Need help with op amp control circuitry for resoelectric guitar

[crazydoc]

I'm building an experimental resoelectric with  magnetic and piezo-undersaddle pickups, and it's
getting time to put in the electronics. The magnetic PU will have the usual volume and tone controls,
and a polarity reversal switch. The piezo will run through a generic preamp mounted in the side of the
guitar. The additional circuitry (except for the pots) will piggyback on the piezo preamp, so the board
has to be small, about 1.5" X 2.5", and power will be from the 9v preamp battery.

Basically I'm trying to keep the mag and piezo circuits separate until the end, so they don't interfere
with each other. I'd like to be able to adjust the phase of the piezo signal, switch the phase control in
or out, blend the piezo and magnetic signals, then to a master volume control to the output jack. I've
attempted to design a circuit using a TL074 to provide buffers and an all pass filter, but it's way past
my simple understanding of electronics, and age related CRS syndrome is catching up with me.
I've tried using non-inverting and inverting configurations, but prefer the non-inverting with
apparent fewer parts count.

I'll post the non-inverting schematic. No component values, except for the allpass pot at 500k (it's the
only kind I have with a push-pull switch) and a 250k blend pot (it's what I have) if they'll work.

A few questions:

Are the resistors from the + inputs to VG needed?

Can the blend pot output go directly to the master gain pot, left as is to a buffer, or should the
outputs be separated and run through a summing amp?

When I'm switching the filter in or out, I'm connecting the input of the unused op amp to the virtual
ground - is this appropriate or necessary?

Will unity gain buffers be stable, or do I need to add something?

I'd appreciate any suggestions or other help you can give me - thanks in advance.

 

[abbey road d enfer]

Are the resistors from the + inputs to VG needed?

Yes.

Can the blend pot output go directly to the master gain pot, left as is to a buffer, or should the
outputs be separated and run through a summing amp?

You don't need a separate summing amp, Passive summing is pretty good.

When I'm switching the filter in or out, I'm connecting the input of the unused op amp to the virtual
ground - is this appropriate or necessary?

You can't connect the buffer output to the APF output. Move the switch to teh outputs, the inputs can be connected.

Will unity gain buffers be stable

Yes

I'd appreciate any suggestions or other help you can give me - thanks in advance.

You could use a single pot for the blend control. Connect the buffer outputs to the top and bottom of the blend pot and the wiper to the output buffer. Actually, you may not need an output buffer. If you used a lower value pot, let's say 100k, the output impedance would be less than 25k, which is perfectly adequate for most amps. Now if you want to connect to a line input with about 10k impedance, you need a 50k or less pot. In order to avoid DC scratch noise, you may want to connect a capacitor from one of the buffers output to the pot.

 

[boji]

Not sure if you ran across this in your searches, but: 

https://cafewalter.com/wp-content/uploads/2014/02/pzp1-sch.pdf

https://cafewalter.com/wp-content/uploads/2014/02/pzp-blend-bvt.png

 

[crazydoc]

Thanks to both for the input. I'd seen the piezo buffer on the CafeWalter site, but failed to notice the combining schematic at the bottom of the page.

So it looks like I can simplify - throw out the buffers and connect the mag and piezo preamp outputs through a single pot that would act like a blend pot, though a small amount of the other's signal would get through  the wiper at each end of travel, Or I could still use an actual blend pot.

So now I need to switch the allpass filter in or out. Can I keep the input connected to the preamp out, and switch the outputs with a SPDT, or do I need to switch both the in and the out of the filter from the circuit?

 

[abbey road d enfer]

though a small amount of the other's signal would get through  the wiper at each end of travel,

If this is a problem, you can buffer the mag p/u.

Or I could still use an actual blend pot.

That would not guarantee the absence of residual crosstalk.

So now I need to switch the allpass filter in or out. Can I keep the input connected to the preamp out, and switch the outputs with a SPDT,

Yes.

 

[boji]

Only other thing I can think of is to perhaps add a pole to the SPDT disconnecting  opa's V+ during bypass (assuming no thump would occur).

 

[crazydoc]

That would not guarantee the absence of residual crosstalk.

I thought the reason to use a blend pot was to be able to ground either of the inputs at the full excursions of the wipers to eliminate any cross talk.

Only other thing I can think of is to perhaps add a pole to the SPDT disconnecting  opa's V+ during bypass (assuming no thump would occur).

That would be easy to do, since the push/pull switch I'll use is DPDT - I was just going to use half of it.

 

[abbey road d enfer]

Only other thing I can think of is to perhaps add a pole to the SPDT disconnecting  opa's V+ during bypass (assuming no thump would occur).

You just can't assume that: actually, quite the contrary, there will be thumps.

 

[abbey road d enfer]

I thought the reason to use a blend pot was to be able to ground either of the inputs at the full excursions of the wipers to eliminate any cross talk.

You're right, I wrote before turning the brains on.

 

[crazydoc]

Thanks for all your help, M. d'Enfer.

Now for me the hard part - determining component values. I've pretty much pulled these out of my ass, depending on what I have lying around. I can easily substitute resistor and cap values because I have thousands. I mostly have 250k and 500k pots from building guitars, but I have some others. I'd prefer to use a 250k or 500k pot for the filter as it's the only value I have with a push/pull switch, but if necessary I could use a separate switch.

My most pressing question, that I can't find the answer by googling, is the relative resistor values for the filter. Are R1, R2 and (R3+R4) supposed to be of equal value? I can see that R1 should equal R2, but am not clear about the filter resistor. In any case, changing its value to change the filter frequency would preclude it being equal to the others anyway, but maybe it should be the same order of magnitude.

Also, are those reasonable frequency values for phase shifting? They're on the low end, but I could use a lower cap value to shift it up (or put in a rotary switch to have multiple cap choices. ;D  )

 

[abbey road d enfer]

My most pressing question, that I can't find the answer by googling, is the relative resistor values for the filter. Are R1, R2 and (R3+R4) supposed to be of equal value? I can see that R1 should equal R2, but am not clear about the filter resistor. In any case, changing its value to change the filter frequency would preclude it being equal to the others anyway, but maybe it should be the same order of magnitude.

Yes, R1 and R2 must be equal. R3 & R4 can be almost anyting, within reason. The values you have put are correct for a JFET opamp, but would be too high for a bipolar opamp.

Also, are those reasonable frequency values for phase shifting? They're on the low end, but I could use a lower cap value to shift it up (or put in a rotary switch to have multiple cap choices. ;D  )

You want to adjust phase of low ferequencies, because that's where it's the most audible. Your values are about correct, considering the range of the guitar. Yoyu may want to increase slightly C1 to 5.6nF, in order to reach the low E at 82 Hz.

Capacitor C1 must be terminated to ground. The half-voltage node must be connected to junction of R1 & R3.
Also move the 1uF cap to the switch's common.

 

[crazydoc]

I'm having a hard time understanding why C1 needs to go to system ground, rather than the half voltage virtual ground for the op amp. It looks like the cap goes to the system ground in a split voltage supply, which would be the virtual ground in a single supply.

https://www.sanfoundry.com/wp-content/uploads/2017/08/linear-integrated-circuit-mcqs-band-reject-all-pass-filter-q13.png

 

[abbey road d enfer]

I'm having a hard time understanding why C1 needs to go to system ground, rather than the half voltage virtual ground for the op amp. It looks like the cap goes to the system ground in a split voltage supply, which would be the virtual ground in a single supply.

The half-voltage node is undecoupled, so its impedance is 23.5kohms; in other words, it's not a virtual ground. That changes completely the response. You could decouple it with a large capacitor (>1uF), but then, you can't connect the node to the junction of R1 & R3, because it would short the signal to ground. The termination of the capacitor must be a point where there is no AC; it could even be the battery positive.

 

[crazydoc]

Thanks. It makes sense it should connect to a point where there is no other AC.

Also, I think, since this instrument is just an experiment, that I will use a 3 way pickup selector switch  connected to  2.2nF and 4.7nF caps to give me a 6.9nF choice so I can hear the differences (if any) sonically. One more thing sticking out of the guitar, but WTF.

Thanks again for your help here.

 

[abbey road d enfer]

Also, I think, since this instrument is just an experiment, that I will use a 3 way pickup selector switch  connected to  2.2nF and 4.7nF caps to give me a 6.9nF choice so I can hear the differences (if any) sonically. One more thing sticking out of the guitar, but WTF.

I believe you should do that only temporary and decide which value is best; there should be enough range of adjustment with the potentiometer.

 

[crazydoc]

I don't know if it's appropriate to post progress info here, but here is a pic of the populated board. Still have to add a number more of routing wires to pots and switches.

And BTW, I am going to use a selector switch for the allpass cap values - it already has 5 knobs and two switches - another one can't be much more confusing.

 

[crazydoc]

Just a note and a pic to bring this to a close. Unfortunately the allpass filter doesn't make any audible difference in the output, in any cap value, pot rotation, by itself or in combination with the mag p/u in either polarity. Oh well, it was an experiment. The guitar sounds OK with good timbre difference between the two p/u's, but not very dobroish. It might be better with a slide, but I can't really play round neck guitar, much less with a slide.

Anyway, thanks for your help.

 

[abbey road d enfer]

Just a note and a pic to bring this to a close. Unfortunately the allpass filter doesn't make any audible difference in the output, in any cap value, pot rotation, by itself or in combination with the mag p/u in either polarity.

The effects of phase tweaking can be heard only at large volume, when non-linearities appear, or when feedback is an issue.

 

[AMZ-FX]

I'm guessing that if you listen to the piezo pickup only, the sound will be thin and tinny.

This is because a piezo is bascially a small value capacitor, and if you couple it into a low to moderate impedance, you will get a high pass filter (thin sounding).

It may be okay once mixed with the mag pickup, but in solo, it may not be what you want.

If the input impedance of the amplifier stage is raised, there will be more bass response. Much depends on your piezo pickup and its rated capacitance.

Best regards, Jack

 

[abbey road d enfer]

I'm guessing that if you listen to the piezo pickup only, the sound will be thin and tinny.

This is because a piezo is bascially a small value capacitor, and if you couple it into a low to moderate impedance, you will get a high pass filter (thin sounding).

The original schemo mentions a piezo preamp that delivers the signal going to the mixer. I would think it is capable of driving the electronics without any issue. Let's hear what the OP has to say about it.