Urgent: LEDs sync'd to sound for Halloween costume

Discussion in 'The Projects Forum' started by 4thisguy, Oct 29, 2015.

  1. 4thisguy

    Thread Starter New Member

    Oct 29, 2015
    4
    4
    Trying to be a good dad here, but put the project off a bit too long and now have to finish his robot costume by the morning. I'm building a small circuit from components to make an LED flash with sound (so his robot costume will make sound and light up). I'm using the circuit labeled A in the picture below, and it works...but only with the volume turned up very load. Since he will have a speaker on the costume and it will be in his school, it can't be too loud. Is there a modification I could make to this circuit (ie change out a resistor or two) to make it more sensitive to the input? Or, I have some PNP and a NPN transistor left over, could I build a pre-amp? Any help would be MUCH appreciated, because it's got to get done tonight!
    Thanks,
    Guy
    [​IMG]
     
  2. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    No promises but, drop the first resistor (1K) to 100 ohms and see what happens.

    This isn't a brilliant design because the first transistor needs a few tenths of a volt to even start, It's not like it's idling along and waiting for a bump. It's dead off and needs the whole first half a volt to even start to sense an input.

    The second circuit is idling along waiting for a bump. That makes it a lot more sensitive to tiny signals.
     
    GopherT, KJ6EAD and cmartinez like this.
  3. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    I've simmed circuit A (though using different transistors, since I don't have the ones in the original circuits in LTspice) and you're right, the LED doesn't turn on until the input transistor's gate reaches 5V. Now I'm going to sim circuit B, see what happens.

    Capture.JPG
     
    djsfantasi, GopherT and #12 like this.
  4. 4thisguy

    Thread Starter New Member

    Oct 29, 2015
    4
    4
    Thanks for the reply! Unfortunately, lowering the input resistor didn't work. I'm thinking about trying option B, because you're right, that might be more sensitive.

    I'm also wondering if it would be too hard to add an extra transistor to the beginning of the circuit to further amplify...
     
    GopherT likes this.
  5. 4thisguy

    Thread Starter New Member

    Oct 29, 2015
    4
    4
    cmartinez and GopherT like this.
  6. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    I don't think you need more gain, you need more sensitivity. Look at how hard that current is slapping on and off in post #3. You already have at least a gain of 10,000 but the start voltage is way too high.
     
  7. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    Now I've simmed circuit B... and the truth is that I don't fully understand what I'm seeing. @#12, could you please take a look at the graph and tell me what's going on? The green trace represents input voltage at R3, and the yellow trace is current flowing through the diode.

    Capture.JPG
     
    djsfantasi and #12 like this.
  8. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    I don't sim, so I am not up to speed with you. I can say it shouldn't take much voltage (2.4 millivolts) at the input to overcome the 2 or 3 microamps coming down the 1 meg resistor. A tiny amount of negative going signal should slap the first transistor off (which turns the second transistor on).

    You're running your ramp at 10 Hz. Way too slow for the coupling capacitor.
     
    djsfantasi and GopherT like this.
  9. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    I lowered V2's cycle to 2V and it makes the LED's current flow at a steady 18mA, so yes, circuit B is more than twice as sensitive as circuit A, though it's working in an inverted sort of ways.

    Capture.JPG
     
    GopherT and #12 like this.
  10. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    I'm just beginning to understand now... Yes, what you're saying is what's happening, though the input signal has to first peak to at least 2V before the LED burns steadily, otherwise it won't receive enough current. So yes, circuit B might be worth a try.
    (small note: I didn't know you didn't like to sim)
     
    djsfantasi, GopherT and #12 like this.
  11. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    More like never learned. For small analog circuits, my brain is way faster than setting up a sim.
     
    cmartinez likes this.
  12. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    There, here's what things look like running at 100 Hz. I'm under the impression that the LED would be off for too short a period to be aesthetic

    Capture.JPG
     
    GopherT and #12 like this.
  13. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    It will be off during the times the music is between beats (if it's adjusted well).
     
  14. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    And the means of adjustment would be the volume knob?
     
  15. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    As mentioned in post #1.

    ps, the coupling cap is equal to R3 at 1.6 kilohertz. Fine for wide band audio but picky for a one frequency sim.

    See that? I did a Frequency sweep analysis inside my head.:D
     
    cmartinez likes this.
  16. 4thisguy

    Thread Starter New Member

    Oct 29, 2015
    4
    4
    Gave circuit B a shot and it is, indeed, much more sensitive. I think it will work well (enough) for the costume (high stake, I know). Seriously though, you guys probably saved me a few hours of playing around, so it's much appreciated!

    Any suggestions on how long my set of AA batteries will last if my son has this thing running all day? Suggestions to improve battery life?
     
    cmartinez likes this.
  17. cmartinez

    AAC Fanatic!

    Jan 17, 2007
    3,574
    2,543
    Glad to hear things are working out for you... since you have a very limited time frame, I suggest you carry a few spares with you to quickly replace the pair that has run dry.
    Let me check the sim, see if between #12 and me can give you an estimate of battery life.
     
    #12 likes this.
  18. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    About 100 hours (based on NiMH @ 1000 mahr rating).
     
  19. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    WTF? A 20 ma LED at 50% duty cycle can run for 100 hours if you have a 1000 ma hr rated battery set.
    You are suffering from delay time caused by too much high tech!:D
     
    Sinus23, djsfantasi and cmartinez like this.
  20. #12

    Expert

    Nov 30, 2010
    16,320
    6,818
    Don't pout, @cmartinez
    You have a new toy and I'm smoking you on simple stuff. In a month you will be simulating circuits that I couldn't do if I had all day.;)
     
Loading...