Looking for DIY onboard piezo bass preamp

[theehman]

I've got a Zeta Prism bass project that's complete except for the preamp.  A few years ago I had several PCBs removed from Washburn basses that would've probably worked, but I haven't seen them since moving.
I'm looking for a bass preamp circuit that has a mixer for 4 piezo inputs (one for each string), Volume, Bass, Mid, and Treble controls.  Any suggestions?

[SolderBoy]

I've never seen that exactly, but I've designed a few custom circuits as you describe over the years.

Here's a simple dual 3-band opamp piezo pre that will work.




As for mixing, well check out Rod Elliot's page on active mixing.

https://sound-au.com/articles/audio-mixing.htm

Check out fig. 4

If you make a buffer for each piezo, like the first stage of the pre I posted, then use trimpots, resistors, and a recovery amp as in Fig 4 on the ESP page, you'll be off...

[bluebunny]

The late Albert Kreuzer had an onboard buffer circuit on his website, IIRC.  Sadly, the website is now defunct, but you should be able to find it via the Wayback Machine.

. . .

OK, I did the search for you:   



Read the notes here about the changes he made for piezo inputs.

[theehman]

So, I took SolderBoy's advice and combined his design with the active mixer circuit he linked to and came up with this.  I've got it using a TL074 and a TL072.  Can anybody see any potential issues with this circuit?

[iainpunk]

The late Albert Kreuzer had an onboard buffer circuit on his website, IIRC.  Sadly, the website is now defunct, but you should be able to find it via the Wayback Machine.

. . .

OK, I did the search for you:   



Read the notes here about the changes he made for piezo inputs.

that 250k input Z is way to low for piezo pickups

check out tihis page, i have had great experience with the charge amplefier in the past. when doing active speaker feedback for a school project, we used that exact topology with great success
https://sound-au.com/project202.htm

cheers, Iain

[bluebunny]

that 250k input Z is way to low for piezo pickups

The "notes" link I posted acknowledges this and Albert made suggestions.

[iainpunk]

that 250k input Z is way to low for piezo pickups

The "notes" link I posted acknowledges this and Albert made suggestions.

i didn't see that, sorry
also, i guess that people who are to lazy to click through might benefit from what i said.

cheers, Iain

[bluebunny]

also, i guess that people who are to lazy to click through might benefit from what i said.

Yep, can't go wrong with Rod Elliott! 

[anotherjim]

So, I took SolderBoy's advice and combined his design with the active mixer circuit he linked to and came up with this.  I've got it using a TL074 and a TL072.  Can anybody see any potential issues with this circuit?

That appears ok to me.

[iainpunk]

So, I took SolderBoy's advice and combined his design with the active mixer circuit he linked to and came up with this.  I've got it using a TL074 and a TL072.  Can anybody see any potential issues with this circuit?

i didn't notice the first time looking over the schematic, but i saw IC2-A, the mixer opamp, should have the non-inv input (+) connected to Vbias, not ground, and the trim pots to ground should have a HUGE value cap to ground to not let DC through but shunt AC (signal) to ground, so that the bias stays in tact

cheers, Iain

[antonis]

+1 for IC2-A Vbias but -1 for decoupling cap(s)..
(MAX DC gain is unity for each of mixing signals so there is no need for DC unity roll-off due to cap on gain branch..)

[iainpunk]

+1 for IC2-A Vbias but -1 for decoupling cap(s)..
(MAX DC gain is unity for each of mixing signals so there is no need for DC unity roll-off due to cap on gain branch..)

but you have 4 times DC, not just once.
10k resistor and 10k trimmer in series is 20k, having that parallel 4x is only 5k.
look at what the opamp does with the Vb as s non inverting gainstage. having 5k and 10k in there would give a gain of 3, thus putting out 3*Vb, which is well above the output threshold.

a solution is lowering the voltage on that particular input to Vb/3.
i propose 15k and 33k to replace the bottom 47k resistor of the Vbias voltage divider, and putting the node between the two resistors to the (+) input of the opamp

cheers, Iain

[antonis]

Sorry Iain but mixers don't work like this way..

Each individual IC1 output is "isolated" from the other GNDs & outputs via respective 10k resistor so both DC & AC gain is just 10k/10k..
(after all, each individual stage "sees" 10k input impedance of IC2-A..)

[iainpunk]

Sorry Iain but mixers don't work like this way..

Each individual IC1 output is "isolated" from the other GNDs & outputs via respective 10k resistor so both DC & AC gain is just 10k/10k..
(after all, each individual stage "sees" 10k input impedance of IC2-A..)

but you amplify the bias if there is no DC decoupling in the mixers inputs. adding bias makes it necessary to decouple the inputs, DC wise.
remember that you have 4 inputs that have -4.5v on them, but with higher impedances. you add that difference 4 times, since its V1+V2+V3+V4.

look at it like this:


cheers, Iain

[Rob Strand]

If you want to keep the volume pots grounded, the economic way to keep the mixer correctly bias at 4.5v is to put a single cap between the mixer opamp -input and the junction of all the 10k input resistors.
Look at the Boss OC-2 for the non-grounded pot case.

[anotherjim]

Yeh, I missed the DC error. Yes, you can use a single common cap at the -input. Always looks wrong though!

[duck_arse]

are those piezo pickups happy having 4V5 across them, or should they be DC blocked?

[iainpunk]

are those piezo pickups happy having 4V5 across them, or should they be DC blocked?

they don't mind DC across them in general.
they are a modeled like a voltage source and a capacitor in series, and a 10Gohm resistor parallel to that, having one side grounded and the other side biased is no problem in most applications where the bias is under 5v (rule of thumb, taught in school).

having a cap in front of the mixing amp is a good idea indeed.

cheers, Iain

[theehman]

Thanks for the input, guys!  Here's my updated schematic with your suggestions.

[antonis]

Without individual output DC blocking caps you'll result into crackling trimpots..
(it's alright for set-and-forget pots but nuisance for working ones..)