How to give my fast and dirty Piezo pick up more b@lls and bass?

[Skruffyhound]

Thanks, that's great, I can work with that.

[Skruffyhound]

I've tried the cap across the piezo now and straight low pass filter without really significant benefits - ok the lowpass does help a bit, but by far the most important factor is mechanical attachment.
I discovered by accident (lots of removing and re-attaching - long story) that placing something heavy like a 9V battery right on top of the piezo (perhaps just some tape in between) then taping the whole lot down improves the low end massively. Looks like I will epoxy one tomorrow with a piece of steel from my workshop and make a harness. I may go for the side of the guitar facing up next to the sound hole as tests suggest this may be a good spot for a bit more bass.

I think I'm going to have a buffer on the perf board followed by a lowpass filter and some pot value probably a 100K to attenuate some signal so I can dial back the fuzz/clipping.
This is not going quite as planned, I was really hopeful for the Tillman/Bandaxall combo, but they really weren't enough, although I'm not exhaustively testing and others may achieve different results.

Any suggestions for the type of lowpass filter or the order of circuit sections gratefully received.

[DrAlx]

Gluing things makes a world of difference.  If this is going in an acoustic guitar then I'd recommend...

1) Putting the piezo right under the bridge next to the low E string.  Most bridges are reinforced (there will be a sort of triangular area of wood between the braces).
    Take all the strings off and put a small mirror into the guitar body then get a flashlight to help you look inside and see what the target area looks like.
2) Take out all the bridge pins apart from the one for the low E.  You'll be doing everything by feel, so try dummy runs with no glue to get a feel for the correct location.
You can have the piezo butt up right against the bridge pin.
3) Use cyanoacrylate rather than epoxy.  It will be easier to break the piezo off with a sharp blade if anything goes wrong with it in future.
4) The wires that go to the piezo WILL break off unless you take precautions, so the piezo needs to be toughened up before you put it in.
Encase the side of the piezo with the wires in a few millimetres of epoxy so that the wires cannot pull away.   (I then put copper foil tape on top of the epoxy and solder it at the edges to the bottom brass plate of the piezo.  That way everything can be totally screened back to the jack plug).
5) Use screened cable, and make sure all wires are taped firmly to the guitar body so no wires are swinging around.

[Digital Larry]

Predicting "what will be a good EQ" for this is pretty tough from the sidelines.  I'd almost recommend hooking it up straight, recording it, then using some digital EQ to see where the payoff lies.  The simpler you can make the digital EQ, the simpler your analog solution will be.  

It would be interesting to investigate how the attachment affects the sound.  I'd expect something rigid like epoxy to accentuate the highs and something like silicone adhesive, more rubbery, to absorb some of the high end, possibly giving you some mechanical filtering to reduce the need to do it electronically.  As with most things I say on the forum, that's a total wild a%^ guess, based, as Leonard Nimoy (RIP) used to say on "In Search Of...", "partially on theory and conjecture".

[ashcat_lt]

Predicting "what will be a good EQ" for this is pretty tough from the sidelines.  I'd almost recommend hooking it up straight, recording it, then using some digital EQ to see where the payoff lies.  The simpler you can make the digital EQ, the simpler your analog solution will be.  

I like that answer better than my own.

I have seen other people talk about how it helps to put a weight on the piezo.  The voltage is generated from the compression of the piezo element.  If this light little disc just rides around on top of the guitar body, you won't get much of anything.  Give it some significant inertia of it's own, though, and it'll push back against the guitar more, and get squished more as a result.  It makes sense that more mass means more bass for the same reason as guitar strings, or maybe more like how a capacitor works.

[DrAlx]

To be precise the signal is produced by changes in the material stress and strain.
There's more to it than simply squashing the piezo.  For example, bending the piezo also has an effect, and you don't need to squash the piezo in order for that to happen.

Here's an interesting note... you wouldn't think the shape of the piezo matters but these guys claim that it does, or are at least trying to make money on that claim.

http://www.soundonsound.com/sos/apr06/articles/guitartech_0406.htm
http://www.google.co.uk/patents/US8088988

I haven't checked the details but I know that a piezo disc will have a set of modes of vibration (eigenmodes is the technical term).
You can think of these as "building blocks" that describe the piezo's vibration, much like the way that a guitar string vibration can be broken down into a fundamental and its harmonics.  So one way of looking at the problem is to consider how energy is coupled into the eigenmodes of the piezo disc.
Changing the piezo shape changes the eigenmodes, but I'm thinking that the scale of a typical piezo sensor is so small compared to an acoustic wavelength that I'd be surprised if it makes much difference.  But who knows.

I'd be inclined to not worry about the frequency response.  Just get make sure the piezo disc is firmly fixed to the guitar body across its whole area, and then EQ till it sounds nice.  Use double sided tape before gluing anything in permanently.

[Skruffyhound]

Really good suggestion Larry, wish I'd seen that earlier, oh well, I can still use that idea.
Now I've made a bit of a last ditch attempt to get it covered by making a state variable filter. I checked out the power consumption on a TL074 and it wasn't that bad so....
I also encased a piezo in epoxy (with the contact side out obviously) and glued a 10mm thick steel disc to the back of it. We'll have to wait until tomorrow to hear the results.

I'm getting desperately short of time though ha ha.

DrAix, I'm filing away your info and ideas on attaching piezo's, but on this guitar (which isn't mine) and which has a tiny body and tiny sound hole I have to go with the strapped on version.

Thanks for the support and ideas ashcat_lt.

Could someone give me a second opinion, I've used this layout, although I had to trawl the German forums to scrape up a schem to verify it.



Everything on the board checks out, but to my mind he's got his pots wrong. We are looking at them from the bottom right? So I get : Gain at Res. Freq. - from pin 5 on the chip should go to pins 1&2 on the pot and the orange wire from R7 should be on pin 3 on the pot. For the Res. Freq. part of the dual pot (bottom left B50K) brown wire from R8 should go to pin 3 and the blue from R5 should go to 1&2 , same with the other part R9 should go to pin 3 and R7 should go to pins 1 and 2. The dual pot sort of makes sense, just like he's looked at it the wrong way round and missed off the pin 1 to wiper connection, but Gain at Res. Freq. pot is just wrong.

Am I tired and talking out of my bottom or what 

[Skruffyhound]

Tomorrow may be my last shot at this. As of now it goes :

Epoxy encased weighted piezo ⇒ 10M input impedance Fet buffer ⇒ State variable filter ⇒ Mindbender Fuzz ⇒ Ruby

I can switch the fuzz out. I'd really hoped to get a chorus in there and with more time I've got a Bad Stone and a Snow White AutoWah sitting on my bench but it is what it is for now.

[PRR]

The low-end (what usually sucks in piezos) for guitar can be 82Hz. Aim for 40Hz to be sure, if you can. A 150Hz cutoff is not wrong because guitar can be tubby and in band-situations you have a bassist to fill that end.

Do you know your piezo's capacitance? If 1,000pFd, then a 10MEG load gives 17Hz, ample for any musical purpose. TL072 can easily work with 10Meg bias resistor.

Shunt capacitance reduces overall output, but extends low-end, so may be useful if too-hot yet gutless.

Parallel piezos won't add unless they get the exact same force. Distributed piezos may richen-up the sound compared to one-point. You might get either increased output or increased capacitance by stacking two disks between bridge and body. They get essentially the same force, so outputs tend to be the same, and you can series for more voltage or parallel for more capacitance. The lower stiffness may (or may not) be a better mechanical match.

[Skruffyhound]

Thanks Paul. Wiring it up now, we travel at 6am ha ha

[Skruffyhound]

Comments on the pot wiring still relevant.

The Piezo really worked with the big disc. by the way

[bluebunny]

we travel at 6am

Enjoy your hols!