• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 62 subscribers
  • Views 2069 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS54360: TPS54360 failing short on input after a year of operation

Mohammad Asgari
Prodigy 55 points
Part Number: TPS54360


Hi,

We are experiencing a strange failure on some devices using the TPS54360 chip. The chips are bringing 54V down to ~4V, the 4V gets regulated down to 3.3V with an LDO. The current draw is roughly .5A and the load is dynamic (LED screen). The devices had been working fine for a year. They are now failing at a rate of 1-2 a week!

There are 15 devices on a 54V power bus and the failure results in a short-circuit at the input of the TPS54360. Consequently, the power fuse blows on the bus and other modules stop working. In some cases, after opening the housing, the chip/PCB is visibly burnt. 

We have no clues as to what could cause this...

  • 0
    TI__Guru**** 191445 points
    Typically those types of failures are due to electrical overstress (EOS). I would start by looking at the input voltage for transients above the abs max voltage then look at the PH node for any excessive ringing, over shoot or under shoot. If your boards have been working for a year and now start failing, perhaps something has changed with regard to your input voltage.
  • 0 in reply to JohnTucker
    Prodigy 55 points

    Thank for the prompt response John.

    I have to correct myself, the chip is not shorting at the input, it is the catch diode that is shorted (punched through?). The short is visible from the input, which is what led me to believe that the chip was shorting. I only discovered this after removing the diode from the circuit.

    I will look for transient responses that exceed the 100V specified in the datasheet of SK810 . But I would appreciate any insight you have into this.

  • 0 in reply to Mohammad Asgari
    TI__Guru**** 191445 points
    So the TPS54360 is undamaged?
  • 0 in reply to JohnTucker
    Prodigy 55 points
    It is damaged. The chip does not work when I swap out the diode with a fresh one. In fact, after I tested with a new diode, the chip is now shorted on input and output. I tested with a lower voltage (24v)

    There are two components failing (Diode and TPS54360), do you have an opinion on which is main failure and which is collateral?

    In any case, I will be probing the switching node soon on a scope, hopefully the probe capacitance will not interfere with the measurements. I will be looking for spikes under various loads.

    Thanks!
  • 0 in reply to Mohammad Asgari
    Prodigy 55 points

    Test results in:

    No unusual spikes, I'm using low inductance probe and sampling at 1G/s with no filtering (8bit). However, what is unusual (to me at least) is that I am seeing about 85Vpp on the SW node (across diode) with a 54VDC input. Is this normal? Shouldn't the pulses be maxed out at the Vin? Even in the datasheet, the diode that is used had a breakdown voltage of 60V... The diode we are using is 100V max

    See screenshot below (using x10 Probe, so multiply y axis by 10)

  • 0 in reply to Mohammad Asgari
    TI__Guru**** 191445 points

    There must be some measurement error.  The maximum SW voltage should be at about Vin.  The negative voltage portion looks too low also.  Maybe your probe is not well grounded?