So the top circuit in post 1, would it work if I removed Q1, Q2, Q3 and R6? The signal would be at 0V I think, midway between rails.
And jumper R2-C5

No - because you would have removed the buffer, and the input impedance would be too low for the pickup to drive.
The input impedance would be approximately equal to R7 instead of R2.
You could replace Q1, Q2, R4, R5 by a single resistor, and keep J1, and that would work.

 

The old NE5532 opamp had a low input impedance for an old phono pickup, not a Jfet input opamp like a more modern TL071.
Also the opamp is a Sallen-Key filter that must be fed from a low impedance.

 

The old NE5532 opamp had a low input impedance for an old phono pickup, not a Jfet input opamp like a more modern TL071.
Also the opamp is a Sallen-Key filter that must be fed from a low impedance.

NE5532 introduced 1979
TL071 introduced 1978
Both still going strong. Any audio desk is still full of NE5532 op-amps, but not quite so many as were used before DSP took over all the processing.
As Self points out, the NE5532 takes a lot of beating, because it has a very low current noise for an op-amp designed for low voltage noise, but the TL071 is better for high impedance inputs.
A more modern JFET amplifier would be preferred to avoid the TL071's phase-reversal problem.

 

Here's another guitar pedal preamp section that runs on +/- 15v and doesn't use transistors, with unity gain. The diodes are just polarity protection I think? Does the pull down resistor set the impedance? or is that R2?

 

This picture from post #1 is also non-inverting with a gain of 1+ 47/2.2 = 22.4 ?

 

Here's another guitar pedal preamp section that runs on +/- 15v and doesn't use transistors, with unity gain. The diodes are just polarity protection I think? Does the pull down resistor set the impedance? or is that R2?
View attachment 311210

RPD in parallel with the series combination of R1 and R2 sets the input impedance.

 

This picture from post #1 is also non-inverting with a gain of 47/2.2 = 21.4 ?
View attachment 311211

Actually it's 1+47/2.2 for what difference it will make.
But it is attenuated by 220k/(10k+220k) by R1 and R2

 

Comparing those 2 with the original circuit in post 1, I don't think the transistor arrangement provides any gain, right?

 

Comparing those 2 with the original circuit in post 1, I don't think the transistor arrangement provides any gain, right?

View attachment 311212

Yes. The voltage gain will be about 0.95, but the impedance is changed from 10MΩ to <1kΩ

 

View attachment 311210

ok this one has an input impedance of 2M2 || 5k6+560k is 450k. Unity gain buffer like the transistors circuit. Does the op amp significantly reduce impedance like circuit in post 28? Is that desirable?

 

Yes. The voltage gain will be about 0.95, but the impedance is changed from 10MΩ to <1kΩ

The <1k you’re referring to is R8 in parallel with R9 and R10, right?

 

The opamp is a 2nd-order highpass filter to reduce 60Hz and lower frequencies.
The original circuit has a few useless parts.
R8, R9 and R10 do not reduce the input impedance because R8 does nothing and R9 produces positive feedback to make the cutoff a sharp corner when fed from a low impedance.

Here are the frequency responses of the original and simplified circuits:

 

OK I see that U1A is a 2nd order HPF, and understand how R8, R10, R11 can create 6db gain. Is it possible to replace Q1, Q2, Q3 with an op amp (U1B) to reduce the impedance like in post #24? Like this? I'm just curious, I'm not sure which is better. Although 3 transistors have less overhead than an op amp from a layout standpoint.

I understand how R1, R2, R4 create the input impedance. I'm not exactly sure what impedance is optimal for acoustic guitar although 1 MΩ is generally good for most pickups.

 

The NE5532 opamp does not have Jfet inputs, instead they are ordinary transistors. Then its datasheet says its minimum input impedance is only 30k ohms which is much too low for a guitar pickup and is even too low for a phono cartridge.

Modern opamps with Jfet inputs have an input impedance of 1 Tera ohms (1 million x 1 million). Then the external resistors determine the input impedance fed by the guitar pickup.

Post #24 uses a TL072 that has Jfet extremely high input impedance inputs.

 

The <1k you’re referring to is R8 in parallel with R9 and R10, right?

No, it's the output impedance of a JFET source follower.

 

I see from the datasheets that tl07x has a 10^12 input resistance, and the NE5532 has a min of 30k with 300k being typical. OK so a TL07x should work as an input buffer for a guitar, like post 24. Both the circuits in post 24, the older LF356 in post 25 and the transistor circuit use a FET, as I understand it, have high input impedance and low output impedance.

 

Still begs the question, which FET is better - transistor or op amp? A 2N5457 and 2N5088 are very common on the transistor side, also the J201. On the op amp side of things, I've seen a lot of JRC4580's and a few TL072's. It does seem like a preference for TL072/OPA2134 for applications that want to minimize distortion like synth and chorus pedals.

 

I see from the datasheets that tl07x has a 10^12 input resistance, and the NE5532 has a min of 30k with 300k being typical. OK so a TL07x should work as an input buffer for a guitar, like post 24. Both the circuits in post 24, the older LF356 in post 25 and the transistor circuit use a FET, as I understand it, have high input impedance and low output impedance.

View attachment 311253

You've got it!

 

Still begs the question, which FET is better - transistor or op amp? A 2N5457 and 2N5088 are very common on the transistor side, also the J201. On the op amp side of things, I've seen a lot of JRC4580's and a few TL072's. It does seem like a preference for TL072/OPA2134 for applications that want to minimize distortion like synth and chorus pedals.

JFET vs. JFET op-amp: The op-amp will give better distortion, the JFET better noise but only if you choose a good one.
A good FET (2SJ3557) will give <1nV/√Hz, but a 2N5457 is more like 40nV/√Hz, which is worse than the TL072. The TL072 (and NatSemi's LF353) have the weird phase-reversal problem which they have fixed in the OPA2134, which also has better noise than the TL072.
For most JFETs and JFET op-amps you can ignore the current noise (quoted at 10fA/√Hz for the TL072 and 3fA/√Hz for the OPA2134). The current noise multiplied by the input resistance adds to the voltage noise.
A bipolar op-amp such as the NJM4580 has lower voltage noise (3nV/√Hz) , but it's current noise isn't quoted. I'd guess at 1pA/√Hz, and that would make it virtually unusable at high input impedances.

 

The 2N4392 (from original circuit) has a noise of 1.2 nV/√Hz, similar to the 2SJ3557. TL07x says 18 nV/√Hz typ at 1khz. So the FET transistor could be better in that way.