Taming piezo pickups

I got this from Sound on Sound magazine november 2004. There's a review on the fishman Aura. A digital modeling preamp for Piezo pickups that's suposed to do something about the 'quack'.

The two most popular approaches to acoustic pickups, the piezo crystal under-saddle transducer (UST) and the magnetic pickup, have one significant limitation in common in that they both almost exclusively 'hear' the string vibration alone, conveying few or none of the other elements. Piezos are also significantly nonlinear in their output ? if you double the input energy, you get about four times the output ? resulting in an overemphasised note attack and a characteristic brittle harshness often referred to as 'piezo quack'. Magnetic pickups have their own problems, with a more limited frequency and transient response and string-to-string balance issues. No 'quack', but also not quite as much of an inherently 'acoustic' flavour either.

Click to expand...

Maybe you could do something about the nonlinearity in a preamp. But it's harder to 'model' the body resonances in an analog preamp.
I guess that's why tube guitar preamps wil kinda 'cure' the problem. You'll probably need to overdrive them a bit so they will squash the attack and give you some more sustain to make up for the lack of body resonances.

As I understand it when the piezo is stimulated it's velocity/ output is not linear in level and leads to spikes of arround 6.5KHz. At first I thought of a notch filter but that does not take signal amplitude into account . But using the side chain idea you can compress the signal only at a certain centre frequency and only above a certain level . I have since read of this idea being used to tame a snare drum that sounds "peaky" (Paul White, Basic mixing).
Steve

I've found that for my guitars the best way to get rid of the quack was to install a system that had both the saddle transducer and an internal mic. For me I like the Baggs Dual Source system. Of course adding an internal mic can be problematic with feedback issues arrising but I think a person can learn to deal with feedback and get a lot of use out of these dual systems. Choosing the right saddle transducer for the guitar is also pretty important. They aren't all created equal. For archtop guitars blending a saddle transducer with a magnetic neck pickup is seriously great sounding. I also like the Raven Labs stuff to use as a blender for the mic and transducer. I also agree that the Calrec EQ is great with acoustic instruments.

Steve

[quote author="microx"]I have inserted into a comp sidechain a mixer channel which has a sweepable mid eq with a "Q" control. By sweeping arround 6KHz with plenty of "Q" I got a hefty gain reduction at the "quacking" point and by playing with attack and decay I pretty well got what I was after.[/quote]

Thanx for confirming that. A simple opto based envelope-controlled tank circuit with variable release time should then do a good job too - if not a few stages cascaded together. This opens interesting possibilitites, not just with Piezo's !

Two ways to look at this, if it's in the studio and you have a comp with side chain well there is nothing really to make . For stage or portable use then a dedicated "stomp-box" could be made with a sweepable eq in the feedback drive to the LED or FET, nice little project.
Steve

A mic source to mix with a piezo is definitely a huge improvement. Even a mag pickup with a mic - Sunrise and a couple other companies make nice little systems.

This topic comes up fairly often over at the Acoustic Guitars forum at Harmony Central. One trick (which I haven't tried) is some kind of clay shim in the slot with the piezo - it apparently damps things down a bit, and probably mutes the vibrations to the piezo. I'd kind of wonder about designing a pickup that can't be played hard - as somebody heavily influenced by Pete Townshend, I sure don't want to have to play softly to get a decent amplified sound!

On my ToDo list is to buy some of those raw piezo buzzer elements from Digikey and get some practical data. One thing I've already noticed is that one of the specs is a resonant frequency, and the smallest one on their catalog is 6.5kHz. Hmmm.... spikes at 6.5kHz?

Well, back to the drawing board I guess ... :sam:

I was thinking of making the single FET buffering circuit mentioned earlier, for piezo pickup applications, then I thought about adding a transformer to make it a DI, then I ran across this Piezo DI circuit by Jenson:

http://www.jensentransformers.com/as/as098.pdf

I was wondering if anyone had tried this, or had comments on the design...

tx
Peter