• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 63 subscribers
  • Views 424 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LM3409HV: Frequency switch too high when LED load is Vin/2

simone casaluci
Prodigy 10 points
Part Number: LM3409HV
Other Parts Discussed in Thread: LM3409

Hi,

I designed and realized the LM3409 driver board with the following characeristic (following the datasheet schematic):

Vin 48V

Fsw= 600kHz, Coff=470pF, Roff=10K

Iou=1A

L=15uH 

led load= from 1 LED (Vf=3V) to 12 LED

When the led load is max (12LED) the Toff is around 160ns and the frequency switch is around 625KHz. When the load is 1 LED, the Toff is around 1.27us and the frequency switch is around 700KHz.

I have found the excessive frequency switch with the half led load (when the Vout is Vin/2). in the latter case the fsw goes to 2MHz !! and I have thermal issue whit the external PMOS (at moment I mounted SQ2361AEES) but I am going to replace the PMOS with a stronger.

Why this happen and how I can solve?

BR

Simone 

  • 0
    TI__Mastermind 28655 points

    Hello Simone,

    This happens because the LM3409 is a variable switching frequency part.  It's highest switching frequency is when the dutycycle is at 50%.  The MOSFET will get hot when trying to switch that fast with 48 volt input.  It can work this way by choosing a MOSFET that has better heatsinking capability.  You can also reduce the switching frequency, to keep the current ripple the same you could use a 33 uH and change Roff to approximately 22 Kohms.  The MOSFET will still run hot due to switching loss and conduction loss at higher output voltage, it's really too small for 1A at these voltages and switching frequency.

    Also note that the LM3409 maximum minimum on-time is 211 ns.  The switching frequency should be lowered to make sure all operating points ton is greater than 211 ns so in this design Roff would need to be at least 13 Kohms.  This will also change the regulation point so Rcs would need to be lowered since the LM3409 regulates peak current and the average is Ipeak - Ipkpk/2.  Lowering the switching frequency will cause the peak to peak current ripple to go up unless the inductor value is also increased.  I would also measure Vin and Vout for 1 LED and 12 LED since the calculations show different switching frequencies than you are seeing.

    Vled toff duty ton Fsw
    3 2.62152E-06 0.0625 1.74768E-07 357616.689
    6 1.1375E-06 0.125 1.625E-07 769232.775
    9 7.28282E-07 0.1875 1.68065E-07 1115639.75
    12 5.35815E-07 0.25 1.78605E-07 1399737.56
    15 4.23862E-07 0.3125 1.92664E-07 1621990.8
    18 3.50623E-07 0.375 2.10374E-07 1782542.3
    21 2.98971E-07 0.4375 2.32533E-07 1881450.64
    24 2.60587E-07 0.5 2.60587E-07 1918744.28
    27 2.30939E-07 0.5625 2.96922E-07 1894438.73
    30 2.07349E-07 0.625 3.45582E-07 1808543.16
    33 1.88133E-07 0.6875 4.13892E-07 1661063.34
    36 1.72176E-07 0.75 5.16528E-07 1452003.07

    Best Regards,