Piezo's in Parallel?

Uncle Krunkus · Last edited by Uncle Krunkus on Mon Aug 18, 2008 6:32 pm; edited 1 time in total

Hey everybody!
There's some very interesting, and yet painfully slow R&D going on in the Uncle Krunkus lab!
Ideas which have been going around in my head for quite a while are finally being made real, and are mostly being returned to the drawing board! Laughing

Still,...... Up from the ashes!

Anyway, I'm onto a new way of doing an old thing, kinda hybridising a whacky idea with a tried and true one, and I was wondering if you used two piezo's as a common pickup, would it be better to wire them in parallel, or series?
I think the answer is in parallel, as that would halve the impedance, which tends to be too high.

Any thoughts?

yusynth · Joined: Nov 24, 2005 Posts: 1314 Location: France

I guess that in series you would add the amplitude (voltage) of picked signal but with a doubled impedance. In parallel, the amplitude would be the same and the impedance divided by two...

frijitz · Posted: Sat Aug 16, 2008 9:35 am Post subject: Re: Piezo's in Parallel?

Uncle Krunkus · Posted: Sat Aug 16, 2008 3:52 pm Post subject:

Thanks Ian,
I'm not 100% sure I understand the ramifications of them being "capacitive devices", but I'm sure I'll find out soon if they suit my needs. Very Happy

With two in parallel, would I still put a loading resistor/s across them before going to a TL07X buffer like in the SoundLab drum trigger? (link added for reference)
http://www.musicfromouterspace.com/analogsynth/SOUNDLABMINISYNTH/drum_trigger.html
Unlike the drum trigger, I do need to keep hum and noise to a minimum.

frijitz · Posted: Sat Aug 16, 2008 8:53 pm Post subject:

Your project doesn't seem too hard. Here's a guy that makes seismometers from piezo discs. Shocked

http://jclahr.com/science/psn/chapman/piezo/index.html

If electrical pickup is an issue, you will need a high-impedance buffer right next to the disc to use it in voltage mode. You could use a single FET mounted close by, mayber even a cute little SMT circuit.

But if it was me, I'd see if I could go current mode (piezo as a current source driving an I-V converter). That way the disc is at zero potential so a shielded lead would guard the signal without any loading capacitance.

But I've never seen anyone do this. scratch

Ian

Uncle Krunkus · Posted: Sun Aug 17, 2008 4:55 am Post subject:

frijitz · Posted: Sun Aug 17, 2008 8:48 am Post subject:

Uncle Krunkus · Posted: Sun Aug 17, 2008 3:42 pm Post subject:

Thanks heaps Ian,
I settled on a particular mechanical setup last night, so the next thing is to build a couple of these right next to my two piezo's. Which I've decided to run antiphase, so I could take the outputs of these I->V converters straight to either side of a third op-amp yeah?
I think I may have damaged my last LM386 as well! Laughing

I didn't have a bypass cap (10uF) across the 12V for the LM386, and it started squealing! Then it got quite hot! So I shut down, put the cap back, powered up, and, no squealing. But also not as much signal to the transformer either! Confused

I think I need a few new ones. Smile

frijitz · Posted: Sun Aug 17, 2008 9:27 pm Post subject:

Uncle Krunkus · Posted: Sun Aug 17, 2008 10:07 pm Post subject:

I tried reversing the leads for an experiment, but for the main design idea I've got little hooks soldered to the brass base, they are not insulated, (you'll see why when I post another pic, soon) so reversing the leads would short one piezo out.

While we're at it, do you know what the efficiency of a piezo is? Effective both ways? For example, if you vibrate one piezo which is electrically connected to another, how much of the original vibration is reproduced on the other side? What kind of "booster" would work between the two in order to keep it at unity?

These might seem like silly questions, but I think it's good to get to know all about a certain "medium" while I'm in the thick of the brainstorming stage.

As a completely separate, and yet related question.
You know the standard pair of clipping diodes around the feedback loop of a classic fuzz circuit?
No,... on second thoughts, that is a silly question, or one I should just verify through my own experiments. I'll try to expand on that thought later. Wink

Uncle Krunkus · Posted: Mon Aug 18, 2008 8:28 am Post subject:

Uncle Krunkus · Posted: Mon Aug 18, 2008 8:45 am Post subject:

No, with the two of them into a third op-amp, I get this high pitched squeal which sounds a bit like heterodyning (or ring mod?). It stops when I touch the output with my finger, but comes back if I just clap my hands.

Uncle Krunkus · Posted: Mon Aug 18, 2008 7:00 pm Post subject:

Now I've got something good happening which sounds stable.
I took the two outputs to the other two op-amps in a TL074. One of them just as an inverter, the other as a summing inverter. I used 10Ks all round except for the feedback on the summer which I set at 51K for a bit more gain. BTW Ian, is there a problem with reducing the 220Ks? As I'd like just a little more gain off the bat, or will that increase noise etc?

Uncle Krunkus · Posted: Mon Aug 18, 2008 7:02 pm Post subject:

Hmmm, actually, to increase the gain I'd increase the 220Ks wouldn't I?

frijitz · Posted: Tue Aug 19, 2008 6:54 pm Post subject:

Uncle Krunkus · Posted: Tue Aug 19, 2008 7:25 pm Post subject:

frijitz · Posted: Wed Aug 20, 2008 2:00 am Post subject:

Uncle Krunkus · Posted: Wed Aug 20, 2008 5:56 am Post subject:

Okay,
I'll be breadboarding this up tonight, and I should have my new LM386s by Friday, so I'll find out if the one I'm using is okay or not I guess.
In the meantime I'm going to put together the mechanics of the second channel, and see if I can break up these lower order harmonics. I'm starting to think I need to re-think the whole mechanical side of the experiment. That's the good thing with talking to you about the electronics involved, it's given me time to really see what was going on with the mechanics of it, and why that was holding back the results.
Thanks again for the brainstorming of my whacky ideas!! Laughing