• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 64 subscribers
  • Views 1137 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.

LM3450 Start Up

Joseph Elek
Prodigy 50 points
Other Parts Discussed in Thread: LM3450, LM3409

At start up there is a half second delay before FLT1, FLT2 and DIM pins are active.  Is there a way to speed up the turn on of these pins?

  • 0
    TI__Mastermind 28655 points

    Hi Joseph,

    What do you mean by active?  There are delay related to filtering the rectified AC, 120 Hz, dim signal to get to a stable DC reference for the converter, is this what you are asking about?

    Things can be sped up but are limited to filtering a 120 Hz signal and the method the LM3450 determines phase angle.

  • 0 in reply to Irwin Nederbragt
    Prodigy 50 points
    Irwin

    There seems to be about a half second delay from the time the BIAS pin goes from 20V to 14V until the DIM pin goes low. Is that the time required by the angle demodulator?
  • 0 in reply to Joseph Elek
    TI__Mastermind 28655 points
    Hello,

    Sounds like start-up delay. Are you using the EVM? I noticed in AN-2150 there is a soft start circuit that could cause what you are seeing, the values will cause a long delay, R20 = 499 Kohm and C12 = 47 uF on the EVM. There is a reason for doing this though it could probably be sped up. Read 9.9, page 21 of AN-2150. Are you using this at 120 VAC? Single stage or two? AN-2150 is two stage, second being the LM3409 circuit.

    Thanks,
  • 0 in reply to Irwin Nederbragt
    Prodigy 50 points

    I'm looking at the pins of the LM3450.  After the BIAS pin transitions from 20V to 14V it takes 0.5 seconds for the DIM pin to become active. Can this delay be shortened?

  • 0 in reply to Joseph Elek
    TI__Mastermind 28655 points

    Hello,

    Can you help me with the questions I asked before?  I don't know what your schematic looks like or if you are using the EVM, single stage or two, etc.  If you read the previous message it mentions a delay circuit.  Look at AN-2150.


    "Sounds like start-up delay. Are you using the EVM? I noticed in AN-2150 there is a soft start circuit that could cause what you are seeing, the values will cause a long delay, R20 = 499 Kohm and C12 = 47 uF on the EVM. There is a reason for doing this though it could probably be sped up. Read 9.9, page 21 of AN-2150. Are you using this at 120 VAC? Single stage or two? AN-2150 is two stage, second being the LM3409 circuit."

    Thanks,

    Irwin

  • 0 in reply to Irwin Nederbragt
    Prodigy 50 points

    Hello,

    It is a single stage, isolated 24V, 96W driver.  Actually, some one at TI did the first pass on the design for Kichler before I started at Kickler.  The soft start circuit you referenced above is not in the circuit.  The soft start circuit above does not work with an isolated design because the 5K resistor connected to 5V on the COMP pin.  

    let me know it you need any other information.

  • 0 in reply to Joseph Elek
    TI__Intellectual 2580 points

    Joseph,

    The DIM pin only changes in response to the "phase dimming decoder" circuit described on pages 19-21 of the datasheet.  There are multiple things that delay the demodulated signal at any time.

    1.  This is a sampled system.  The samplign interval is 3 out of every 25 twic line cycles.  At 60Hz input (120Hz rectified) the status of the DIM pin will change every 200ms roughly. 

    2.  There are RC filters attached to FLT1 and FLT2 pins.  These are to filter out the twice line frequency ripples affect on the decoder circuit.

    3.  The VADJ pin sets the amplitude of the demodulated signal before it comes out of the sampling block.  It is then filtered and compared again to a triangle wave to generate the PWM.  If there is a big capacitor on this, it may be slowing the response as well.

    In your case, once the BIAS pin transitions you know you have the internal circuits working (in fact, it started when you first pass the VCC threshold).  So all of these potential delays sum together.  Without waveforms and a schematic, it is hard to be of more help here.

    -Jim

  • 0 in reply to James Patterson
    Prodigy 50 points

    Jim,

    I've been out of the office.  Thanks for the explanation.  The start up delay makes sense now.

    I do have another question on burst mode operation.  In the data sheet it say if the COMP pin drops below 1.4 volts the chip enters burst mode.  What is the tolerance on the 1.4 volts?

    I had to move the scope trigger up to 1.55V to get the scope to trigger on the COMP pin but you can see the burst mode on the gate pulses.

    Thanks

    Joe