protection and the OVP

Does "OCP go away" mean it essentially turns OFF the output? I modeled this in TINA using the TRANSIENT model. I saw the frequency decrease in powers of 2 just like the datasheet says, but the output tried to keep up.

In another (not TI) SMPS, the OCP feature will cause the chip to reset. It will stop driving for a "cool-down" period, then it will restart the soft-start. In the meantime, the output droops below the voltage monitor reset voltage, and the system restarts.

In the TINA model, the TPS54531 keeps driving the output until the bulk capacitance is charged, then it returns to normal.

It's as though the TPS54531 TINA model turns into a constant current source while it charges the bulk capacitors, then returns to a voltage source after the inrush. I can understand that if this high inrush lasted a long time, the switch junction temperature would increase and eventually trip the over-temperature protection.

Further, section 8.3.9 talks about the frequency being divided by 1, 2, 4 and 8 during OCP; however, it does NOT state that the frequency is ever stopped.

OCP go away means no overcurrent during the operation.

The frequency will be decreased but not stop. And if the  junction temperature increase and trigger the OTP(over temp protection), the IC will turn off.

So how long of the overcurrent will you have? And what's the current(OCP) ?I'll help you to check it on the bench.

Hi 

When the OCP ,the IC still work and the output will be pull-down, the frequency will decrease.

When the IC trigger the OTP, the IC will shutdown. 

If the time of OCP will be long ,the temo of IC will increase. Besides, the diode and Inductor will be very hot, even damage.

The details data you TPS54531.zipcan see the attachment. Thank you!

I used the TINA Transient model to simulate this. The event lasts for 1-2 msec. I expect this to occur once every 0.5 to 1 hour, but no more than 20 events a day.

It seems to me that the OCP will kick to limit the frequency/output current, but the IC will only turn OFF if the over-current condition persists long enough to elevate the switch junction temperature causing an OTP event.

Do you have any data/experience to say if the 1-2 msec OCP occurring 20 times a day will cause any degradation in performance? WEBENCH calculations and thermal model suggest that at +40degC and 3A out, I have 50 to 60degC of temperature margin to Tj max (+150 degC).

I'm willing to accept your answer as opinion or, at best, an educated guess. Of course we will test this before releasing this design. I'm looking to answer the question of: "Will the IC stays ON, and the short burst of high-current will hardly affect the thermals?"