Buffer design for matching impedance from piezo to guitar amp or piezo to mixer?

[iandy4]

Anyone ever build or know of a really good circuit for a buffer with high enough impedance for a piezo input and an output to either a guitar amp or mixer?  I realize that guitar amps and mixers have different input impedance's but I'd be happy with a schematic for either! It would be great if it had the ability to get the output gain near the signal of a normal guitar pickup.

[R.G.]

Most piezos are happy with 1M and up as a load. It's pretty easy to get that from any JFET-input opamp, including the semi-standard TL072. All you need to do to make this work is (1) use the noninverting input for the signal input, (2) keep the combination of input pulldown resistance and opamp bias resistance over 1M, or more, and (3) accept the noise that the large resistances give you.

Piezo pickups tend to give more signal out than magnetic pickups, or did back when "ceramic cartridges" were used on cheap record players. They used to put out line level directly. Depends on the exact pickup, I guess. The TLO72 opamp version can give any gain you want, including less than unity, with Rfeedback>Ri where Ri is the gain resistor on the inverting input, and then a volume control or resistive divider after the output.

Most mixers have guitar-level inputs, which tend to be about 1/10th (i.e. -20db) compared to line level of -dbm, or about 1V peak, 0.7xx volts rms.

[iandy4]

Most piezos are happy with 1M and up as a load. It's pretty easy to get that from any JFET-input opamp, including the semi-standard TL072. All you need to do to make this work is (1) use the noninverting input for the signal input, (2) keep the combination of input pulldown resistance and opamp bias resistance over 1M, or more, and (3) accept the noise that the large resistances give you.

Piezo pickups tend to give more signal out than magnetic pickups, or did back when "ceramic cartridges" were used on cheap record players. They used to put out line level directly. Depends on the exact pickup, I guess. The TLO72 opamp version can give any gain you want, including less than unity, with Rfeedback>Ri where Ri is the gain resistor on the inverting input, and then a volume control or resistive divider after the output.

Most mixers have guitar-level inputs, which tend to be about 1/10th (i.e. -20db) compared to line level of -dbm, or about 1V peak, 0.7xx volts rms.

Thanks R.G.. I wrote a very long response that got erased but basically, in my personal experience the piezo straight to amp sounded like it lacked a lot of low end and I had to crank the amp all the way up to get a small signal (aren't most amps having around 1M input impedance?).  In that case I was using a piezo in an electric bass to a bass amp.  Then I tried the piezo through a Tillman FET preamp (which I knew worked well for guitar) then to the amp. This barely improved the signal.  I know I've heard piezo's in guitars that sound just fine.  If there isn't really an impedance mismatch and a piezo produces a stronger signal than a pickup then what am I missing or not understanding that explains why the signal has to be cranked up so much more?

[R.G.]

Then I tried the piezo through a Tillman FET preamp (which I knew worked well for guitar) then to the amp. This barely improved the signal.  I know I've heard piezo's in guitars that sound just fine.  If there isn't really an impedance mismatch and a piezo produces a stronger signal than a pickup then what am I missing or not understanding that explains why the signal has to be cranked up so much more?

Piezos vary. They're effectively capacitive, with lowest source impedance at high frequencies, the opposite of magnetic pickups. So they lose bass frequencies selectively when used into too-low a load impedance, just as the inductive nature of magnetic pickups cause them to lose highs. Same effect, opposite ends of the spectrum. So it is an impedance mismatch, but it's a frequency selective mismatch, opposite to magnetic pickups.

If 1M isn't high enough for a load, go to 10M. This gets tricky because 10M or higher input impedances start getting corrupted by things like dust on the PCB and humidity.

[iandy4]

Ahhh, that makes sense.  I'll give it another shot in the coming days.  Thanks for the help R.G.!

[chi_boy]

The cigar box guitar guys frequently use piezo pickups in thier creations.  From what I read, the Tillman preamp is often used sucessfully.

[R.G.]

The cigar box guitar guys frequently use piezo pickups in thier creations.  From what I read, the Tillman preamp is often used sucessfully.

Yeah, it struck me as odd that 1M wasn't enough. That's why I guessed piezos vary.

[iandy4]

I ripped it out of a greeting card so maybe it wasn't optimized for the task at hand.  I've ordered some piezo's off of ebay described as "for cigarbox guitars" so I'll see how those work out.  The other thing was that it is installed in an electric bass and the piezo is just pinched between the bottom of the basses neck pickup and body.  I'm not sure how they are supposed to be installed to work optimally but I imagine that being pinched could also prevent the piezo from transducing the lower frequency vibrations. I'm doing research on placement now.

What I thought would be very cool is to use these piezo's as drum triggers (more like drum mics) that could be run to a mixer (or custom summing amplifier) and then routed through guitar effects which then go to the PA at a venue.  This could be a cool way to use my guitar effects on the drums.  Is that a plausible idea? or would it make more sense to get those little cheap mic capsules and try it with those instead of piezos?

Thanks for your input guys.  I've created some incredibly useful things which I'm starting to use in live settings just because I can bounce ideas off of people in the forum and get help when I'm stuck.  Non-stop soldering since I joined this forum!

[Gus]

Something I posted at another forum



You can omit or adjust C1 value(it is a capacitance voltage divider with the piezo) it is to keep the SF from clipping.  From out of C3 to R10 is a limiter and low-pass filter, this can be omitted.  I just picked 500pf for the piezo.  C1 would need to be changed if used for different piezo capacitance

I have not built it but it sims OK.

Edit  you want about 4VDC at the "top" of the 2K (2ma) SO the IDSS of the JFET needs to be higher than 2ma at 5VDC drain to source test voltage.  I would look for one around 10ma.   
OR
you could make R2 3.9K with the same 4VDC (just over 1ma less drive) and use a JFET with >1ma IDSS

[iandy4]

@Gus

Thanks! This looks pretty nasty!  I would love to build it.  I think it's a few too many variables for me tackle at my current understanding of electronics but I'm always up for a challenge.  I'll be saving this one for a time when I better understand what everything in the circuit is actually doing.  Thanks for posting it! I'll definitely check it out more in depth.

[Gus]

I put notes in the screen shot about what the parts do and what needs to be adjusted.  This is a text book follower circuit

It seems the LT spice model of the 2n3819 has an IDSS of about 11ma at 9VDC D to S(I simmed another circuit in LT spice to check)

Remove the diodes and R4

Look up measuring IDSS of a JFET

Decide what drive you need, say 1ma  Make sure the IDSS is higher than that (a 2n3819 selected for >= 5ma IDSS might be something to try)don't use a J201

Make R2 3.9K and adjust R1 value for about 4VDC at the "top" of R2

Omit C1 at first

Adjust C7, R3 to taste (RC low-pass)

Have fun

[iandy4]

Alright then, I'll be trying it sooner than later.

[Ben N]

In response to your initial post, I thought placement might be the culprit. The strength of your signal is going to be proportional to how much string vibration the piezo picks up. I know folks have experimented with different locations AFA getting different tones from the piezo, especially for mixing with magnetics, but in terms of signal strength, optimum would have to be either at the bridge or the nut. It may not be the worst, but the neck pickup slot is surely not the best place for this.
There is (or was) an instrument builder's forum out there that had some info about piezo placement and mounting.