• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 62 subscribers
  • Views 248 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.

LM5101A oscillates and burns up

Roy Hodkiewicz
Prodigy 40 points
Other Parts Discussed in Thread: LM5101A

We are using an LM5101A gate driver in a buck converter regulator for a battery charger.  Several times we have experienced the driver chip burning up unexpectedly.  Usually we can hear an audible noise and then it's dead.  Can you help resolve this issue?  I have attached a pdf file of the converter schematic.6787.lm5105a driver.pdf

  • 0
    TI__Expert 5900 points

    Hi Roy,

    I just looked at your schematic.

    I had concerns at below points，

    1.Does VHB-VHS exceed its maximum value? in spec the maximum value is 18V, we recommend you use it not exceed to 14V.

    In you schematic, i saw the 24V dc is connected to HB.

    2. I do not know your working frequency, but I am afraid that the capacitor between HB and HS is too large.

    3. I am confused at the net of schematic, For example where should W40,W41,W2,W24 Connected?

     

    Regards

  • 0 in reply to Kening Gao
    Prodigy 40 points

    This circuit is for a battery charger with a programmable output voltage from 7 to 20 volts. In our circuit, the boot strap bias is clamped to 15 volts with a zener diode, and we will change it to the recommended 14 volts.   Switching frequency is 50kHz.  The nodes W40, W41, W2, and W24 are tied together through a current transformer primary turn.

    The 475 ohm resistor to +24V (actually is +28V) is there so the gate driver will have sufficient boot strap voltage when a 20V battery is connected to the output.  With a battery connected, the gate driver will not turn the switch FET on because the source node needs to be low enough to charge the boot strap capacitor enough to overcome the boot strap UVLO (and Vgth of the FET).

    Please tell me about how the boot strap cap size affects operation?  Also, we could add a schottky diode, from Vcc to HB, to bypass the diode in the driver chip so it doesn't get too hot?  I thought that once the boot strap cap is charged, regardless of it's size, the FET would require only little sips of current from the boot strap cap, and then in turn the diode in the chip would see those same little sips of current to recharge the boot strap cap.  The peak output current from the driver is limited to only 0.5 amps.  If we were driving 2 amps or more, I would then see there could be a power dissipation problem with the chip.

    The 100uF boot strap cap size was chosen because that part is also used in other circuits, and is a low esr long life cap.  We wish to make this thing last a long time, so we would like to not heat up the boot strap cap, or any other cap (you can see there are many input and output caps).

  • 0 in reply to Roy Hodkiewicz
    TI__Expert 5900 points

    Related to the size of bootstrap capacitor, the needs for the bypass capacitor should be low ESR and placed as close as possible to the driver.

    So a good choice is to place a ceramic capacitor near the driver.

    This need also aplys to the decoupling capacitor of VDD.

    Of course, you could add a diode from VCC to HB if its conduction voltage is lower than the internal one.

     For the damage of driver, the possible cause could be the bootstarp diode damaged inside driver, the negative or positive voltage exceed the spec, the output stage being wrongly shorted.  
    BTW, why LO(w40,w41) is connected to a floating node? I think it would cause some trouble because LI is low.