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.

LM5122 Blows mosfets

Other Parts Discussed in Thread: LM5122, PMP7992

I'm designing a prototype power supply to using an LM5122. It's designed to work for a 18 - 28VDC input and a 24V 6A output, but for some reason both my mosfets go into short circuit as soon as the supply is enabled (as soon as the turn on threshold on UVLO is reached)

I've checked and double checked my design but I can't find out why. The mosfets should be able to handle up to 80V and 60A. This should be enough

The schematic (still prototype, could be neater)

Am I missing something why this is happening?

Hope somebody can help me with this.

  • Hi

    If you are sure there is no layout mistake, please check the direction of your flyback transformer. 

    Regard,

  • Yes, I checked this a lot of times.

  • Hi 

    Looks like everything okay. 

    Have you check if any excessive switching spike at the drain of your two switches ?

    For a test purpose, SS pin can be used to control the output voltage. Please try to increase  output voltage slowly using SS. 

    Also, please refer http://www.ti.com/tool/pmp7992

    Regards,

  • I didn't check this, but thought about it. 

    My calculations say 32nF is big enough. The pcb is fitted with 100nF. I will try it with a much bigger capacitor.

    EDIT:

    just checked it with a 1uF capacitor and one single led on output (on pcb). There is  power for a very short time and then I hear a small spark and power is off en input in short circuit.

    Controlling with SS pin, you mean applying a voltage and slowly increase it instead of a capacitor?

    EDIT2:

    Just checked again with an extra power supply on the capacitor pads (capacitor removed).

    1. I set the SS power supply to 0.00V
    2. Add power on VIN, 23V. No problem.
    3. Turn up the SS power supply, first step (0.04V) and FETs go into short circuit again.

  • Hi 

    I am still thinking there is an error on your board, either layout or component.

    Looks like the sound you heard  was the sound caused by short circuit. 

    If you can , please provide some waveform. You might be able to capture waveform if you decrease your power supply current limit and start-up at no load. 

    Regards, 

  • It possibly is, but I can't find it.

    Yes is probably is.

    Currently I'm out of FETs, I will order new today.

    The supply itself is running. These are the HO and LO signals without any FET's installed.

    Speed is ~500KHz as expected.

    EDIT:

    the inductor is calculated set to 1uH

    (23/((192/23)*0,25))*(1/500000)*(1-(23/24)) = 0,918403uH

    My analog experience is not very good. Is this enough?

  • Hi

    It is too small. Within one switching period, the current will be increased up to 56A and saturated. The switches can be damaged if your magnetic is saturated. Please refer http://www.ti.com/tool/pmp7992

    Regards,

     

  • Oké, I hope that is the problem. I will order bigger one. How to calculate the correct value for this supply?

    And how did you calculate the 56A?

    Ken

  • It was from 28V / 1uH / 500kHz = 56A. 

  • Finally the circuit is working. Needed a lot more cooling then expected. I can't find the power loss calculations for this flyback setup.

    Next problem is the coupled inductor. Now I'm using the 74485540440 , it works great but it gets hot! 150 degrees celcius after a while, then I turned it off.
    This is ofcourse told in the datasheet, but for prototyping now it was oké. I need to change the inductor to a higher current type but this seems to be a problem. The 12A rated current and 24A saturation current at 4 point something uH is the highest coupled inductor I can find.

    DcDc or flyback transformers have a much higher inductor value.

  • Hi

    Looks like you have to contact magnetic manufacture and ask a custom design. If it is okay, you can configure two phase flyback using two LM5122 boards. 

    Regards

  • Hi

    Two phase will decrease the rms current with roughly 50%? And because the extra switching is 180 degrees off I can half the frequency for the same 'speed', decreasing switching losses by 50%?

  • Hi 

    Yes, RMS current will be about 50% per phase.

    The switching frequency can be half, but in your case, my recommendation is to keep the current switching frequency since you need a small inductance of transformer.

    Total switching loss will be about the same, but it will be split into two.  

    Regards