TPS1641: blow up when connecting 24V power supply at the input.

Hello Samuel,

What's the total output cap and the starting load?

At what voltage downstream converter turn ON ?

Thanks 

Amrit 

Hello Amrit,

The output cap is 80uF (you can see it in the schematic Alimentation.pdf) and the 33nF cap on dVdT pin limit the inrush current at 250mA.

The load is an i.MX8 based SoM (including PMIC, DDR, eMMC and a phy Ethernet), the current during startup (after the inrush current) is about 300mA.

 The enable pin of the buck is at 1V and there is a soft start of 4ms.

Thanks

Samuel

Hello Samuel,

In schematic I can see two TPS1641. which one is getting destroyed?

Is it U22? Can you place output cap near to the pin? This helps in slowing down the current slew if there is short circuit.

Can you help with waveforms to analyze? We can have a online meeting if you need urgent debug

Thanks

Amrit 

Hello Amrit,

I think I have solved the problem in practise by adding a TVS at the output of the common mode choke (just before the two eFuses with a clamping voltage below 42V which is the maximum absolute rating for Vin).

From my understanding, at the startup during a hot plug, there is a big inrush current caused by the 80uF load and when the eFuse sees this current (>10A), the internal MOS is opened and then the circuit that limit the current during the inrush current event is acting (according to the image below).

This creates a high dI/dt with the inductive elements of the filtering circuit that blow up the eFuse.

I would like to simulate the whole circuit in LTSpice but the TPS1641 PSpice model is encrypted on your website, can you provide me the unencrypted PSpice model please ?

Thanks for your time

Samuel