Minimalistic piezo buffer for passive blending?

[vin97]

Hello!

I am planning on putting some piezo saddles in my fretless bass and am looking for an appropriate 9V buffer.
I want to keep the magnetic pickups fully passive, so the buffer is only needed to match the impedance/level of the piezos to the mags.
The buffer circuit should be as simple and transparent as possible.

What circuit can you recommend for this purpose or are there even readily available buffers like this on the market?

[Gus]

I did a search at this forum using "piezo buffer" and got 134 results

[vin97]

I've gone through a couple of threads but the links are either dead or talking about the much more common design of an active blend where both the magnetics and piezos are affected.

While I have the necessary soldering skills, I don't have enough knowledge to modify those circuits to fit my needs.

[idy]

Its kind of an interesting question. Most blends are two active or two passive. What you have is an impedence matching question. No?

https://www.seymourduncan.com/blog/latest-updates/active-and-passive-in-the-same-guitar-can-it-be-done
asks the question and comes up with:
buffer the passive pu
only use one at a time
two seperate volume controls/output cables/signal chains

[Rob Strand]

The buffer is the easy part.  There's quite few examples on this forum and many on the web.   Many are simple JFET amps but some are opamps.

How many pickups do you have on your bass and what are the existing controls?
What controls do you expect to see in the final setup?

The decision that needs to be made is how much impact on the passive sound are you willing to accept?

For example if your bass is has two pickups, you can switch between the piezo and either of the existing pickups then use the existing controls.   That's a relatively simple way to combine active and passive.     It's not 100% benign but since you are adding the piezo things aren't unchanged anyway.

However, if you want to add additional controls on top of what you already have that will generally cause some additional loading on the passive pickups.  In other words when you add on the new circuit it will affect the passive stuff even when the new circuit is dialed out.

If you want avoid the loading then you would need to feed the output of your passive set-up into an on-board buffer/mixer.    Doing that will keep the sound passive as possible but it's not really passive in that the passive output is buffered.  So if you plug the bass into a low-impedance  load the output of the buffer will keep the passive sound the same as a high impedance load but on a true passive set-up the low-impedance load changes the sound.  Also for the buffer to work the battery must be working.

If push comes to shove you can always add an active/passive switch.

[vin97]

I know the circuit is relatively simple to build but I am a complete novice and don't know how to customize the existing ones for my application.

Some extra loading will probably not be a problem as long as the output impedance of the buffer is roughly equal to that of the mags. The bass has enough highs. The problem is, all the buffers I found lower the impedance too much (all the way to "active levels") and therefore would affect the passive mags very negatively or drown them out completely.

It's a standard Fender Jazz bass with two single-coils (blended by a 100K M/N pot), 250K master volume and 250K master tone.
My plan for the mag/piezo-blending was to simply add another M/N pot after the mag-blend and in front of the master tone and volume.
I would mod the blend pot to "no-load" by cutting the traces at the outer ends, so that I can get the piezo part completely out of the circuit.

I definitely want to avoid making the whole bass active. Ideally, a loss of power should not influence the signal of the mags.

[Digital Larry]

I'd vote for some kind of JFET input op amp in a non inverting configuration.  Then add a series output resistor, to start with, on the order of the DCR of your pickups, probably around 10k.  There's probably already a series output resistor in combination with a DC blocking capacitor for most designs you'll find that run off a single voltage.

[vin97]

So it's just a matter of fitting the right resistor in series with the buffer out?
Wouldn't this affect the sound of the piezos negatively after a certain point?

[Rob Strand]

It's a standard Fender Jazz bass with two single-coils (blended by a 100K M/N pot), 250K master volume and 250K master tone.
My plan for the mag/piezo-blending was to simply add another M/N pot after the mag-blend and in front of the master tone and volume.
I would mod the blend pot to "no-load" by cutting the traces at the outer ends, so that I can get the piezo part completely out of the circuit.

In that scheme, what Larry said should do the job.

So it's just a matter of fitting the right resistor in series with the buffer out?
Wouldn't this affect the sound of the piezos negatively after a certain point?

It's more likely to affect the passive pickups when the blend is towards the piezo side.   It's a two-way street in that the piezo is affected by the passives when the blend is towards to passive side.   That's life with these blend controls.    Even two passives interfere with each other in a blend configuration compared to being wire singly.

Yes, in fact you could play with the value 10k to 33k to tweak the sound.   Higher values will attenuate the piezo a bit less and the affect the passives less.   

The only down side of the buffer is it doesn't give you any opportunity to boost the signal.   You need to make sure the piezo is wired in-phase with passive.

With the your constraints there's only a few options.  The buffer is the most straight forward.

[vin97]

What do you think about the Tillman buffer?
It's the simplest circuit I could find and the output impedance is 6K. That would already be pretty close to the mags, wouldn't it?

[sergiomr706]

Which piezo model are you planning to use?

[vin97]

GraphTech Ghost saddles with individual trim pots for equal string volume.

[samhay]

Where do the trim pots sit? Are they inline with the piezo?

My solution to this situation is a little different. I don't like batteries in guitars, so I run the piezo to the ring of a stereo output jack. The mag pickup (p bass, so only 1) is passive and goes to the ring like normal. If you plug in a typical mono cable, the piezo is shorted to ground and the mag pickup does its thing. If you plug in a stereo cable and take this to a preamp with a stereo input, then you can separate the two pickups and do with them what you will.
I expected there to be more crosstalk between the pickups, but this is negligible and may not be a problem if you blend them anyway.
I'm not sure how this will work with your piezo pickup as it has 6 outputs right? You should be able to connect them in parallel, but that's where my question about the trim pots comes in.

[vin97]

Yes, I'd put the trim pots in series with the individual transducers and connect them in parallel to the buffer which is then mixed passively with the mags.

I don't like batteries either. I was planning on supplying the necessary power through a stereo cable. I also don't want to rely on external mixers or preamps, that's why I would like to do all the blending onboard.

[amptramp]

What do you think about the Tillman buffer?
It's the simplest circuit I could find and the output impedance is 6K. That would already be pretty close to the mags, wouldn't it?

One of the things you can do with a Tillman buffer is mount R1, R2 and Q1 in the guitar and the rest of the circuitry in a separate box so the guitar cable sees the impedance of the drain resistor R3 in parallel with R4 so when Keith finally knocks off and the Stones tap you to play guitar, you can use stadium-size 30 metre cables and not worry about excess capacitance.  No battery is required in the guitar but the box with R3, R4, C1, C2 and the power supply or battery must be used.

One enterprising company has come up with a Tillman cable with R1, R2 and Q1 on one end and the rest of the preamp components on the other end.  This means you do not need to modify the guitar beyond mounting the piezo pickups but you have a 50% chance of plugging in the cable the wrong way.

[Rob Strand]

What do you think about the Tillman buffer?
It's the simplest circuit I could find and the output impedance is 6K. That would already be pretty close to the mags, wouldn't it?

The Tillman might work.   You would need to make sure the Tillman doesn't clip.   However, from the later posts I think you need to revisit the whole structure of the preamp as a result of the individual piezos.

As far as the output impedance is concern.   You don't need to *find* a preamp with a specific output impedance.   You can use just about any preamp then add a 10k to 33k (or whatever) in series with the output.

FWIW,  the Richter BassXX  piezo preamp seems to do what you want.     That preamp has 100k output impedance  (another page showed 47k!)   I saw one claim you can blend it with Passive pickups but they don't give a wiring diagram.

Yes, I'd put the trim pots in series with the individual transducers and connect them in parallel to the buffer which is then mixed passively with the mags.

I've got serious doubts that will work.    Even if it appeared to work I don't think it's going to work well.  It would mess with the frequency response of the pickups or also add noise (due to the resistors).

The configuration I see is *each* pickup going to a JFET buffer (or maybe a Tillman preamp).   Each output has its own volume trim, then those outputs are mixed together.     If you mix 4 inputs passively the output will drop by a factor of 1/4.   So it is highly likely you will need to use an active mixer, or a passive mixer followed by another gain stage to recover from the mixing signal loss.

[Rob Strand]

Check out this pic of a Fishman piezo bass preamp,

https://reverb.com/item/47536-new-fishman-piezo-4-string-bass-guitar-preamp-tone-volume-controls-and-more

At the top right you can see the Piezo board.   Each channel has it's own JFET (don't know if it's a buffer or a gian stage).   It's high likely the IC is being used as an active mixer.    The board has an extra trimpot for overall gain.

[vin97]

That Richter Preamp looks perfect for my purpose. I just contacted Andreas to get some more information on it.

Meanwhile, in case I do have to build my own, would it really be necessary to have individual buffers for each piezo or could you simply design it for more gain and connect the piezos in parallel to the input? The trim pots would be relatively low resistance, with the quietest string having no extra resistance on the piezo whatsoever.
The problem is mostly space here, as I have to fit everything into the small Fender electronics cavity and I cannot etch PCBs.

[Rob Strand]

Meanwhile, in case I do have to build my own, would it really be necessary to have individual buffers for each piezo or could you simply design it for more gain and connect the piezos in parallel to the input? The trim pots would be relatively low resistance, with the quietest string having no extra resistance on the piezo whatsoever.

The main thing individual buffer lets you do is adjust the individual volumes.

The simplest scenario is to connect all the pickups in parallel (as samhay mentioned earlier) and feed the output into something like the Tilman preamp.   The GraphTech preamp looks very much this type of thing.   So what you are giving up here is the ability to adjust the volume of each string.

I have my doubts adding small series resistors is going the adjust the volume.   Also adding resistors messes with the frequency response and could add noise if you need to add large series resistors.   Piezo pickups look like capacitors they don't work like normal pickups.

One way to *reduce* the level of an individual pickup in a parallel wired pickup is to add a series capacitance to one element.   You can do an experiment to see just how much adjustment you need maybe you only need to tweak a couple of strings.   So that means just adding a couple of capacitors in series with those pickups.
[Boosting a single string would need a set-up with two Tilman preamps.]

This configuration is vastly simpler than a multi-channel buffer set-up with individual pickup levels. 


FYI, some background on piezos:

Piezo pickups look like capacitors.   When you put them in parallel the frequency response doesn't change but the output drops by a factor  of  1/n; so two parallel pickups = half signal, 4 parallel pickup = one quarter of the signal.   You don't get this signal drop with individual buffers.   The signal drop can be managed. 

The resistance load presented to the pickups determines the bass cut-off.  If you load the pickup with a low resistance  you get bass cut-off.  Note however paralleling pickups doesn't change the bass response because they are all capacitors.  What happens though is putting all the pickups in parallel makes the overall capacitance of the combined pickup higher.  That lets you use a lower resistance load for the combined pickup than a single pickup in order to get the same bass cut-off.

Pickup specs rarely give the capacitance or recommended loads.   The GraphTech documents for the guitar pickups show 6 pickups in parallel driving 5Meg ohm load.   The preamp would be 10M input impedance, maybe lower,  so the load for 6 parallel pickups ends up being 3M ohms.   Which is the same as 6x3M ohm = 18M ohm per pickup.   The Richter preamp presents 2.2M ohm.    So you can see there's a big uncertainty in what the pickup needs for a load!.

It's possible to measure the pickup capacitance with a capacitance meter but you probably want one that can measure 100pF *resolution* (not range).  Many DMM's only do 1nF resolution which isn't enough.
----------------------
Here's some more info,

https://www.richardmudhar.com/blog/using-piezo-contact-mics-right/
https://sound-au.com/project202.htm

Reference 4 on the last link might also be helpful.

[vin97]

Ok thanks for all that info!


What kind of impedance do regular single-coils have anyway? I've seen vastly different values quoted.