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.

ISSUES WITH LM3464 FOR 4 CHANNELS 700mA LED DRIVER

Other Parts Discussed in Thread: LM3464, TL431, LM3466

We are using a power supply with output 40V x 3A and we have problems to adjust the current in 4 channels.

Using only one channel we can adjust the current (700mA) and it stay stable, but when we increase more channels the current goes down (380mA) and isnt stable.

we are using it to connect 12 leds with Vf 2,9V for each channel and the circuit is mounted as AN-2071

  • Eduardo:
    LM3464 is a dynamic headroom control 4 channel LED driver. To make it work, it has to control the AC/DC front end power supply by connecting the feedback pin of LM3464 to the voltage feedback circuit of the AC/DC power supply and make the output voltage of the AC/DC power supply close to the LED string voltage. Did you make those changes to the AC/DC power supply?
    The LM3464 eval board AN-2071 is 4 channel 350mA board. Can you tell me what changes you made to drive the LEDs with 700mA?
    David Zhang
  • Hi David Zhang.

    We made that changes to the AC/DC Supply. We connected  using R1=150k and R2 = 11k. And TL431 is connected in D1 (SL04)

    The changes made in the AN-2071:

    RISNSx = 0R28

    RFB1 = 36k

    Q1, Q2, Q3 and Q4 = IPD200N15N3.

  • Edurado:
    I am glad you did the change to the AC/DC power supply. What was the Vrail voltage when you connected 4 strings of LEDs? Can you please check if Vrail is stable? If it is not stable, you need to adjust the control loop. One suggestion is to increase the CDHC cap to slow the loop.
    I have a few concerns on the resistor values you picked and the changes you made on the eval board. First Vrail(nom) is too high.
    With R1=150K and R2=11K, Vrail(nom) = 36.6V which is higher than the LED voltage. Vrail(nom) should be a few volts lower than the LED voltage. Second VDHC_READY is too low. With RFB1=36K and RFB2=2.67K, VDHC_READY = 36.2V. This voltage should be a few volts higher than the LED voltage.
    Another concern is the MOSFET overheat. With 700mA, the MOSFET IPD200N15N3 could get overheat. If the voltages on all 4 LED strings are the same, the MOSFET drian to source voltage is 0.7V which is OK for 700mA current. But in real applications, they are not thes same. The MOSFET drving the lowest voltage LED string sees the highest Vds. 1V difference on the LED voltage leads to 1.7V worst case Vds and 1.2W power loss which may be too high for the MOSFET.
    David Zhang
  • David:

    Ok, i made the following changes:

    R2=13k

    RFB1=40k2

    Vrail=31,35V

    VDHC ready = 40,10V

    Simulating the led string with a DC power supply with current setted as 700mA we had 34,9V (Vf total of the leds)

    After that we tested the circuit and we had the strobo effect in the led.

    We changed CDHC (220nF) to 470nF and to 1µF.

    Nothing changed. The strobo effect remains in the led.

  • Eduardo:

     

       What is the Vrail voltage after the LEDs are turned on? Can you show me the waveforms of the Vrail voltage during the shart up and after the LEDs are turned on?

     

    David Zhang

  • David,

    See attached the waveform of Vrail after the LEDs are turned on.

  • Eduardo:
    It looks like the AC/DC power supply is not stable. It is turned off after the LEDs are on. You need to find out what makes the AC/DC power supply turned off. Is it over current, over power or unstable control loop? LM3464 can not regulate LED current without stable Vrail.
    David Zhang
  • Hi David.

    In the start it had more current (2,7A) and more voltage (48V) that we are using now. But we tested the power supply again and it really dont have issues with over current and over power.  We will try to check now about control loop.

    While we do this i want to know if you have a suggestion in another power supply that we can use. Something with 130W and output around 165V @ 0,7A or a power supply with output 36V that we can use in 4 channels with 0,7A

    Do we need to use one specific TL431 for this circuit? We use in this circuit TS2431A.

  • Eduardo:
    Did you build your own AC/DC power supply? Because LM3464 changes the power supply control loop, it is hard use an off the shelf power supply and simply modify it. You need really understand the power supply to make it work.
    David Zhang
  • David

    We are using one evaluation board with small changes in the transformer to correct our output voltage.

    The original evaluation board had 48V x 130W. After the changes we have 35V x 2,7A. Then we adjusted the circuit as you said before.

    You have a suggestion to exchange LM3464 to another component? Or one power supply that you know that we can replace? 

  • Eduardo:
    Can you tell me what AC/DC power supply eval board you are using? If you don't want to design your own AC power supply, you can use a standard constant current AC LED driver like mean well's HLG-100H-36 and use LM3466 to do LED current sharing. If you are not going to make many of this product, that might be a better choice than designing your own AC power supply.
    David Zhang