help to build precision LED driver?

Discussion in 'The Projects Forum' started by smilem, Apr 27, 2009.

  1. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    Hi, I have this circuit diagram
    http://www.edn.com/article/CA316054.html
    [​IMG]

    This schematic shows 5V output "5v to circuit" do I need to connect it to multiple 5V "inputs" shown as empty circles shown as "5V"?

    Also the LM324 is 14pin chip. What pins I need to connect to "+", "-", "R7"?
    "1IN +" "pin3" connect to "+"
    "1IN" "pin2" connect to "-"
    "1out" "pin1" connect to "R7"
    "GND" "pin11" connecto to ground.

    What and where do I connect "TTL input"?
     
    Last edited: Apr 27, 2009
  2. David Bridgen

    Senior Member

    Feb 10, 2005
    278
    0
    Yes, IC1's output supplys all those points marked 5V.

    You have a choice of four op-amps within the LM324 i.c.

    The pins you mention are fine.

    The TTL input enables you to switch the current source on or off. It's labelled TTL because, one supposes, that the designer used a TTL output to perform this function. You can use a mechanical switch if you want. Or, if you don't need to switch the source off, simply connect those two points together.
     
  3. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    [​IMG]

    Seems like major overkill for an LED. Why all the extra circuitry?
     
  4. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The circuit and accompanying write-up was authored by Dick Cappels and published in EDN.

    The OP, for unknown reasons, started a new thread. In the previous thread which went on for some 60 or 70 posts to get them to where they needed to be using an LM317L and a couple of resistors to make a constant current source, suddenly mentioned that they wanted to power it from a USB port.

    The OP now wants to power this new circuit from the 5v of a USB port.

    If you have the time and patience, give him a hand. My time is just too limited nowadays.
     
  5. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    Because this will be used to drive a led that I will use to measure and plot light spectrum graphs using very sensitive equipement. I can't risk LED intensity changes.

    And yes I would like to use 5V (from usb because I will have PC handy when I use this device). Otherwise I just have to use normal wall adapter for +12V to power it.
     
  6. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    What color? Red is no problem with USB, the blue and white are a bit more problematic. This is because different colors drop different voltages.

    Looking at the diagram, it is a precision current regulator. I don't see any provision for temperature control or optical feedback, so I suspect it could be simplified radically. Are you needing the remote on/off feature or PWM? This is the real reason for the complexity, he wanted to created super stable pulses.
     
  7. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    I started new thread because this is entirely new circuit design than I was suggested in my other thread.
    54 posts, not 60 or 70.
    My last thread title is: "Very stable power supply from USB to UV LED"
    So how can you state "suddenly mentioned that they wanted to power it from a USB port."

    Thanks for help.
     
    Last edited: Apr 27, 2009
  8. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    This is for UV LED.
    Led PDF is here:

    OA-260019 LED UV 5mm (781)
    http://www.elfa.se/pdf/75/07500366.pdf

    -------------------------

    In this link here they say it's stable with temperature changes: http://www.edn.com/article/CA316054.html

    "Voltage at the wiper of R6 results in an equal voltage across R9 because of feedback to the op amp. Because transistor Q2 has a high alpha, most of the emitter current that produces the voltage across R9 comes from the collector of Q2. Because alpha varies little with temperature, this current remains stable."

    ------------------------

    Optical feedback would be nice IMO. Is it possible to implement this in this circuit or do you know better ones?

    "Are you needing the remote on/off feature or PWM? This is the real reason for the complexity, he wanted to created super stable pulses."

    I need to have on/off button on the device is that what you call "remote on/off"?
     
  9. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    You have to understand it is work writing these posts, and to have parameters shift under you is fustrating, probably more fustrating than you realize. I've helped groups before that did the same thing, like the light tree I built for the robotics group. They wanted simple, so I built and finished simple using 120VAC, then they wanted low voltage controls, so I built and finished low voltage controls, then they wanted to match another guys computer controls (which used different voltages and polarities), so I did that..

    My point is define what you want up front. If there are technical issues we make changes, but a moving target is 10X harder to hit.
     
  10. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    We have a posting lag going on here...:rolleyes:

    For on off a simple switch will work. If all you want is stable current it can be much, much simplier.

    UV is equivalent to a white light LED, since the core is the same.

    All you really need is a really stable current source, which isn't too many parts. Variable light intensity is a bit more complicated, especially if you want to hit the exact same points every time.

    Will the LED be shining though flesh?
     
  11. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    I need to power this UV led, from computer usb port:
    OA-260019 LED UV 5mm (781)
    http://www.elfa.se/pdf/75/07500366.pdf

    The LED power supply must be very precise. The brightness can't change. I will use it to make measurements using precise specrophotometer. I'm not making a light tree or a torch here.

    Optical feedback is welcome if the sensor that reads the spectrum of this wavelength (peak 370nm) is reasonably priced and somebody can help me with schematic that uses one in the first place.
     
  12. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    The LED will be used to perform Fluorescence tests, as UV light source.
     
  13. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    No, but you are the one asking for help. Think about it.

    The comment about the light tree is a bit obnoxious, and you got oodles of help previously that you blew off. I just reviewed the previous thread...

    http://forum.allaboutcircuits.com/showthread.php?t=13191

    Why should anyone waste time helping when their advice is ignored? My time is as valuable to me as yours is to you. You want perfection, their isn't such a thing this side of the grave.

    I'm a firm believer in this being a no flame zone, so I'll leave it at that. Common courtesy goes a long way.
     
  14. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The LM324 is a low voltage quad opamp. The LM358 is the same but is a dual opamp in an 8-pins package.
    The LM324 has pin4 for its positive supply. Don't forget about it. Connect it to +5V.
     
  15. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    Thanks, :)
     
  16. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    This is no flame, it's 3AM at my timezone now do you really think I would waste my time like this if I wasn't looking for solution.

    Please tell me what solution that allows me to drive the specified LED from 5V at I overlooked in my previous thread?

    The only solution I received was to use LM317, then LM317L then the circuit diagram I posted in this new thread. None if these work for 5V. And since USB uses 5V can you explain me what did I ignore then?
     
  17. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    My one attempt for a solution. Use this if you want. The 140Ω resistor would have worked.

    This schematic should compensate for various USB voltage variations, there are always variations to one degree or another, if you are a scientist you know this. Adjust the pot for the current you need, measuring voltage on the 10Ω resistor. 0.2V = 0.02A (20ma)

    [​IMG]
     
    Last edited: Apr 27, 2009
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    ALL LEDs degrade over time. Their brightness gradually decreases with age and use. Can't give you specifics, but the more current you run through them, the more quickly they get dim.

    You're going to need some kind of control sample, where you can test the intensity periodically over time.
     
  19. smilem

    Thread Starter Active Member

    Jul 23, 2008
    132
    0
    Yes I know that, for control sample I use ceramic white tile without brighteners.

    I use only 8mA for my LED(maybe could use even less), more amps and it shines too bright. I only switch it on when takind a sample. So I expect it to be quite stable and work for a long time without brightness changes. The LED is to-18 package metal body with glass window type.
     
  20. DickCappels

    Moderator

    Aug 21, 2008
    2,651
    632
    The original application was a biology experiment in which the cells from a retina were excited by optical pulese. The experimenter wanted fast rise and fall times and wanted the amplitude to be well controlled. The TTL source was a bench-top pulse generator.

    http://www.cappels.org/dproj/ledsw/Fast_LED_Driver.html


    (Thank you for the reference, [​IMG] SgtWookie)
     
Loading...