Post Go back to editing

LT3471 fail

Hi we have a circuit using a lt3471 to generate +/- 16.5 V from 12v supply..   Load draws approximately 200mA each rail

We actually have several such circuits in a row on our prototype PCB

Unfortunately we have had several failures of the part, specifically the negative rail causing the par to go into high current mode, where it gets very hot, 110 oC.

We suspect some of the failures may be cascaded from adjacent lt3471 circuits due to heat transfer. But at least three circuits so far have spontaneously failed.

Modelling suggest the steady state load is causing less that spec limit of 1.3A  into the device.

We have one theory that it is due to the startup profile of the 12V input, We had earlier board failures elsewhere where the 12v psu only came up to 2-3 volts and the lt3471 inputs may have rested this low voltage for some time.

Do you have any thoughts.?

my circuit is attached

  • Hi, CNash596, 

    I suspect you might have an inrush current situation during startup, (I know, this sounds redundant, but, oh well).

    Please read the "Soft-Start and Shutdown Features" on page 7 of the datasheet, particularly the last paragraph. You might have an overstress condition during startup, with the negative rail. Try re-sizing the SHDN RC values.

    By increasing the values of both, resistor and cap, slow down the switch current such that there is no switch current overshoot. You might have to increase the values 5X or higher of what you have now. This can be checked first in simulation. 


    I noticed your C203 is rated for 25V. This is not enough. This cap should be rated for >30V.


  • Thankyou for this, I will do what you suggest.