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.

LM3404HV: Thermal Shutdown - LM3404HV

Part Number: LM3404HV

Hi all,

We are facing some thermal issues with a prototype of 40W LED driver based on LM3404HV (SOIC8).We are trying to drive a 38V 1A COB LED from an integrated 52V Flyback SMPS.The Flyback SMPS and LED driver is on a same PCB inside a plastic enclosure.Its working fine as open frame but when the enclosure is closed,working for a few hours the LED starts to blink.We also tried separating the SMPS part but it doesn't helped.We now doubt both the frequencies,switching and dimming.Please help !

VIN - 52 V

Vout - 38V

Iout - 950mA

Rcs - 0.2 ohm

Fsw - 250Khz

Fdim - 500hz (3.3V)

Ron - 1meg

Diode - MBRS4201T3G (SMC)

Inductor - 100uh 1.5A 10x15 Drum

Thanks and Regards.

  • Hello,

    Have you measured the temperature of the components when enclosed and not enclosed? What is the temperature in the enclosure? Why are you questioning the dimming frequency and switching frequency if this works for a few hours then starts blinking? Is the enclosure sealed? How does heat get out?

    Best Regards,
  • Hi Irvin,Thanks for your reply.
    Yes,we checked the temperature with and without enclosure.Without enclosure the temperature is about 75dC and when closed it rises up to 90dC and undergoes shutdown.Is this 70dC normal in without enclosure condition?we had given some more copper area in the PCB than recommended in the datasheet.We doubted the frequency because in long run (24 hrs) operations the temperature keep increasing,So we thought that some frequency mismatch between the Fsw and Fdim may be the reason.Do we have to cross check the copper thickness of the pcb?or do we have to decrease the output current to a safe limit?
  • Hello,

    Frequency mis-match between the switching frequency and PWM frequency shouldn't cause issues.  What is probably happening is the LM3404HV is in an environment that causes it to run on the higher temperature side, this causes the RDSon to be higher making it dissipate more power.  Where are you measuring 90 degrees Celsius?  Is it on the LM3404HV?  Is it the ambient in the box?  Is it the PCB temperature?

    I'm not sure why you are using "Diode - MBRS4201T3G (SMC)", I would look for a lower voltage part that is a schottky possibly with a lower current rating so the die area is smaller, just have to make sure it doesn't run too hot and has low leakage at higher operating temperatures.

    Are you certain the LM3404HV is thermally shutting down or can it be another part of the circuit shutting down?  How long is it off when it shuts down?

    The layout may require more heat spreading for the LM3404HV to reduce its operating temperature.

    Best Regards,

  • Hi Irvin,

    Sorry for the incomplete information,90C is the temperature I measured on the PCB.For LM3404HV in worst condition we got as high as 150C and we also checked the input supply (50V) when the LED blinks but its steady.Off time during the shutdown is around 500 ms and On time is around 3 seconds.Will update you after getting the issue again.

    After we changed the switching frequency to 100Khz ( with 10M Ron and 200uH inductor) we didn't see any shutdown issue.But the temperature of IC is near 90-100C.Is there will be any improvement on thermal performance if we use PowerPAD SOIC instead of normal SOIC?

    Thank you Irvin.
  • Hello,

    At your output current and duty cycle I assumed you were using the PowerPAD SOIC. Yes, it would probably be better to use the exposed pad part to get more heatsinking. However you may want to figure out how to also keep the PCB and other components running cooler as well especially if you have electrolytic capacitors or other heat sensitive components. There is also safety to consider when operating at elevated temperatures, that is something else to look into.

    Best Regards,