I've been going down the rabbit hole to make a small board that enables using a guitar effect pedal either in series or parallel. I put together a prototype on a solderless breadboard, but the damn thing was so flimsy I got bad connections at every step.
Now, I've made a PCB prototype of the schematic above which largely works, but with a few issues:
1. when applying power, the "series" setting (which is just a bypass of the board) becomes significantly noisy compared to its unpowered state.
2. the "parallel" setting seems to have less overall volume and subjectively seems to lose some bass (though it's a bit difficult to gauge with certainty, volume drop affecting perception a bit)
For the first problem, I'm thinking that the issues lies in my switching scheme which leaves U1A's non-inverting input connected to Vref and its output still connected to the pedal input. I should probably use a 4PDT switch and not only route the input signal to the pedal input or through U1A, but also route the pedal input from the switch itself or U1A output. I'm thinking this kind of arrangement would completely take the opamp out of the equation when going using a series setup.
For the volume drop, I don't have any solid working theories. Maybe the RV1 + R15/R16/R17 arrangement is wrong. I took my inspiration from Rod Elliott's active mixing circuit (see Fig 4), but maybe I got it wrong.
Any thoughts on both issues would be much appreciated.
Now, I've made a PCB prototype of the schematic above which largely works, but with a few issues:
1. when applying power, the "series" setting (which is just a bypass of the board) becomes significantly noisy compared to its unpowered state.
2. the "parallel" setting seems to have less overall volume and subjectively seems to lose some bass (though it's a bit difficult to gauge with certainty, volume drop affecting perception a bit)
For the first problem, I'm thinking that the issues lies in my switching scheme which leaves U1A's non-inverting input connected to Vref and its output still connected to the pedal input. I should probably use a 4PDT switch and not only route the input signal to the pedal input or through U1A, but also route the pedal input from the switch itself or U1A output. I'm thinking this kind of arrangement would completely take the opamp out of the equation when going using a series setup.
For the volume drop, I don't have any solid working theories. Maybe the RV1 + R15/R16/R17 arrangement is wrong. I took my inspiration from Rod Elliott's active mixing circuit (see Fig 4), but maybe I got it wrong.
Any thoughts on both issues would be much appreciated.
Looking at C3 and R5 going into an OPA1662:
Firstly thats an RC time constant of about 0.22s, OK for an instrument without deep bass I think.
However the OPA1662 input bias current is 1.2µA worst-case, giving a DC offset of 1.2V, which isn't great (although that stage is AC coupled on the output).
You are switching the instrument input directly onto the output of U1A (at AC), which won't work as the opamp output is very low impedance, effectively shorting out the input, and maybe setting U1A oscillating. Only route signals to inputs, not to outputs.
Firstly thats an RC time constant of about 0.22s, OK for an instrument without deep bass I think.
However the OPA1662 input bias current is 1.2µA worst-case, giving a DC offset of 1.2V, which isn't great (although that stage is AC coupled on the output).
You are switching the instrument input directly onto the output of U1A (at AC), which won't work as the opamp output is very low impedance, effectively shorting out the input, and maybe setting U1A oscillating. Only route signals to inputs, not to outputs.