Building a driver for 12v 10w LED

Discussion in 'General Electronics Chat' started by Arius007, Sep 2, 2011.

  1. Arius007

    Thread Starter New Member

    Sep 2, 2011
    14
    0
    I've searched quite a bit for a circuit that might do this without success so I'm hoping someone is willing to guide me in the right direction with this.

    I've bought some high power LEDs, rated at 12v, 1amp with 700 lumens rated output.

    I've tested them and found as follows:
    They have no internal resistor and will draw infinite current;
    They will turn on at approx 7vdc and with as little as 1mA they begin to light;
    They provide good brightness from 10mA upwards to 750mA so I expect to drive the LEDs between 10mA and approximately 750mA.
    The light output near or above 1A is dangerously bright and may need to be limited if I don't go for a diffuser.
    I expect to use 4 LEDs in parallel - 4 x 750mA = 3A = half of a TIP41A rating. The TIP41A NPN transistor is rated for a continuous collector current of 6A so I've bought some of these too.
    Supply voltage will be approx 13vdc - 14vdc.

    Here's what I'm trying to do:
    Step 1> I'd like to create a small driver circuit using a TIP41A. I was thinking of limiting the base current with a 10R in series with a 100R pot.
    Q1: What else should I include in this part of the circuit apart from the TIP41A and the base resistors?

    Step 2> I'd like to switch this driver with a MOSFET for flashing (approx 5Hz max, low duty cycle). I was thinking of switching the base of the TIP41A using a BS170 together with a pull-down resistor to do this.
    Q2: Will this work?

    Any other thoughts?

    I have a good knowledge of components but only a very limited knowledge of circuit design.

    Thanks in advance for any assistance.
     
  2. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    I think that I would use a separate constant current driver for each LED & switch all with one PNP or P-ch FET. See Bill Marsden's Blog for constnat current drivers.
     
  3. iONic

    AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    The make and part number of the LED's would be most helpful or a datasheet for reference.
    Many here often refer members to a device call a BuckPuck. I've used then with great success and they are a darn safe and efficient way to protect your LED investment.

    BuckPuck

    It can operate up to 12 1W LED's or 6 3W LED's. With the addition of a simple PWM circuit you can be ready to rock!
     
  4. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    4,770
    970
    This comment scares me.. High power LED's are very dangerous to look directly at for more than a few seconds. Permanent damage may/can occur. Please think twice about looking into them again and protect whatever you are doing with a proper diffuser.
     
  5. Arius007

    Thread Starter New Member

    Sep 2, 2011
    14
    0
    Thanks for the replies.

    I read through all the LED articles there before I posted but didn't see anything that looked like I could easily adapt to this. I know there's not much to what I'm looking for but my circuit design experience is limited.

    I have seen comments about 1 LED per circuit and I'd have no great difficulty with that assuming I can drive multiple transistors together.


    Unfortunately I didn't get one at the time and it doesn't appear to be available, thus the testing I carried out. This is the best I can do for now:
    http://www.ebay.co.uk/itm/ws/eBayISAPI.dll?ViewItem&item=220439870435&ssPageName=ADME:L:OU:GB:1123

    I'll have another look at the BuckPuck.
    I'm expecting the PWM will make it too complicated at this stage - I'm assuming the PWM would go on the output to each LED, which could be quite a few, and then I'd need to switch that same output on/off using a MOSFET or similar.
     
  6. Arius007

    Thread Starter New Member

    Sep 2, 2011
    14
    0
    No need to be scared...There won't be anyone looking at them until the light level is either low enough or suitably diffused to be safe. That said, thanks for the concern/advice.
     
  7. iONic

    AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    You are using a heat sink I hope! It's not really an option.


    Nope, just one PWM circuit 555 based is all that is needed to vary all LED's at the same time...unless your planning differently.
     
  8. Arius007

    Thread Starter New Member

    Sep 2, 2011
    14
    0
    Yes and No. Yes, I will but not just yet - I've only run one LED for a few seconds a couple of times to check current draw.


    It looks like I'd need a BuckPuck for each LED as I expect to run them at up to 750mA. I don't doubt your recommendation of the BuckPuck but buying ten of them is a bit pricey.
     
  9. iONic

    AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    I think with a 24V - 30V supply you will be able to drive 2 - 3 in series.
    Your original post only mentioned 4 LED's. But I do believe that the dimming can be done all at once, in other words, you would not have to build a dimmer for each LED. It might even be able to operate all 4 in series with the 1000mA version and only driving them to 750mA. I can't say for sure on that. It may be worth calling them about it.
     
  10. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    Plain LEDs, high power or convention, doesn't have resistors and doesn't limit current. This is fundamental to LEDs in general. I'm thinking of started round 3 in developing a SMPS current limiter, but for general purpose a transistor current limiter would work OK. Need a schematic idea?
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Transistor aka linear current limiters/regulators for 1W LEDs are going to waste a lot of power. There's not much of an in-between; either linear, simple and wasteful, or switching, efficient, complicated and hard to understand. There are switching current controllers available that are low parts count, but unfortunately most of them are tiny SMT's, so difficult for hobbyists to use.
     
  12. Arius007

    Thread Starter New Member

    Sep 2, 2011
    14
    0
    I tested a simple circuit today with the TIP41A0. I put a 100R in series with a pot to drive the base from an input switch connected to the +12v line. I also put a 10k from the base to ground to pull down the base when the switch is open. The LED ground was connected to the collector and the emitter directly to ground. This worked quite well. The pot varies the LED current and light output as expected.

    I couldn't achieve fine control because I don't have the correct pot yet (I want to use a 100R but only have a 1MΩ at present), it should arrive Monday, but it appears to be more or less working.

    I then connected a separate flasher circuit that I intend to use. I put this in place of the switch and again it worked well. It provides a +12v pulse so I fed it through the 100R & pot into the base.


    There was no noticeable heat whatsoever in the TIP41A. What I did notice though was the current the circuit is drawing is much higher than I would expect, pulsing up to 2.5A or maybe more - I'm wondering is there some kind of inrush into the base of the TIP41A? Should I try to limit this with a coil or something?

    Another test indicated that the base should be getting about 15mA when the collector is passing about 600mA. However when I replaced the 100R with a lower wattage resistor, the resistor started to heat - my calculations show the resistor should be seeing about 0.015w - not enough to be heating. Does this support the inrush theory or am I missing something else?
     
Loading...