Amplification Circuit Design Issue

[BARD]

Hey all,

I am designing a drum trigger circuit to clock sequencers.

I started by designing a dbounce circuit using the 7414 and a momentary push button. This part of the circuit is represented in the attached schematic on the right side. The circuit worked well with the momentary switch and so I started to design an auxiliary input to allow piezos/mics to trigger the 7414 as well.

I introduced a 3904 transistor and this worked well. I could tap the piezo and it would generate a 5v pulse out of the 7414. I went to go use this with a bass drum and found that it would only trigger off the more intense/direct vibrations from the drum. I wanted both the rim and drum head to trigger the 7414 so I decided to design an amplification circuit to increase the gain coming from the piezo.

I designed an amplification circuit with an LM324, represented on the left side of the schematic. Because I had already built/soldered a prototype of the 7414 circuit, I made a prototype of the amplification circuit separately. They each had their own power supply headers coming off the 5v rail of a distribution board (I am designing it for eurorack). I put a jack at the output of the LM324 and patched it to the input of the 7414 circuit with a patch cable. This worked very well. I was able to use a potentiometer to increase/decrease the gain of the piezo and was able to tune in the trigger amount to where I wanted it.

At this point I decided the circuit was complete. I designed the schematic attached to this post and sent it off to get made into a PCB. I got it back in the mail and build it up. It appeared to work well. The momentary switch worked perfectly and I was able to generate triggers with a piezo going into the input.

I went to go test the board with a bass drum as I had with the protoype. But I found that it was still only triggering from direct vibrations. It was acting as if the amplification circuit wasn't doing anything.

I checked the board against the schematic and it was spot on. So I decided to breadboard the circuit off the schematic from eagle. I was still having the same issue after breadboarding. I decided to check the output from the LM324 on a scope and I noticed it was only outputting 1V with each peizo hit. But when I disconnect the wire from the LM324 to the trigger circuit, the output shoots up to 5v with each hit as it was designed to do.

I need the output of the LM324 to output 5v with each hit in order for it to properly trigger the 7414 but when I connect the two circuits it cuts the LM324 output down to 1v. I'm not sure why this worked when I had two separate prototype circuits (amplification and trigger) and patched them together with a patch cable.

Is it a buffering issue? Any ideas would be helpful.

Thanks so much. This is my first post on here 
-Zach

[PRR]

Welcome!

> output from the LM324 ... only outputting 1V with each peizo hit.

You have it driving a naked transistor B-E junction. It would take obscene current to pull Base over 0.6V.

However the 0.6V slam at Q3 Base *should* fling Q3 Collector a full 5V.

The whole plan seems odd to my eye. You have a 5V rail. Audio (even drum-slam) swings both ways. We fake this by biasing things to half-supply, here 2.5V. The '324 "will" work with both inputs and output near zero. But the audio is horribly mangled. Which we may not mind for a Trigger, but it may be mangled too much to work as expected.

No part-values also makes it hard to guess what's going on.

I have no idea why it worked in proto-patch. You might re-visit that and see if any detail got lost in translation.

> made into a PCB

Sorry to hear that. Makes changes difficult.

My first suggestion is--

http://oi66.tinypic.com/9k5o5k.jpg

Lift the "Gnd" end of R4 out and air-wire a 2.5V supply for it.

Lift the "Gnd" end of R9 and air-wire a cap in there.

Lift the transistor Base leg out of the PCB. Air-wire a cap to '324 and diode to ground.

This makes the '324 a linear audio amp. With gain of 100 (you don't say) and a high input impedance (you don't say), this should be idling at 2.5V and slamming 2V jolts up and down. The 2V jolts will really be limited to 0.6V jolt each way by transistor/diode clamping. The transistor will idle "off" and be forced "ON" once each cycle. That gives a 5V swing into the 7414 which is more than enough.

I do not understand why you are feeding +5V to the input. Most piezos won't mind, though they may be stressed which may affect sensitivity. Or by "piezo" do you really mean electret capsule, which does need power, but is a very different thing?

[BARD]

You have it driving a naked transistor B-E junction. It would take obscene current to pull Base over 0.6V.

Would you suggest a resistor in there?  That junction is where I originally patched in the two circuits together.  Out of LM324 into 3904.

The '324 "will" work with both inputs and output near zero. But the audio is horribly mangled. Which we may not mind for a Trigger, but it may be mangled too much to work as expected.

There is no need for clean audio.  The rapid transients of a drum are already insane and will be smoothed out with the dbounce circuit. 

I do not understand why you are feeding +5V to the input. Most piezos won't mind, though they may be stressed which may affect sensitivity. Or by "piezo" do you really mean electret capsule, which does need power, but is a very different thing?

GOOD POINT!  I will take that out.  Thank you haha



I might try out your suggestion with the amplifier.  Still curious as to how my proto-type works so well.  I'll take some more measurements when I get into the shop again and see if there's something I'm missing.