An apolar LED tester

Discussion in 'The Projects Forum' started by SPQR, Feb 18, 2013.

  1. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Hi all,
    Not a "great" project, but something I liked doing.

    It's always bothered me to have to line up the appropriate leads on an LED to the appropriate leads on an LED tester :D - I sometimes forget - "long is what?":)

    So as an exercise I thought about making an apolar LED tester such that the leads can be inserted either way and the LED will flash if it is good. To learn more about transistors (doing my "transistor phase" now), I started drawing diagrams and building examples, and after a few iterations got this:
    [​IMG]

    Hooray! It works.
    I learned about choosing a transistor based on CE voltage/current needs, and learned about calculating the value of the base resistor.
    You hook up two leads from a microcontroller, one to each of the connected bases, and then alternately pulse the leads 1,0,1,0 etc. Connect an LED between the two poles and "on" "off" "on" "off" no matter what the polarity.:)

    The next issue is a "simple" transistor circuit to alternately turn the bases on and off. The smallest "flasher" unit I could find is the classic two transistor flasher like THIS one.

    Do you think that I could take a wire from each of the LEDs and connect it to the two bases of the LED tester to alternatively turn the bases on and off? (ground goes to ground across the two circuits)
    Any other ideas?

    Thanks in advance.
     
  2. #12

    Expert

    Nov 30, 2010
    16,298
    6,811
    OK...a 2 transistor oscillator to replace the microcontroller. Then your phrasing leaves me stumped. "A" lead from each LED to the bases of the large circuit displayed above? You lost me.

    ps, congrats on the miniature H bridge.
     
  3. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Ok! So I've reinvented the wheel!
    Is that its formal name? H-bridge?

    Here is the same diagram as above, I've just redrawn it with the flasher on top of it.
    [​IMG]
    My question is whether I can take a lead off of the collectors of each of the transistors and have enough current to drive the transistors on the H-bridge.

    If you think it's worth the time, I'll build the flasher and connect it to the tester.

    If not, then maybe there is another "flashing" circuit where I can drive the two bases alternately?
     
  4. #12

    Expert

    Nov 30, 2010
    16,298
    6,811
    Yebut...remove the LEDs from the oscillator so they don't waste half your 5 volts.

    Then the 470 ohm collector resistors are the limiting factor. You can test this idea by just connecting a 470 from +5 to a base input. If it fires up the LED you are testing, the oscillator will work.
     
  5. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Ok, gotcha.

    My confusion in situations like this is always between "voltage" and "current".

    I understand that you need current to drive a resistor's base, but a "voltage" is created by a drop across a resistor in a CE circuit.

    I'll play with the circuit a bit, and see what is there.

    Thanks again!
     
  6. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    He could leave one in somewhere as a "oscillator working" light. :confused:
     
  7. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    We used to make things like that with a 555 timer. It has a push-pull (half bridge) output that is fine to operate (flash) a LED in either direction.
     
    SPQR likes this.
  8. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    I thought about the 555, but I was trying to keep it all in transistors as a learning experience.

    I looked for single transistor oscillators, and couldn't find one, then found the two transistor above.

    The bad news is that I connected both of the common base pairs together and tried to drive it with a single lead from the microcontroller, and it wouldn't work.
    Need to think about why - that'll teach me more about transistors.
     
  9. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    There aren't any common base transistors in your circuit above.

    4 common emitter for oscillator and "low side", and 2 common collector/emitter follower for high side.

    To figure out which is used, cover the input with one finger, and cover the output with the other finger. Whichever pin is left is the common (Base, Emitter, Collector).
     
  10. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Isn't a half bridge marginal with a 5V supply? How would you make it work with a 3V LED?
     
  11. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Just threw this together. Why not cross connect the LEDs in the classic Multivibrator?

    Very Sloppy, Real components and stuff would help, but it gives you the alternating polarities.

    [​IMG]
     
  12. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Funny, I was going to propose the same circuit, but I would probably use 300Ω collector resistors and 3k base resistors. 300Ω would yield about 10mA through a 2V LED, which is plenty to test it. The 3k base resistors ensure saturation.

    Keep in mind that there is only one LED in the circuit, and it will light regardless of polarity.
    Do any LEDs have reverse breakdown<5V?
     
  13. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Not that I know of. I'm sure we'd have heard of it with all the Charlieplexing going on with every type of LED made.
     
  14. RamaD

    Active Member

    Dec 4, 2009
    254
    33
    That is a very good point.
    I just remembered my first testing of 50 LEDs purchased fresh from the store, with a cheap analog multimeter -9V battery powered, during my college days (dont ask me how many decades back!).
    I was not expecting anything defective, but to my surprise I have been getting defective ones; I started keeping the good ones and bad ones in two separate piles. I was wondering whether I was testing them correctly on seeing that the defectives are growing as much as the good ones. Then I retested the good ones(!) and moved some more to the defectives. Then I was suspecting myself whether I was putting the good and the defectives correctly! Then I stopped and went back to college to check the databook (so expensive, could refer it only in college) the next day to find the PIV to be 5V!
    I had destroyed more than 2/3s!
     
  15. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Sorry for my tardiness, I was out of town for a day.
    I'm sure I'm using the wrong nomenclature, but by "common base" I just mean the two transistor bases on the left connected together so I can turn both of the transistors on at the same time.

    I'll give everyone the iterations I went through to make this thing.

    1- One NPN transistor, just to see if I could turn on an LED - I was surprised that I could turn an LED on above and below an NPN transistor.

    2 - Two transistors (NPN and PNP) trying to get current to flow through one section in both directions - no luck

    3 - Saw a "push-pull" amp and noted the line connect between the transistors and had an "Aha" moment - one wire, current flow both directions! - but couldn't figure out how to make the opposite end of the LED change to positive and negative with the push-pull transistors.

    4 - Then thought of two transistors with the LED between the collector of one and the emitter of the other - it seemed to work logically - and it worked when I built it with two NPN transistors.

    5 - Tried same in PNP transistors, and it worked.

    6 - When I put the two "sides" (PNP and NPN) together, it didn't work.

    7 - thought that I could use two "sides" of both NPN transistors just by changing the wiring on one side - it worked! (that's the diagram at top)

    The key to getting current to flow through the one pair of transistors was turning them on at the same time - so I connected the bases together, and it seems to do fine.

    But when I took the next step - conntecting all FOUR of the transistor bases together - the circuit just sits there.
     
  16. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Now THIS is interesting!!!!
    One wire in the midst of two transistors - like the push-pull I mentioned above.
    I'll try it when I get home!

    I should be able to crank down the voltage, adjust the non-RC resistors, and not torch any LED's - do you think?


    Excellent discussion as usual!
    Thanks to you all!!
     
  17. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    You may have to lower the base resistors to get the transistors to saturate. Bear in mind that, with only 1 LED (not two antiparallel as shown in the schematic), your duty cycle will not be 50%. This should not be a problem, as the flicker is invisible anyway.
     
  18. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    Ok, I'll experiment a bit.
    I think the "flicker" is actually required, wouldn't it be? - since the LED in either polarization would only be on for 1/2 cycle.

    I could slow down the RC circuit to oscillate at about 500 msec.
     
  19. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Yes, the LED will only be on for a portion of the cycle. Why do you need to see the flicker?
     
  20. SPQR

    Thread Starter Member

    Nov 4, 2011
    379
    48
    I think I see your point.

    I was thinking about running the oscillator at a low frequency so you could "imagine" the current flowing forward (on) backward (off). But that really isn't necessary.

    So I'm guessing your idea is to have it oscillate fast so you can't see any flicker, and the LED seems to be on all the time.

    I can put pots and different caps in the RC portions and see how things go.

    Thanks.
     
Loading...