• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 61 subscribers
  • Views 533 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LM2655 Synchronous Switching Regulator

John Wilce
Prodigy 30 points
Other Parts Discussed in Thread: LM2655

HI


I am having a problem with a power supply based on the LM2655 switching regulator operating in synchronous mode.

The design is based very closely on the example circuits in the data sheet. We have 12V input and 1.2V output. We have many of these boards running with no problem but I have one board that refuses to work. The regulator fails with a short between the Pvin and SW pin. The input supply goes into current limit.

The device fails immediately the power is applied but I can find no short on the output. This has happened twice on the same board - when first built and following a replacement of the regulator. I have also had the lower FET, Schottky diode Cboot, Css and Cdelay replaced just in case one of these was faulty/wrong value.

I can find no obvious differences between a good board and this failing board.

Can anybody suggest a failure mechanism that would cause the device to fail in this way; normally these devices are pretty well self protected to prevent this sort of thing happening.

  • 0
    TI__Intellectual 2225 points

    John,

    A short from PVIN to SW suggests that the internal high-side (HS) FET has been shorted.  This internal FET is protected by current limit, so a failure means that the current limit was not able to protect the device.

    The cases in which the current limit do not work are:

    1. A short from the SW node to ground

    2. If the inductor saturates and the inductance value drops, or if the inductor is somehow shorted.

    I would try verifying there is no short from SW node to ground, and also try replacing the inductor.  A shorted inductor can occur due to a solder bridge during assembly or cracking, and can be difficult to debug because an inductor has low dc resistance anyway.

    Hope this helps

    Perry

  • 0 in reply to Perry Tsao96839
    Prodigy 30 points

    Hi Perry

    Thank you for your reply and suggestions. Sorry for taking so long to get back to you - other projects taking priority (I guess we are all used to that).

    I have checked, and there is no short between SW and ground, and I have checked the condition of the inductor - no short found. I replaced the inductor (and the regulator once again) anyway, and the failure re-occurred.

    We have since had another board fail in the same way except this time it took about 30 seconds for the regulator to blow.

    A little further investigation (the board is not my design) and I have found that the regulator is supplying a very light load (less than 100mA). On a 'working' board I have found that the bottom FET is mostly in the switched on state, with regular off periods every ~3.4us (as expected with 300kHz oscillator).The off period most often for 8.4ns but occasionally the off period is 840ns. On average there will be two or three short off periods to every one longer off period.

    The very light load concerns me. The data sheet isn't too clear about how the device handles light load and in any case the compensation circuit values are as per the application schematics so unlikely to be set for the conditions of a light load.

    What are your feelings about this light load condition?

    Thank you

    Regards

    John

  • 0 in reply to John Wilce
    TI__Expert 4100 points

    Hi John,

    There should be no reason for device to have issues with light load conditions. As you noticed under light load conditions the low side FET will be in switched on state and if this FET is properly sized device should operate without any issues.

    You mentioned that board inspection did not find anything out of ordinary, but considering described behavior we think that another board check could provide some clues. We would suggest to check FB and COMP pins (and corresponding circuitry/ if any) for any short circuit conditions or anything out of ordinary.

    Thank you

    Regards

    Brani