Question on High Powerr LED flashers...

Discussion in 'The Projects Forum' started by pcon2009, Dec 24, 2012.

  1. pcon2009

    Thread Starter New Member

    Aug 5, 2012
    18
    0
    First off, I know there has been discussion on flasher circuits for high power LEDs, and it is something I have been working on for awhile (with a couple of previous posts on this forum looking for help). I shelved the project for awhile because I simply got too frustrated with it. I went back to it today, and I have some questions. NOW I am NOT looking for a solution, per se, but the education/knowledge as to WHY I can or can't do things certain ways. So let me start with a general synopsis of my experience so far, and maybe a few of you can answer some of my general questions?

    In my first attempt, I wanted to do something based around an LM555 like such:
    [​IMG]

    I understand that this works for standard LEDs becuase they don't draw a large amount of current like high-power LEDs. The 555 can only handle a max of 200mA I believe. So when I look at the spec sheets for the LEDs I am using (which are THIS and
    THIS I would need it to be able to hand 350-400mA on the RGB (color-dependent) and 750mA on the white. So, this solution doesn't work simply because the 555 can't handle it, correct?


    The second solution I looked at would be something like this:
    [​IMG]

    So, I built one of these with some spare parts and low-power 20mA red standard LEDs and it worked well. So out of curiosity, I hooked in the high-power LEDs (subbing in the appropriate resistors instead of the 470Ω) and it worked. But I noticed the flash rate went WAY up. I thought about it, and presumably this is because the higher current-draw was draining the capacitors faster. So, I opted for high value caps and tried some 100μf and 47μf values just to see what would happen. As predicted, the 100μf caps slowed the flash rate considerably, too slow for what I wanted, but the 47's did the trick. So all seemed to be well, it seemed as though I had everything figured out.

    Next day, I hooked the circuit back up again to continue working on it, and wanted to try something more. I built 2 of the "fast" (47μf cap) circuits and a "slow" (100μf cap) circuits each on their own mini perfboards so that I could see if connecting the fast ones to each "output" of the "slow" one would work as expected - flash two LEDs quickly on one side on the slow board, then the other two LEDs quickly on the other. This lead to some interesting results. The LEDs flashed, but in a seemingly random pattern that was all over the place. Why would this be? Is it just that the drain on the caps isn't consistent enough to keep the flash rate consistent?

    Final question: when I went back to the project the next day, now NONE of the circuits seem to flash the high power LEDs anymore... they will still work with low power LEDs, though, so I don't think I "fried" anything (like the transistors or caps), but basically, no matter which circuits I use the LEDs just all light up simultaneously. At this point, I have become frustrated and am ready to shelve the project again - i just can't seem to wrap my head around any of this. I am new to the whole idea of building my own electronics. I am quite capable when it comes to taking a schematic and building it, and my soldering skills are decent enough to not be making lots of bad connections or anything, but when it comes to reading about MOSFETs and PuckBucks and the difference between an NPN and a PNP transistor with sinking and sourcing currents, my head begins to spin.

    Can anyone help me out and try to set me straight here? Am I just going down a never ending road of projects I should never have started?
     
  2. droggie

    Member

    Oct 21, 2012
    137
    1
    I just finished building a circuit with that exact same layout, except the two 100k resistors were replaced with two tri potetiometers.
     
  3. pcon2009

    Thread Starter New Member

    Aug 5, 2012
    18
    0
    What were the results with the potentiometers? What did adjusting the resistance do, did it change the flash rate? Were you using the same 10uf caps?

    Also, the big question - what LEDs are you using? Are the also high power or just regular? Because this circuit does work well for the low power LEDs but for me has unpredictable results with the high power chips.
     
  4. droggie

    Member

    Oct 21, 2012
    137
    1
    Yes, I used 10uf caps and each pot had its own LED and the flash rate changed. I used regular powered LEDs. I attached the schematic.
     
    • aa.jpg
      aa.jpg
      File size:
      161.7 KB
      Views:
      24
  5. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    You can see the circuit as below, it's working fine with 9V power and 3V LED.
    3V LED/20mA(rating) but used it around 16mA.
    If the MOSFETs change to the PNP,NPN transistors is also ok, but when you want to use the LED have heavy current, then you better use the MOSFETs.

    [​IMG]
     
    Last edited: Dec 26, 2012
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    ScottWang,
    Your schematic has a few problems; the first of which is that the gates of the MOSFETs are tied together, you have a high-value (4.7k) resistor in the gate charge/discharge path, and the drains of the two MOSFETs are connected together. Very generally, MOSFETs take a good bit longer to turn OFF than they do to turn ON; so the way you have them wired, you will wind up with the dreaded "shoot-through" condition where both the upper and lower MOSFETs are turned on at the same time, limited only by the Rds(on) of the MOSFETs and the impedance of the power supply. This is a very un-good thing.

    I've attached a modified version of your schematic.
    On the left, you'll see C1 from the threshold and trigger inputs to ground as usual, but then you'll note C2 from the threshold and trigger inputs to Vcc/9v. Since C2 is roughly 1/2 the size of C1, the threshold and trigger inputs will start off at 1/3 of Vcc instead of ground potential. This makes the first pulse about the same as all subsequent pulses, rather than the first pulse being much longer.

    Note R3/D1 and R4/D2; these cause the MOSFET gates to discharge more rapidly than they charge, which causes them to turn OFF more quickly than they turn ON. Note the addition of R7. A bjt (transistorized) 555 can get within ~200mV of ground, but can't get above about 1.3v less than Vcc, due to the Darlington follower output configuration of the 555 itself. The addition of R7 provides a current source to enable the output to rise to nearly Vcc.

    The drains of the MOSFETs are no longer connected together, which removes the possibility of shoot-through entirely.

    Not shown are the bypass capacitors across the 555 supply pins. The datasheets indicate that you should place a 0.1uF poly or ceramic cap in parallel with a 1uF or larger aluminum electrolytic cap across the supply pins. If you omit the bypass capacitors, you are likely to have intermittent or consistent problems.
     
  7. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Whenever I see a "9V" supply on a schematic I think of a little low current "transistor" battery that goes dead in a few minutes when it has a high current load.
     
  8. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    Thanks SgtWookie.
    I also thinking over about the circuit what you posted, the similar way is the most time that I did, and now the circuit that I posted is another idea from internet, in the beginning that I had a little worried about the circuit, so I decided to do a testing on breadboard, but every thing seems ok, even the power side is didn't affected at all, it could be the current is low less.

    When I saw what you said, I went to redo again and checked the MOSFETs temperature, it's really hot, the dead time could kill the MOSFETs, but the last time that I forgot to test the temperature, that is the most important thing that I always care about it.

    So this is a really bad idea, I apologize for this.
     
  9. pcon2009

    Thread Starter New Member

    Aug 5, 2012
    18
    0
    ScottWang and SgtWookie thanks for your ideas!

    SgtWookie, is the circuit you posted (updated from ScottWang's) something that would then work for the higher powered LEDs?

    It looks like to me that it is kind of a "combination" of the two circuits that I originally posted, and that the 555 is driving the transistor circuit to power the LEDs?

    Couple of questions: you mention using MOSFETs... What exactly are these? If I understand they are a type of transistor that is special for some reason? I am new to the electronics world so any explanation here would help.

    In addition, I wanted to clarify that I will be using a 12V supply - would this just be a matter of changing the values of R5 and R6 (the ones that connect to the LEDs directly) to a value for 12V? I usually use THIS calculator for my LED resistors.

    From here it definitely looks promising and is something that I would like to try, I just want to make sure before I go buy more components that I actually understand totally what is going on here, because as I have said before, the project is more for the learning experience than the end result .
     
  10. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    SgtWookie's posts isn't update from me, that is the normal way to do so, and I were used combination, in the low current that I already figure out the way to reduce the current through the Up Mosfet and low Mosfet, but if using in the big current that still not solve the problem of dead time, the simple circuit can't do that, but I will still keep to try.

    The calculator for the resistor of LED, the formular is oK, but when concern about the brightness and the life of LED, we do not using in the rating current of the LED, as you have a 3V/20mA LED, you can use 80% of 20mA, it will about 16mA, so when you use the calculator, you just set the current to 80% of the rating current to calculate.

    The calculator just calculate the Quiescent Current for the LED, when a LED working in the flashing circuit, because the duty cycle maybe not in the 100%, the current of LED is less than Quiescent Current, so the resistor of limited current can be decreasing some more.

    You can read the below articles, it will help you to know more how A Bjt and MOSFET using a as switch, and the MOSFET is good for the high current load, post your questions when you reading the articles.

    Transistor as a Switch.(BJT)
    http://www.electronics-tutorials.ws/transistor/tran_4.html

    MOSFET as a Switch.
    http://www.electronics-tutorials.ws/transistor/tran_7.html
     
  11. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    No.
    Most ordinary LEDs have a maximum allowed current of 30mA to 50mA. Their brightness and many hours of lifetime is spec'd at "only" 20mA so operating them at 20mA is fine.
     
  12. pcon2009

    Thread Starter New Member

    Aug 5, 2012
    18
    0
    Cool! That's good to know Audioguru.

    So do we have a consensus that the above schematic posted by SgtWookie should work for the higher powered LEDs? And which resistors would I need to change then for the 12v supply instead of 9v? Just the ones connected directly to the LEDs?
     
  13. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Yes, the value of the current-limiting resistors must be increased. But since the supply voltage is much too high then the resistors will get extremely hot.

    Power in a resistor (amount of heating) is the current times the voltage. Since the voltage is increased then the power (heating) is also increased.
     
  14. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    You can decide the lifetime of your LED, you want them to used for one month, two month ,or six months ,or over year, I would prefer keep them logner.

    If you want to save money then you should care about the light fades for LED light.

    http://www.simaoled.com/english/news_show.asp?mid=6&nc_id=1&n_id=479
     
  15. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
  16. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    Unless you tell the OPs to buy the LEDs has high quality, then it will like you as mentioned that I knew, but many situations is that OPs didn't know that or hard to buy or too expensive, as I met on the internet, many people has the same problems that the LEDs easy to burn out or light fades.

    So the situation maybe won't happend on you, but too many people on the internet always has the similar problem everyday.

    I ever heard someone overcome the heat problem of the LED, and also he applied and passed the patent, so he can used many LEDs in the over rating current as much as more he can.
     
Loading...