Post Go back to editing

LT8705 Failures - INTVcc short circuit and strange behaviour when heated up

Category: Hardware
Product Number: LT8705

Hello everyone,

I am writing this because lately we have had two issues with LT8705 component. In both cases changing the component resulted in the fix.

The final application is a small PLC for earth-moving machines. The supply is a battery, usually 24V (could be also 12V) and the output is 12V 1A. The mosfets are BSC070N10NS3G and we put a 3,3Ohm gate resistor. The inductor is 47uH and the switching frequency around 350kHz.

- First problem: two boards with INTVcc in short circuit. We got back two (in around 200 boards) and found that the INTVcc pin measured a short to ground. Substituting the component fixed the short. We are not able to replicate the problem on the lab and. Does anyone had the same problem ?

- Second problem: we have other 2 boards that function normally at ambient temperature. BUT if we heat the LT8705 up just a little the start behaving erratic:

                                      - One board's output voltage oscillates between around 22V (with 24V input) and zero. Measuring INTVcc voltage we see that while we heat it up the voltage rises. When it reaches around 7,9V the IC starts to oscillate. Looking at the CLOKOUT pin the die                                                               temperature sould be around 91°C

                                      - The other board instead we heat it up and nothing happens until we try to switch off and then on the board. At this point the board is dead and the output of the converter is 0V. Only when the chip cools down the board starts to work back properly. This behaviour                                                   appears far below the "Overtemperature shutdown" since by heating other boards the same we do not see this behaviour. Looking at the CLOKOUT pin the temperature of the die sould be around 111°C.

Any idea of what could it be? 
If you need any other information let mw know. Thank you

  • Thank you for the detailed write-up. The INTVCC voltage of 7.9V is definitely a concern. Could you confirm that measurement?

     

    Just to narrow down if this is an application issue or a board issue, are you able to share: 

    - A scope shot of INTVCC waveform during the fault

    - Schematic

     

    If you prefer not to share the schematic publicly on the forum, we can also open a support ticket and continue debugging that way.

  • Thank you for the quick response. 

    Here in the following you can see a screenshot of what happens when i heat up the board with the IC that starts to oscillate

    Detail when working properly:

    Detail when oscillating:

    Here also the schematic of the part of interest:

    Schema LT8705.pdf

    Moreover regarding the two damaged ICs I took some screenshots of the Switching node 1:

    Can the negative voltage be a problem for the LT8705 ? The datasheet is a bit vague and does not provide a definite voltage. 

    Thank you

  • Hello,

    I have taken some screenshots of the BOOST circuit. 
    What I found is that there is a strong oscillation on the ANODE of D4 and D5 ( the boost diodes). The diodes used are CMMR1U-02 like in the evaluation board. The oscillation presents itself when the LT8705 drives the top buck mosfet ON (MF4 in the schematic). On the GATEVCC pin the voltage is much lower but I am afraid that in some particular situations the voltage could go above absolute maximum limits. 
    Here are the screenshots:

    D4 ANODE:




    GATEVCC:

    Do you see anything strange?