Results 1 to 5 of 5

Thread: LT3474 as diode driver?

  1. #1
    Join Date
    Aug 2009
    Location
    Bristol, England
    Posts
    333

    Default LT3474 as diode driver?

    Anyone tried the Linear tech LT3474 as a driver for these diodes yet?

    The part is designed as a switch mode driver for LEDs, operating in buck mode from up to 36V and outputting an amp maximum to the diode using current mode drive. The nice thing is that the high supply voltage means that we could run the drivers off the +24V scanner rail, saving on a power supply.

    Switching can be up to 2Mhz, so a hundred Khz control loop bandwidth should be possible.

    It can be modulated via a control input (but there is a question about how fast the control loop is, and if there is any overshoot)), and IF the output is sufficiently clean it might make a good driver with appreciably better efficiency then the usual linear pass transistor affair.

    I am thinking I might get a few and have a play if nobody else has done so yet.

    Regards, Dan.

  2. #2
    Join Date
    May 2009
    Location
    UCSB
    Posts
    715

    Default

    With careful design it could be used to drive a 445nm diode (for reference, the driver in the projector is a buck converter similar to this) but the modulation performance of such a driver will be dismal.

    If someone wants to tackle the project you might be able to get it modulated at 20kHz from a 2MHz clock, but designing the feedback loop will be a nightmare.

  3. #3
    Join Date
    Aug 2009
    Location
    Bristol, England
    Posts
    333

    Default

    Yea, I am thinking that probably the right way to go is a simple voltage output switcher followed by a linear current control stage.

    Just trying to do away with the additional 5V rail that I really don't need for anything else.

    Regards, Dan.

  4. #4
    Join Date
    Dec 2008
    Location
    Vezon, Belgium
    Posts
    1,017

    Default

    why not use a shunt modulator in parallel with the diode? a mosfet current sink should be useful for that

  5. #5
    Join Date
    Aug 2009
    Location
    Bristol, England
    Posts
    333

    Default

    Won't work with that chip, the output filter cap for the buck converter is effectively across the diode (well there is a current sense resistor in the way, but...).

    The thing is not really a constant current source, more like a voltage source with current sensing feedback so the impedance at the output is not high at typical modulation speeds.

    I strongly suspect that dv/dt will be too low to modulate well, even given that we only need about half a volt swing at the anode.

    I think the better answer is just to run a switcher to give a constant 6V or so rail, then do a standard low side driver.

    Regards, Dan.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •