Multi-sensor guitar piezo installation - looking for advice

I'm looking to rig up a Johnson Trailblazer with a 'permanent' internal piezo pickup installation, and would like some advice...

This is the guitar (sounds very sweet, if a little quiet...) www.jklutherie.com/johnson-travelguitarwgigbaginstruments.aspx
I have this kind of sensor in 35mm, 27mm and 20mm:
http://cgi.ebay.com.au/5-35mm-Piezo-Elements-Sounder-Sensor-Trigger-Drum-Disc_W0QQitemZ120526869145QQcmdZViewItemQQptZUK_BOI_Electrical_Components_Supplies_ET?hash=item1c0ff61299

The plan is to use a 35mm about an inch from the bridge on the bass side (found what seems to be a sweet spot by 'listening' with fingertips P ), a 27mm right up against the bridge in the middle, and a 20mm on the underside of the bridge itself on the treble side. The idea being to pick up as much tone as possible by not just having a single sensor which could miss a vibrational mode of the guitar top by being placed on a node...

Questions:
1) Should the sensors be glued with the ceramic side in contact with the wood?
2) Can the three sensors just be hooked in parallel straight into a single buffer, or should they be buffered first, then mixed?

Any other comments or suggestions would be welcome

Answers to your questions -
1. Yes, I think so.
2. No. Piezo sensors are pretty high impedance devices. I think they're at least 100 megohm if not higher. I don't think hooking them in parallel will work the way you want. Buffer each one separately first and then mix the outputs of those buffers.
DW.

You may want to search for info on the K&K line of transducers. The better models use 3 elements attached under the bridgeplate. I believe they are wired in parallel.

It is perfectly feasible to connect piezo pickups in parallels, but the respective balance may not be satisfactory (the larger element, having the highest capacitance will dominate) and adjusting it is very difficult.

Thanks fuzzhiemer - the K&K transducers seem to be glued brass-to-wood rather than ceramic-to-wood, which I found interesting. They have three pickups the same size, which would avoid the balance issues 'abbey road' mentioned.

This guy seems to be going for the same effect, wired directly in parallel, and doesn't seem to be using buffering (which probably explains why he's using foam, to make the pickup sound less harsh) http://www.nuklhedcentral.com/html/piezo_pickup.html

Found some other good links here
http://www.godeater.com/misc/contact-microphones/home?do=backlink

So, the plan is: straight buffer for each piezo, with a mixer stage, all using a TL074 op-amp, mounted inside the guitar - external power box with a pair of 9V batteries in an external altoids tin... might try bridge pickups in addition to soundboard pickups, and put a mixer in the tin... maybe jam a PT2399 reverb in there if there's room... (I'm a SW engineer by day, does it show much?)

abbey road d enfer said:

It is perfectly feasible to connect piezo pickups in parallels, but the respective balance may not be satisfactory (the larger element, having the highest capacitance will dominate) and adjusting it is very difficult.

Click to expand...

I'm not looking to start an argument here. But even if it can be done and will work, is it sensible to do so? I say no. I bet there will be significant phase differences and cancellation when two or more elements are connected in parallel. DW.

I would definitely buffer each one and then mix them. But I would also seriously consider cutting down complexity, some of the guys who have put a lot of effort into piezo's (Rick Turner to name one) say that being farther from the bridgesaddles doesn't necessarily equate into a more acoustic sound. the piezo's in my Electroline bass are in the saddles and it is very acoustic sounding despite being a solid body.

Tubemooley said:

abbey road d enfer said:

It is perfectly feasible to connect piezo pickups in parallels, but the respective balance may not be satisfactory (the larger element, having the highest capacitance will dominate) and adjusting it is very difficult.

Click to expand...

I'm not looking to start an argument here. But even if it can be done and will work, is it sensible to do so? I say no. I bet there will be significant phase differences and cancellation when two or more elements are connected in parallel. DW.

Click to expand...

There is no difference in terms of phasing issues between passive mixing (parallels connection) and active mixing.

Would phasing issues really be a problem? I was under the impression that guitar tops (and stringed instrument soundboards in general) resonate in various standing wave modes depending on the pitch - with the longitudinal sound velocity in wood being upward of 5000 m/s, and the pickup 5cm from the bridge, the propogation time would only be 0.01 ms, so for a 40kHz sound wave, the phase difference would be fairly significant (80 degrees or so - carnt be arsked working it out just now), but only for the < 0.05ms or so that it takes for the wave to reflect off the edge of the soundboard...

Thanks for the tip Jarno - have seen similar things on user reviews of the K&K - I might even try under-saddle first, see how I like it... that would leave the other channel free for an electret or a magnetic neck pickup...

madact said:

Would phasing issues really be a problem?

Click to expand...

Not really, but still, you need to take care about it. Just the same as an electric guitar with 2 pick-ups on, there are are cancellations and enhances, that may not be nice (some guitars have areally honking sound with the switch in mid-position).

I was under the impression that guitar tops (and stringed instrument soundboards in general) resonate in various standing wave modes depending on the pitch - with the longitudinal sound velocity in wood being upward of 5000 m/s, and the pickup 5cm from the bridge, the propogation time would only be 0.01 ms, so for a 40kHz sound wave, the phase difference would be fairly significant (80 degrees or so - carnt be arsked working it out just now), but only for the < 0.05ms or so that it takes for the wave to reflect off the edge of the soundboard...

Click to expand...

Wood is an anisotropic material, sound travels at different speeds depending on direction, spruce, in particular is 5 times slower across than lengthwise. Anyway, the concern is not pure solidian propagation speed; combination of mass and elasticity is similar to LC, so the frequency response is twisted and group delay becomes important. Just like in a transmission line, propagation speed is much slower than c.