Kingslayer: OD drops volume

[forsbach]

My Kingslayer sounds great, but when I engage the overdrive the volume drops substantially. I've got BAT46's in for the OD. Anyone else have this issue?

i did alter the switch from the original BOM. Instead of using a spst switch listed in the bill of material I used a dpdt stomp switch. One side of the switch is activating an LED.

[jkokura]

One of the reasons that a clipping choice switch is often on a toggle is that it does create a large volume boost to go with no clipping diodes. When you engage the diodes (turn your switch on), it reduces the volume as a matter of course. There is no way to not reduce the volume, it's part of what clipping diodes do by nature.

I'm not sure, but there may be a way to reduce the volume when you have the diodes off. Perhaps introducing a Resistor in line with the output? Might mean you need to switch to a 3PDT switch...

Jacob

[midwayfair]

I'm not sure, but there may be a way to reduce the volume when you have the diodes off. Perhaps introducing a Resistor in line with the output? Might mean you need to switch to a 3PDT switch...

Here's one way:
Remove R12

Use a DPDT for the OD switch. Use one pole as normal. For the second pole, run a wire from lug 3 to one R12 pad, and another wire from lug 1 to the R12 pad. Now put a pair of resistors from the outside poles to the common: 1K on the side where OD is ON, and let's say 22-47K on the other side, when it's off. You'll probably have to experiment. When the switch is flipped either way, it'll short one resistor and leave the other in-circuit. Of course, that may wreak all sorts of havoc with the hard clipping because it cuts signal before the diodes.

So maybe a better way is:

Remove C11

Use a DPDT for the OD switch. Use one pole as normal. For the second pole, run a wire from lug 3 to one C11 pad, and another wire from lug 2 to C11, then to the other C11 pad (C11 will be floating in space ...). You want those on the side where OD is OFF. Put a resistor across lugs 2 and 3. When you flip the switch to OD on, the resistor will be shorted, and signal will just pass through C11 with no added resistance. When the switch is flipped off, the resistor is in series with C11. Again, try 22-47K.

This method avoids changes to the amount of clipping available but could actually produce some strange low-pass interactions with C12.

[forsbach]

So this is a common problem with this circuit? It's odd to me that volume only drops when the OD is engaged. It doesn't change volume when selecting no diode, D8/D9, or D4/D5 on the diode select switch.

What about adding a 3PDT switch that activates a resistor in parallel with R24? The switch would connect the resistor to the circuit when the OD is off to bring the volume down. Would adding more impedance at that stage of the circuit alter the sound much?

[jkokura]

So this is a common problem with this circuit? It's odd to me that volume only drops when the OD is engaged. It doesn't change volume when selecting no diode, D8/D9, or D4/D5 on the diode select switch.

Forgive me, I'll try again. I think a good part of your issue is that you're looking at the volume drop as a problem. What I'm telling you is that it's not a problem, it's how this works. The volume is going to be different when you engage clipping diodes, it is that way because of how the op amp feedback loop diodes work.

The difference is that the other diodes are in line with the circuit, whereas the OD switch is in the feedback loop of the op amp. In the Op Amp stage, the gain is being determined. In other words - you'll get a volume cut in the gain, and more distortion obviously, when you introduce diodes. Remove the diodes, and the volume goes way up, and the distortion goes away.

When you get to the clipping to ground (clipping switch) area, the volume is already determined. You then have a set of diodes which again clips the signal, but it's not affecting the volume as drastically.

The OD switch is not meant to be put, as you have done, on a footswitch. It's not that you can't figure out how to make it work and both you and Jon have made some suggestions on how that might happen. Your suggestion of putting something in parallel with R24 may work, however you'll need to work out your calculations. Resistors in Parallel don't add to each other, they divide, so the math is a little bit complicated. For example, a 10K and a 10K in parallel is about 5K, and a 560R and a 10K turns out 530R. Long story short, I think using a switch to choose between two resistors for R24 would work better than trying to add a resistor in parallel.

Jacob

[midwayfair]

So this is a common problem with this circuit?

I wouldn't call it a "problem" -- it's how clipping works.

To produce distortion, a wave needs to be amplified until it exceeds the voltage restraints of the circuit. Anything that exceeds that restraint is clipped -- the top of the wave is lopped off because it can't get any bigger. Any additional gain simply pushes up more transients and harmonics to have levels similar to the fundamental (which we hear as distortion). If your signal exceeds the power rails on an amplifying device, it distorts. That's "headroom." If you exceed the headroom, your signal won't get louder. (This is why running some pedals on higher voltages or having an amp with more wattage produces less distortion at the same volume levels and has more output volume -- because you've increased the headroom.) Diodes artificially create less headroom by limiting the peak signal size allowed to continue along in the circuit. In this case, the diodes in the feedback loop of U2 limit the signal to .6v peak-to-peak before they are clipped, instead of the maximum 9v/-9v swing of the op amp (I'm not sure what the resulting signal size is, because my math isn't so good, but the gain looks like it's x40 for some frequencies and x100 for others, so it must be at least 15v ... I think). The maximum signal signal size with the soft clipping diodes is MUCH smaller than it would be without them.

It doesn't change volume when selecting no diode, D8/D9, or D4/D5 on the diode select switch.

Either you're mistaken, or your switch isn't working. With the diodes lifted, your signal's max size is big enough to oscillate the following op amp (U3), meaning it's HUGE. With the germanium diodes in the circuit, the signal is clamped to ~.3v, which is tiny compared to with the diodes lifted. [Edit: It's 1.7v peak-to-peak with the LEDs, or almost 6x the size of the germanium diodes, and thus should be half again as loud. Voltage gain-to-decibels is a logarithmic scale.]

Now, it's possible you're not hearing too big of a change in this switch with the OD on ... that's because the OD is also clipping the circuit before it reaches these diodes. You should have more distortion total.

[forsbach]

Thanks guys that makes a lot more sense. I think I need to play with the pedal a bit more to see if I want to put the OD back on a switch or try and even out the volume and use a footswitch.

Here's a couple questions for the brain trust.
1.) If I did opt to bring down the volume when the OD is off which would have the least impact on tone? Adding a resistor in SERIES with R24 or adding a resistor to C11 as suggested earlier?
2.) Would it be easier/possible to raise the volume when the OD is engaged? I'm thinking something involving a transistor...? I know a transistor will impact tone, but the OD is supposed to be different.

[midwayfair]

1 - doing it in such a way as to have minimal impact on tone is super difficult. Basically everything has an impact on tone ... The weather in China ... Etc. You're going to affect something, whether it's the frequency response or the impedance or the gain levels.

2. Much harder/more complicated. You're talking about splicing in a transistor stage somewhere in the padal in such a way that it won't affect the distortion or overdrive the transistor ... Very tough.