Is there something wrong with this circuit?

Discussion in 'The Projects Forum' started by GRNDPNDR, Apr 4, 2012.

  1. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    I've built this circuit, and it works ( could use some minor tweaking) but I'm wondering if it's actually built "correctly".

    This is someone elses design, and it's meant to take a piezo transducer signal from a kick drum pad, and convert it to an on or off signal.

    It does in fact seem to do just that, however in simulation I get errors when it runs and it's part of the main design of which this is just a sub-circuit.

    It simulates fine by itself.

    More importantly in reality, when there is no diode on the output the LED will light up constantly, just by touching a wire, or moving it, or plugging it in. when it's completely plugged in the LED will stay on constantly without the diode.

    so is this built correctly? is there anything that can be done to stop the LED from flashing when it's touched or moved?

    when it's fully plugged in without the diode it is somehow completing a ground loop from the drum module that is basically shorting J19.

    I would really appreciate some input here, even alternative designs that will convert the piezo signal to an on or off.
     
  2. #12

    Expert

    Nov 30, 2010
    16,304
    6,814
    Without D1, continuity to ground at J19 will allow the LED to be on.

    The LED is constantly enabled by R25. It's just waiting for a ground at the point by the label U4. Are you missing this idea?
     
  3. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    No I didn't miss that, but my understanding of transistors isn't great.

    1.) the output into the BASE passes through the transistor to the collector?? that goes against what I do know of transistors.

    2.) I can't really see why the designer put that LED cathode to the collector WITH the signal output other than to avoid using another transistor (not a terrible idea)

    If I'm not mistaken isn't feeding rail voltage to the signal output (J19) so that with the signal, is also 9V?

    It might be part of the design because the module that uses this signal is simply waiting for an on or off signal, so that might be a way to boost the voltage maybe?!

    If I'm not mistaken the piezo transducer thats plugged into this circuit will put out around 10V.

    At any rate I don't think the 9V being fed into J19 is nessecary for the correct operation.

    Lastly, whats the reason for it being so sensitive to touch and being plugged in without that diode? that in and of itself can't be good?
     
    Last edited: Apr 5, 2012
  4. #12

    Expert

    Nov 30, 2010
    16,304
    6,814
    When the 555 chip puts positive out, it goes through the transistor from base to emitter. That allows a flow from collector to emitter. The LED is powered and waiting for the transistor to ground its current flow. When the transistor conducts, then The LED comes on. When the transistor is not turned on by the 555 chip, (most of) Vcc is available at the junction of the LED and the diode, but no current flows from this circuit to J19 because the diode stops it. Apparently J19 is also waiting for a grounding flow through the transistor.

    9v is not being fed into J19 from this circuit. The diode blocks that from happening. Something else is feeding voltage to J19 and waiting for the transistor that grounds the LED circuit to ground J19.

    This is not a voltage boost circuit. The piezo volts do not get through the timer chip. Being sensitive to touch means you have something loose in your wiring.

    Did that cover it?
     
  5. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    Pretty much covers it, but being sensitive to touch only happens with the diode not in place.

    I would think that because of that it might be picking up static or something like that.
     
  6. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    Sensitivity can be adjusted by R20. I think it is there to adjust false triggering.
     
  7. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    R20 is supposed to "increase pulse width"

    R31 adjusts sensitivity to some degree. It's evidently to help with double hits registered from beater bounce back.


    Those pots I think are what need adjusting, because R20 does work for adjusting sensitivity, if it's lower, I have to hit the drum harder for it to register, and higher will set the drum off with barely a breath. However it doesn't adjust as "finely" as one might like it to, and whenever the pot is at either extreme no hits will register.

    R31 also works in conjunction with this but I'm not entirely certain to what degree, when setting it, it does in fact affect wehther or not I'm getting a single hit with one kick, or multiple.

    On a breadboard I haven't really been able to tune it in to the perfect spot where its correctly registering all hits.

    It should be able to work down to about 4 beats per second (32nd notes?) and ideally should be able to pickup 64th notes but that isn't really a concern right now, just getting it to work correctly with 32nd notes would be perfect.

    I want to avoid the use of a multi-turn pot though because it could get really hard to adjust it again if the settings ever needed to be changed. at least with a single turn you can roughly remember what spot it was in before it was changed.
     
  8. #12

    Expert

    Nov 30, 2010
    16,304
    6,814
    The diode isn't on the "picking up static" end of the circuit. It's on the low impedance end. Put the diode back and figure out what is coming in J19 that is waiting to be grounded through the transistor.
     
  9. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    U need a filter circuit like a LC tank tuned to the beater
     
  10. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    How would I tune that to the beater?
     
  11. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    LC is calculated with a specific frequency in mind. After that u have to find those components which basically consists of 1 or 2 caps with inductors.

    The tuned circuit will only allow that frequency to pass through it and block the rest.

    Google for LC or tuned filters as such. Read about to get the idea.

    U will need this between the opamp and the 555. it does not require power . It's passive
     
  12. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    Wait a sec, is tht a 386. Why the heck is the circuit using a power amp.
    Won't a simple opamp would be a better replacement.

    Try replacing the opamp with something like a LM741 or LM4558. Check the data sheet for proper pin connection.

    It can be replaced without a problem.
     
  13. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    It didn't seem to work well with the LM324 or LM358.

    I don't know the exact frequency range to tune a bandpass filter to.
     
  14. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The LM386 is a little power amp with a voltage gain of 20. It has internal biasing and internal negative feedback. Its DC output is half the supply voltage. R21 does nothing and can be removed.
    The LM386 amp probably oscillates at a few hundred kHz because it is missing the RC network at its output to 0V that is shown on every circuit in its datasheet.

    An opamp has a voltage gain of 200,000 or more and its output will be saturated high or saturated low in that circuit. The opamp will need to be biased properly and have negative feedback added.

    EDIT:
    The maximum allowed input to an LM386 amp is only 0.4V which is being exceeded in this circuit so maybe the LM386 is destroyed. That is why nearly ALL e-drum circuits use a Schottky diode at the input to clamp the negative input voltage to about -0.25V.
     
  15. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    The circuit worked fine when built in real life.... though I did have my suspicions about R21, but if i remove it I get no output.


    So could someone please explain to me then how I should go about adding the RC network needed and add the schottky or how to build this circuit correctly? Because according to the responses, it is not build right even though it does work.

    So how do I get a piezo signal into an on/off signal?
     
  16. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    You have too many threads about your e-drum.
    I posted an e-drum circuit with a Schottky clamp diode at the input on one of your other threads.
    The datasheet for the LM386 power amplifier has about 30 schematics and EVERY ONE has the RC network to 0V at its output.

    R21 in this circuit does absolutely nothing at the output of the LM386 power amplifier.
     
  17. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    I've tried the schottky amplifier design, I didn't get amplified output, anyway that aside I thought this LM386 was used as a comparator, it doesn't amplify the signal it makes it square on the output. If this is supposed to just amplify it then yeah it's probabbly wired wrong.

    I only get half wave output on the schottky design. I need the entire wave.
     
    Last edited: Apr 7, 2012
  18. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The e-drum circuit I posted with the Schottky input clamp diode is used by many people and it works fine. It has a note about simply increasing is voltage gain.

    The LM386 that you show here is an amplifier with a voltage gain of 20. A comparator has a voltage gain of about 200,000 or more. If the amplifier has an input signal too high then it makes a square-wave output.
     
  19. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Did you try reversing the wires from your piezo sensor? Maybe it is connected so the drum hit produces a negative pulse instead of a positive pulse. Following its negative pulse then there will be some lower level positive pulses from resonance that you don't want.

    EDIT:
    Your LM386 is connected as an inverter so that a positive pulse from the piezo causes its output to go low which triggers the 555 mono-stable. The other circuit I posted is non-inverting so a positive input pulse causes its output to go high, not low.
     
  20. GRNDPNDR

    Thread Starter Member

    Mar 1, 2012
    435
    7
    When it isn't connected as an inverter it clamps the voltage at 675mV, I'm not really sure why.

    I have had this working connected as an inverter without inverting the signal or at least would be led to believe so from multisim.


    I have no problem using the amp circuit you posted except that it produces the half-wave. This isn't a mtter of having the piezo wires connected backwards because for the time being I was simulating it in multisim using a plain AC source. I'm not going to go ahead and build a circuit if I don't have some idea that it's working and I don't own a real scope to check a breadboarded circuit.

    Is there a way to stop it from chopping off the negative portion of the wave?
     
Loading...