LM25010: burns out

Hello,

Thank you for sharing the schematic and board image. It looks as though the IC is damaged on the input voltage pin side.  Would you be able to answer my additional clarifying questions for me?

Which capacitors are the input voltage?  Is the LM25010 directly tied to the 24V power supply?  Is there any additional filtering components or protection components? Also what is the power supply being used? 

Thank you for your help.

Best Regards,

Katelyn

Hello Katelyn,

All capacitors can stand up to 50V.

There is an input protection circuit (please, see attached picture).

24V is generated by an expernal power supply not designed by the customer. We have to consider 24V nominal.

Current required on +12V: no more than 100mA

Current required on -12V: no more than 50mA

Bursts and Surges are limited by D2

Thank you for your help.

Regards

Riccardo

Hello,

Can you share your layout as well?

What is the inductance of L3 and L4? It is possible to have ringing from the large inductances and parasitic loop of the input of the DC-DC. This rining is created after the TVS (at the input of the DC-DC) and not at the power supply. This would lead to a possibility of your TVS to not see it. Do you have a waveform of VIN, SW and VOUT of the LM25010?

Thank you,
Katelyn

Good morning,

LM25010 replaced with a new one and hereafter the signals:

Yellow: Vin (+24V)

Green: Input Current

Purple: Vout +24V (should be +12V)

Light blue: Vout 0V (should be -12V)

We are completing the check of the board

Thank you

Riccardo

Good morning,

we have tested other samples bringing the power supply over the limit because the customer suspecs there are input slow transients over 50V.

The input protection (TVS Vbreakdown = 36V) cuts fast transients but can stands up to 50V with slow transients.

The LM25010 easily supports the overvoltage.

The diode between ISEN and SW pins (Vreverse = 40V) is damaged by the input overvoltage and becomes a short-circuit. Subsequently, the LM25010 burns.

We need to reduce the TVS Vbreakdown and increase the diode Vreverse.

Let me know what you think.

Regards

Riccardo

Hi Riccardo,

I want to make sure that I am understanding you correctly. I think that I agree with your analysis.

I understand that the first component damaged is the diode between ISEN and SW pins. What is the input voltage during this time?

Best Regards,
Katelyn

Hi Riccardo,

I wan to follow up with you. Would you be able to confirm what is the input voltage?

Thank you,
Katelyn

Good morning Katelyn,

The oscilloscope screenshot published a few days ago was taken by measuring a board in which the LM25010 did not look damaged.

Testing a brand new board with a power supply of 24V the outputs are correct.

We cannot confirm the input voltage on field.

Neither our costomer nor the final user measured the input voltage.

We can only assume that there are strong variations of the power supply.

 

Next step, in agreement with the customer, is to bring the power supply over 50V and to show that the diode between ISEN and SW pins burns and consequently also LM25010 burns.

Best Regards

Riccardo