TL072 18V single supply pickup buffer to 25K MN blend pot

[jesterbass]

Hello all,

I have come up with this:

It is a 18V single supply (bottom left) dual magnetic pickup buffer, based on the TL072 op amp (one half of the chip shown here, for one pickup).

Pickup is modeled very simply at the bottom.

Since the buffer uses single supply, we need to generate a mid-supply bias voltage using a voltage divider (left). This voltage is used to bias the non inverting inputs of the TL072. I am injecting the bias voltage to the inputs using Rbias1 = 1MΩ.

The op amp is obviously in non inverting configuration and gain is unity, no amplification.

Since the op amp is working with the signal at a 9V bias level, input and output coupling capacitors are necessary, to remove the DC component from the previous (pickup) and next (potentiometer) stages. They are 0.22uF and 4.7uF respectively.

I also added Rin1 = 470Ω as a small input protection and Rz_out1 = 510Ω as some output protection.

The main question I have is what do I do after the op amp output. The output coupling 4.7uF cap is certainly a necessity. But the two resistors I have next is pure speculation on what would be best to lead this output into one of the wiper inputs of a 25KΩ, MN taper potentiometer. I mention the potentiometer specifically because the main intention behind building this buffer is a smooth blending experience across the whole tonal spectrum between the two pickups. Something like a home brewed active blend control a la EMG ABC.

What do you guys think would be best post the op amp output and into the blend pot? If you have other recommendations/comments about the circuit they are very welcome. I 'd like to start prototying this on a zero breadboard so I would like to have a better idea as to what I will need to wire up together before a start.

Many thanks!

 

[ceetee]

Not competent enough to answer any questions and not convinced, I've understood everything you are describing but highly interested in upcoming input!

Btw... TL071 or TL072? Schematic vs Text.

 

[jesterbass]

Not competent enough to answer any questions and not convinced, I've understood everything you are describing but highly interested in upcoming input!

Btw... TL071 or TL072? Schematic vs Text.

I had to make a schematic with half a TL072 somehow!

 

[gerryjazzman]

Here’s something similar that RobbieK did a while back. I’m guessing this is what you are aiming for:



He uses a single linear pot for this. MN isn’t necessary in this design. Take a look. He has a pdf and strip board layout for it. Maybe that’s all you need.

 

[ceetee]

I had to make a schematic with half a TL072 somehow!

Sure. Just wanted to clearify. People find & use a schematic and wonder what went wrong.

 

[gerryjazzman]

Wanted to add one more thought. I guess you could try an MN pot at the output of the opamps but you’d have to put a series resistor at the output of each opamp, maybe something like 2.2k or so since at center detent on the MN , both opamp outputs would be directly connected together which won’t be good since the opamp outputs are essentially zero impedance. You could give that a try if you want. You could probably see what difference the two pot configurations would make in simulation.

 

[byacey]

Since you are running 18V, why not split the supply, +/- 9V? I would also increase the input impedance to 10 meg.

 

[jesterbass]

Here’s something similar that RobbieK did a while back. I’m guessing this is what you are aiming for:



He uses a single linear pot for this. MN isn’t necessary in this design. Take a look. He has a pdf and strip board layout for it. Maybe that’s all you need.

Thank you for this! It will be studied carefully as I love seeing how other people approach this stuff. The idea that the outputs may be effectively blended with a simple linear potentiometer is fascinating and an unexpected bonus.

 

[ctmullins]

I like it! Reminds me a bit of late-period Kandinsky, but with the orthogonality of Mondrian.

 

[jesterbass]

Wanted to add one more thought. I guess you could try an MN pot at the output of the opamps but you’d have to put a series resistor at the output of each opamp, maybe something like 2.2k or so since at center detent on the MN , both opamp outputs would be directly connected together which won’t be good since the opamp outputs are essentially zero impedance. You could give that a try if you want. You could probably see what difference the two pot configurations would make in simulation.

That's a good point. This specific op amp can tolerate having its outputs grounded, although I'm sure it would be easier on the chip to always see some resistance at its outputs (part of why I added the external output resistors). Plus, other op amps might fry if connected this way.

 

[jesterbass]

Since you are running 18V, why not split the supply, +/- 9V? I would also increase the input impedance to 10 meg.

I agree that a 18V single supply looks odd when you can have two independent batteries for the two rails. In fact, one of my first prototypes was as you describe, with such advantages as having the input/output signals referenced from true ground, less need to isolate the buffer stage from actual ground level, no need to use a voltage divider for bias voltage. In short, a much simpler design.

But, there's a catch. You need to disconnect two batteries when pulling your plug out, and it can't be done with a single contact between ring and sleeve. I did s little searching and a latching dpdt relay can be used, with some minimal driving circuit to close/open it when the cable is inserted/removed. I will explore that in time because I like your concept, but I am not willing to deal with its complexities (and potential reliability issues) right now. And I really can't justify an additional dpdt toggle on the bass in order to turn a simple buffer on/off. Thank you for your input nonetheless!

 

[Jerry Catanescu]

I use a 2.7k series resistor on the output of my onboard stuff. It makes it easy to parallel 2 or even 3 gadgets without resulting in too low of a combined load for the op amps, and it interfaces nicely with common 25k pots downstream. No particular reason why I used 2.7k, as opposed to say 2k or 3.3k, other than the fact that I already used this value elsewhere in the circuit. Alembic also uses 2.7k on the output of their filter preamp, FWIW.

Your buffer looks good to me. I personally wouldn't bother with Rin1 in an onboard design, since the input never gets disconnected from the source so there's no need to limit any input current surges that might happen when you plug in a cable in a pedal scenario, but this is really a minor thing. I would, however, use a lower power op-amp onboard, such as OPA196 or OPA205. Longer battery life, and also better headroom (rail-to-rail output) - but they're SMD only, if that's an issue, and a few bucks apiece.

For a unity gain buffer, I don't see the point of 18V power, especially if you'd use a rail-to-rail opamp. It will never clip.

 

[jesterbass]

I use a 2.7k series resistor on the output of my onboard stuff. It makes it easy to parallel 2 or even 3 gadgets without resulting in too low of a combined load for the op amps, and it interfaces nicely with common 25k pots downstream. No particular reason why I used 2.7k, as opposed to say 2k or 3.3k, other than the fact that I already used this value elsewhere in the circuit. Alembic also uses 2.7k on the output of their filter preamp, FWIW.

Thanks for your input!

So, what you're describing is:
op amp output > output coupling capacitor > 2.7k resistor 25k pot wiper lug?

I'm asking because I've arbitrarily added a 25k resistor to ground after the output coupling capacitor, just because it looked nice.

PS. I will gladly remove the input resistor to remove parts count and save space.

 

[Jerry Catanescu]

So, what you're describing is:
op amp output > output coupling capacitor > 2.7k resistor 25k pot wiper lug?

Yes. The 2.7k value is negotiable, but seems reasonable to me. Too much and you lose some volume to the resistive divider formed with the pot that follows.

 

[jesterbass]

Yes. The 2.7k value is negotiable, but seems reasonable to me. Too much and you lose some volume to the resistive divider formed with the pot that follows.

I'm having trouble trying to estimate the input impedance each output will "see" (post resistor), going into each 25k MN pot wiper, with the typical passive-pickup-blend wiring applied on the pot. I am almost certain that a) it is a function of the pot's rotation and b) it could be quite low, so the voltage divider effect you mention could be very pronounced, meaning losing valuable volume. Can you help with giving me a better idea of what the input impedance will be?

Also, from an earlier project posted by a forumer, the designer there just connects the two outputs to the outer lugs of a B10k potentiometer and takes the output at the wiper. I wonder how this would behave compared the previous solution with series resistors plus blend pot.

 

[Jerry Catanescu]

^ You're right about point a) above. The lowest impedance the op amp sees is the value of the output series resistor alone, if the blend pot is in a position where it pulls it all the way down to ground.

As for b), you do lose a couple of dB to the divider formed by 2.7k and 25k, assuming the pot is at 50% or more rotation, where its resistance from wiper to bottom log is the full 25k.

The solution from the image above works fine and is actually more elegant I didn't think of it because my circuits tend to have a series output resistor so they can be easily passively summed through a pickup selector, for instance. That resistor would prevent completely turning off one or the other side with a single gang pan pot. This might still work fine in practice, though, since if one pickup is, let's say, 20dB below the other, it's virtually as good as being off, you likely won't hear it over the other.

 

[clickclack]

Just want to mention my fav opamp, the TLC2262 for bass & guitar fx. Low power, low noise, rail to rail(more swing / headroom with lower voltages). Surprisingly overdrives gracefully if you design it to. Disregard distortion and slew rate. Much much worse opamps on paper are in legendary status now because of how they sound on a guitar/bass circuit. You want some distortion anyway even if you don't know it.

 

[Jerry Catanescu]

Just want to mention my fav opamp, the TLC2262 for bass & guitar fx.

I think I've briefly used it before. Nothing wrong with it, except maybe that it doesn't do 18V. I just found something I like better. OPAx196, or if I need even lower noise (for a bit more power), OPAx205.