Due to the U.S. Thanksgiving holiday, please expect delayed responses during the week of 11/22.

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.

PMP6024: Start up circuitry overheated

Part Number: PMP6024

Hi, 

I have built a prototype base on the reference design PMP6024 and have some issues with the start-up circuitry, i.e. Q1 (schematic of Tidu958, page4) which got overheated at power on. What can cause this?

Br,

Conny

  • Hi,
    I have built a prototype base on the reference design PMP6024 and have some issues with the start-up circuitry, i.e. Q1 (schematic of Tidu958, page4) which got overheated at power on. What can cause this?
  • Hello Conny,

    Did you use the same transformer as the PMP6024 uses?  Can you look to see if the Q1 circuit is getting disabled after power on?  If it is supplying IC power through Q1 instead of the primary bias winding it is not working correct.  Also, Q1 needs to have heatsinking, it is a linear regulator from rectified AC so it will dissipate power.

    Regards,

  • Hi Irwin,

    Yes,I do use the same transformer.

    I have found the root cause to the start-up. The FET that I used in my design is not equivalent to the original one when switch back to the FQN1N60CTA the start-up seams to be more stable. For the moment It is not a start up  issue. 

    Another question that I have asked previously,  how does the soft start works in this design?

    Br,

    Conny

  • Hello Conny,

    Soft start, or overshoot prevention, is done by C19. It is also limited by C5. It is a PFC so it will respond naturally slow.

    Regards,
  • Thanks Irwin!
    Now back to the start-up again.

    I have now been able to perform som more measurements and test with proper load i.e. LED light. I have experienced that at no load the Q1, FQN1N60CTA is also getting overheated. That can reproduce by starting with load (PWM the LED-light > 10% duty cycle) and reduce the load (PWM the LED-light = 0% duty cycle). When measure the voltage levels @ D9_pin C=13.3V @D11_pinC = 11.4V, that means the Aux winding voltage is too low and not able for Q5 to sink the Q1_Gate voltage.

    Questions:

    Can the Aux winding be modified with thicker wire to achieve more power when having very light load/ no load ?

    My application requires universal input voltage 85-265VAC and this reference design is aimed for 100-240VAC. What is need to modify to for to meet this requirement? Specially the start up circuitry and the transformer.

    Appreciate your support!
    Br
    Conny
  • Hello Conny,

    If you plan on trying to dim this to zero output current the IC will have to run from the Q1 circuit or the output will need more complicated circuitry to keep current flowing to the secondary (this includes loading to the output that is not going to the LEDs).

    No, making the wire large will not change the bias voltage.  The problem is that you are regulating current.  If the LED current is lower the forward voltage of the LED stack will be lower.  When the LED stack is lower the primary bias will be lower by the ratio of the secondary number of turns to the primary bias number of turns.  The current is low in the primary bias winding so wire size really won't affect it much.  The aux winding can be modified by adding more turns, or the output voltage (LED forward voltage) can be increased.

    Universal input from 85-265 is possible but the transformer and some of the design may need to change.  At 85V input the peak current in the transformer will need to be higher.  If the magnetic flux in the transformer is near saturation the transformer may need to be redesigned or the output power reduced.  The start-up circuitry isn't really affected by the input voltage since this is a flyback.  It is to some extend due to duty cycle and such but that's different than the dimming problem.

    Regards,

  • Hi Irwin,

    I send you email, please check your mailbox!

    Br,

    Conny