Hi,

I made a simple guitar signal booster for leads. It works, but now I want to add dry / wet mixing. That is, mixing the input with the output. I know proper mixing should be done with an opamp but I don't have space for that. I've also read that sometimes you can get away with mixing two audio signals, each through a resistor.

So naively I added a pot, RV1, as shown in the attached diagram. Stuff added is in the blue rectangle. This eats almost all the signal. Maybe I wired it wrong. Should this in principle work?

Thanks,
hrs

 

If the Box is strictly a "Booster",
then there will be no difference between Dry and Wet,
therefore, what You would have is a 2-level, preset, Volume-Control-Switch.

However, You do appear to have a Filter in the Circuit, which
I haven't taken the time to calculate the response of.

U1A should be moved to the Input to provide a Low-Impedance-Drive for the Filter-Circuit.
Where U1A is located in your Schematic will not have
any effect on the performance of the Circuit.

R5 does nothing.

The RC4558 Op-Amp is an "updated" copy of the "infamous" 741 Op-Amp,
and both of these Op-Amps are a very poor choice for any Audio application.

There are probably better Filtering configurations that could be used.

What specifically do You want the Filter to do ?

What is the Voltage of your Power-Supply ?

Where is the "Stomp-Switch" ?

I would setup a simple 3-Band-Tone-Control instead of a fixed Filter,
or, better yet,
2 separate 3-Band-Tone-Controls,
each with their own separate Output-Volume-Controls, to switch between.
.
.
.

 

If the Box is strictly a "Booster",
then there will be no difference between Dry and Wet

It's a bandpass (~800Hz to ~2400Hz) and a gain stage. There will be a difference.

Where U1A is located in your Schematic will not have
any effect on the performance of the Circuit.

I see what you mean. The bandpass filter is often shown with a voltage follower but I didn't consider that it's not needed here.

R5 does nothing.

I added it to avoid plopping. But R3 and R4 will do the same job, which I failed to see.

What is the Voltage of your Power-Supply ?

9V.

Where is the "Stomp-Switch" ?

The last one here minus the led.
What about the passive mixing with a pot, should that work? What value for limited attenuation?

 

"True-Bypass" is a "Gimmick", and plenty of people buy into it.
A properly designed Amplifier is completely "transparent", and
has zero need for "Bypassing".

Cheap Circuits, built with a bare minimum number, of the absolutely cheapest parts available,
are the problem which creates the apparent "need" for a "True-Bypass" switching arrangement.

R3 and R4 do nothing to reduce switching noise.

Do You want to know how to build a far more useful Boost-Box ?
.
.
.

 

Do You want to know how to build a far more useful Boost-Box ?

No. I want to know about passive mixing with a pot.

 

You will need a Buffer-Amp on each end of the Pot.
The Pot needs to have a minimum of 2K-Ohms Resistance,
a higher resistance-value will require a third Buffer-Amp for the Output-Jack.

Learning by making mistakes is the only way you'll really learn anything, so have at it.
.
.
.

 

The simplest way to start designing a Circuit is to use "Building-Blocks".

Each "Block" must have a High-Impedance-Input, and a Low-Impedance-Output.

Each Block must also be designed to operate within a well defined Voltage-Range.

Then You can connect the Blocks in various arrangements to achieve your end goals.
.
.
.

 

Since this wont work without additional opamps I went with an alternative option of wiring a 500k pot across R3. This gives some control between on vs off. Performance is acceptable for this stopgap circuit until I have time to make something better.

As you noted the buffer after the bandpass does nothing so I may use it for an active mixer, something like the revised circuit in the attachment. Though I'm not too sure if I should pick the input from before (as shown) or after the input coupling capacitor C6.

 

J1, ( Input ), is not High-Impedance.

Your 1/2-Supply-Ground may not have a low-enough-Impedance,
Current flowing to Ground from the Filter-Network
could affect ( bleed-into ) the Output-Signal from the Final-Output Op-Amp.
This may, or may not be, significant, considering the "Low-Fidelity" Guitar application.

A "Wet-or-Dry" "Fader" is not the best plan,
unless You intend to actually adjust it while You are playing.
Switching between 2 preset-levels is a more useful plan.

And, adjustable Filter(s) are much more useful than a single fixed-Filter.
.
.
.

 

J1, ( Input ), is not High-Impedance.

I understand. It is driven from a buffered output so it's good enough for now.

A "Wet-or-Dry" "Fader" is not the best plan,
unless You intend to actually adjust it while You are playing.
Switching between 2 preset-levels is a more useful plan.

And, adjustable Filter(s) are much more useful than a single fixed-Filter.

Would you mind expanding on this? Why adjust while playing? Can you provide a diagram example of 2 preset-levels?

My work in progress is the attached chorus derived from the Little Angel Chorus. But it may need some tone shaping and gain/volume control.

 

.
.
.

 

Another use for the extra opamp is to make the filter do what you think it is doing, but is not.

With two R-C sections in series, the second section appears as a load in parallel with the shunt element of the first section. IOW the two filters interact, causing both corner frequencies to be *not* what they are calculated to be individually. There are two ways around this. One is to put a buffer stage between the two sections. It can be a gain stage without affecting the filter characteristics. The other is to use a true bandpass filter topology, such as multiple-feedback, that takes these interactions into account.

AND - don't worry about using the 4558. It was the backbone of 70's and 80's amp design. Peavey used them in many (most) of their amps. Other than some input-stage noise, it is fine for limited-bandwidth instruments such as guitars. Note that some "purists" don't like it because it is *too* transparent. That's why 301's and 709's are cherished by pedal crazies (and I mean that in a good way).

If you want to build something that sounds right for that era, stick with the 4558. If you want something truly transparent, the NE5532 (LM5532, MC5532) is a drop-in replacement. I resurrected a KB-60 for my brother, and Peavey had the spectacularly good sense to use DIP sockets. I changed all of the 4558's to 5532's, and improved the system noise by 9 dB.

ak

 

Another use for the extra opamp is to make the filter do what you think it is doing, but is not.

With two R-C sections in series, the second section appears as a load in parallel with the shunt element of the first section. IOW the two filters interact, causing both corner frequencies to be *not* what they are calculated to be individually. There are two ways around this. One is to put a buffer stage between the two sections. It can be a gain stage without affecting the filter characteristics. The other is to use a true bandpass filter topology, such as multiple-feedback, that takes these interactions into account.

For this reason I made R2 much larger than R1. Did I get that the wrong way around, i.e. should I have made R1 an order of magnitude larger than R2 to reduce loading? It sounds a bit like how phones used to sound. If I decide to redo the circuit I may widen the band a bit.

Regarding the 4558 there is some chatter about how the magical sound of this or that amplifier is due to the limited slew rate of the 4558 or stuff like that. I'm not too concerned with that. I happened to find it in some discarded electronics and decided to use it and save my better opamps for another occasion.

 

For a band-limited instrument playing through an even more band-limited circuit, the 4558 is perfect for the job.

Except for the magic part, the chatter is correct; there are very real sonic differences among circuits. Note that I am *not* saying all audio claims are equally valid, which is why my speaker cables are in fact lamp cord with lotsa oxygen.

ak