Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

LED's in Series Vs LED's in Parallel

  1. May 8, 2013 #1
    This should be a simple question yet I am struggling to see the benefit of using one over the other.
    I was building an LED and LDR based sensor head for a line following robot, and I needed to choose a method of connecting 3 LED's inside of it. I figured that the 3 options are:

    1) All in series with a resistor

    2) All in parallel, with the parallel bit in series with 1 resistor

    3) All in parallel, each with their own resistor

    4 important points:
    .The LED's are rated at 3V, 20mA, 60mW. This is what I am aiming for.
    .The LED's should all be as bright as possible, and more importantly they all need to be the same brightness (so that the sensor head is accurate).
    .The 5V is set, this cannot be changed, however the resistors can be any value at all.
    .I'm guessing the LED's are 150 Ohms as 3V/20mA = 150 Ohms, but the resistances of each may slightly vary due to manufacturing inaccuracies.



    The voltage and current through the resistors can be controlled equally well in each case as the value for the resistor in each case can be changed.

    I was told that option 3 is the best option, and option 1 is completely useless, but I don't see why. Is it something to do with the fact that a diode does not have a linear I-V characteristics?
    0VGk7F4.FiPeOdnS12jRdw.jpg
    So the relationship between current and voltage is linear like with a resistor. I have always assumed that this doesn't matter for an LED, as the main reason there is a voltage across this is because energy has been converted to light, (like a resistor does as heat). This is unlike a diode whose "resistance" changes as you change the voltage across is.

    Please help!

    Thanks!
     
    Last edited: May 8, 2013
  2. jcsd
  3. May 8, 2013 #2

    phinds

    User Avatar
    Gold Member

    You are saying that each LED requires 3 volts, yes? If you put 3 of them in series, how many volts would you need?
     
  4. May 8, 2013 #3

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award


    Do you see the problem with this scheme ??

    look at your supply voltage 5V, Vs you voltage drop for each LED 3V and the resistor


    thats why parallel LEDs are better if there's going to be a lot of them


    Dave
     
  5. May 8, 2013 #4

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    hey phinds SNAP LOL

    you typed less ;)

    Dave
     
  6. May 8, 2013 #5
    Im trying to understand this:

    Why would you ever put lights you wished to be of equal "brightness" in series?
     
  7. May 8, 2013 #6

    Integral

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    A instrument I build has LEDs in series. But this is to guarantee that either both are working or neither.
     
  8. May 8, 2013 #7

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award



    Ohhh its no problem 3 LEDs in series and with a current limiting resistor work well
    all the same brightness.
    But as the OP will discover he will need more voltage ;)


    Dave
     
  9. May 8, 2013 #8
    Ah yes, it would be impossible to have 3 Volts across each in series, as this would add up to more than 5 Volts. I see now, thanks for answering.

    Although I'm still not sure why each having their own resistor is better than just having one in series with the all the parallel resistors? What is wrong with having one resistor that splits the voltage between itself and the parallel part as 2V and 3V?

    If they are all of equal resistance (or at least very close) then the voltage across all of them will be the same, and as they are in series the current through all of them will be the same, so they will all be equal brightness, no?
     
  10. May 8, 2013 #9

    berkeman

    User Avatar

    Staff: Mentor

    With newer "matched" LEDs, you can put them in series. You do however, need enough voltage headroom to make it work, as pointed out by other posters.

    The disadvantage is that matched LEDs tend to be a bit more expensive. The advantage is that you can run them from a higher voltage, and lose less power in the series current-setting resistor for each series string.

    Series LED configurations are common in LED lighting applications, where you may have several strings of series connected LEDs (each with its own current setting resistor or low-side current source transistor) placed in parallel.
     
  11. May 8, 2013 #10
    Yeah, I have to much WCAD (who cares attention deficit) to work the simple math stated by the OP. I'm going to guess equal "brightness" is on the less than probable result of the hypothesis
     
  12. May 8, 2013 #11
    Guess what I'm really trying to say is in the situation where the "brightness" is so paramount to the problem itself that I would never consider series with a 5V supply. We can assume your supply is VDC? Yes, I know you can also go AC with LED. Just curious...... what happens to the equation as your supply begins to deplete it's ampacity???????
     
  13. May 8, 2013 #12

    berkeman

    User Avatar

    Staff: Mentor

    Actually, the best way to do this for good battery life would be to do a boost DC-DC circuit to drive the 3 LEDs in series, and use current-mode control (with a small low-side sensing resistor) to keep them operating at 20mA... :smile:
     
  14. May 8, 2013 #13
    Yeah the supply is from a 5V DC battery (a pretty big rechargeable one). This battery is also powering a motor drive circuit (in parallel with the LED circuit).
    As the source depletes the voltage across it will decrease, and therefore the voltage across everything in the circuit will deplete, and hence the current through everything will deplete, so they will become dimmer.

    I've never heard of a boost DC-DC circuit, but my robot is running about of space to mount new circuits anyway so it probably couldn't be done. But out of curiosity, how do they work?
     
  15. May 8, 2013 #14

    berkeman

    User Avatar

    Staff: Mentor

    http://en.wikipedia.org/wiki/Boost_converter

    They are one of the topologies of DC-DC converters. Normally you use voltage feedback from the output voltage supply to control the duty cycle of the switching action. But in LED applications, you put a small-value resistor at the bottom of the string of LEDs, and use that sense voltage to regulate the current out of the boost DC-DC to the desired value (20mA in your case).
     
  16. May 8, 2013 #15

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award


    google DC - DC Boost Converters.
    And for your interest also look at DC - DC Buck Converters, for dropping and regulating voltage
    They are much more efficient than linear regulators like the LM78xx series.

    Dave
     
  17. May 8, 2013 #16

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award


    also for the obvious reason .... to get 3 times the brightness out ;)

    I have a commercial product I produce that has 3 sets of 3 LEDs in series
    all hi-intensity 5mm LEDs, 3 sets of red, of green and of blue


    Dave
     
  18. May 8, 2013 #17
    I will do thanks, I should have done that before asking really haha.

    Just one last question though:

    I was told option 3 is better than option 2, do you know why this is? To me it just seems like a waste of resistors.
     
  19. May 8, 2013 #18

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    well yes more resistors ... but the cost is irrelevent, they are so cheap.
    but if you are just going to use 1 resistor supplying ALL the parallel branches, it will be a higher wattage to handle the extra current. and if it fails then your whole display dies where as if you have individual resistors in each branch and one failed then you would only that one branch


    Dave
     
  20. May 8, 2013 #19
    Let's do this for fun please????

    Debate this topic from the true stand point of modern engineering. No, I don't mean academic theory vs real world experience.

    I mean, the cheaper the better!!!!

    Now pretend you own the market share and like you are trying to gain market share.

    Accounting balance sheets lol
     
  21. May 8, 2013 #20
    Dave, why not the1970's christmas light argument on the series option? Which is more efficient? Do the math, I don't want to. Which loses the most due to heat? Which ever one is mathmatically using less wattage is the option I pick. Would you consider that a load resistor? :devil:
     
  22. May 8, 2013 #21

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    several of us have already said why series option is bad( not as good)


    Dave
     
  23. May 9, 2013 #22

    berkeman

    User Avatar

    Staff: Mentor

    But you said you use that configuration in your product...no?

    And I mentioned how many LED lighting applications use series connected matched LEDs, with DC-DC converters regulating the current of each series-connected string.
     
  24. May 9, 2013 #23

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Why not? If they're in series, they would have the same current through each - which would make them equally bright - if they are of the same type.
     
  25. May 9, 2013 #24

    davenn

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    No, not a christmas tree light configuration where there would be many many LED's in series

    My units are a calibrated sound monitors for clubs and pubs etc to monitor noise level and keep it under the level required by local council ordinances.

    My sets of hi intensity LED's 3 red, 3 green, 3 blue are on a remote display case separated from the main base unit by 15 - 30 metres of cable ( depending on the installation circumstances). They are also being driven by 24VDC that overcomes cable voltage drop over that distance

    Anyway this is getting way off the original point .... the OP not realising that he wasnt going to power 3 x 3V LEDs from a 5V supply ;)

    Dave
     
  26. May 10, 2013 #25

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Re-reading this thread I can't see there's any argument about how it must be done. If you only have 5V available then you can only connect 3V Diodes in parallel. Using separate series resistors for each will probably ensure that their brightnesses are nearer the same value and you won't 'waste' any more power.
    There are practical reasons not to bother with a DC-DC converter (cost for a start).
    If you are worried about battery life and what happens when it starts to die, the volts will dip, of course but most rechargeables, these days, maintain a good voltage until they are nearly dead so I don't think it's worth worrying about that too much. Just buy a bigger battery which will last long enough.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook




Loading...