Bass/guitar signal level references

[marcelomd]

Hi,

Is there any reference for signal levels that are useful for amp/effects design?

Let's say I want a clean-ish preamp that does some low gain overdrive. The first stage is a soft clipping opamp and I want the edge-of-breakup to be at +- 50% gain. What is the resistor/potentiometer values I should use? The answer is different for a 10mV and a 200mV signal.

In the same vein, how do I know if I have to attenuate some signal before an active tone control?

Reason I'm asking is I won't have access to a breadboard for a while, so I'm simulating a lot and ordering assembled boards with a few spare resistor/capacitors to mod my first idea. It's kind of wasteful, I know, but it's what I can do for now.

Thanks!

[Matthew Sanford]

I'd think with various signal levels even from one guitar depending on pickup switch settings, and considering the inaccessibility of breadboards, you should use a trimpot as variable resistance for Rf on the piano to figure what value resistor to replace it with

[antonis]

Brute solution..

Use the gain for the lowest signal amplitude you wish for with input pot at 100%..

[Fancy Lime]

Guitar and bass pickup output voltages are all over the place. Peak voltages go from under 100mV to several Volts. If you really want to cover all cases, I would try to make the range of the gain pot as large as it reasonably can be. With an audio taper gain pot you can easily do 50/1 (highest over lowest gain factor). At 100/1, the pot resolution at the low end becomes shaky. I would go with 50/1 and add a 3-way switch that allow to do -12db/0db/+12db. Or even -20db/0db/+20db. This should cove everything that you can reasonably expect to encounter.

But for a more middle of the road approach, I would assume 100mV to 2V peak voltages and 20mV to 500mV during most of the sustain phase. I sometimes put two antiparallel red LEDs to ground before the input stage to tame extreme signals.

HTH,
Andy

[ElectricDruid]

+1 what Lime said.

Actual guitar levels are hugely variable, from low-output pick-ups on classic instruments to super-hot modern active pick-ups on some total metalhead's favourite axe...you have no idea what is coming in, in short.

The best you can do is try and be as flexible as you can. That doesn't necessarily mean trying to accomodate *everyone*. Covering 90% is a reasonable position, and if you reckoned on 100mV to 2V as "typical" signal levels, you'd probably get *most* people. That's "only" a 20:1 range, a "mere" 26dB of gain.

The type with the super-hot metal pickups probably won't care that they get distortion through your pedal (might not even notice, let's face it), and the one with the late-50's original tele is probably used to it being too quiet on a lot of modern setups. They can use some extra gain somewhere. They can afford it

[Rob Strand]

Even the definition of level is up for grabs:

The type of level measurement:
- peak to peak, peak, rms (,average, quasipeak ...)

The time scale you are interested in:
- peak over whole signal (ie. the initial peak)
- rms over initial peak region
- rms after the peak has largely passed

Typical plucking vs worst case hard plucking.

The peak over the whole signal with hard plucks are used to prevent clipping but measurements after
the peak with typical plucks would be useful for a distortion.  For a compressor you need to check both.

https://www.muzique.com/lab/pick.htm
https://sound-au.com/articles/guitar-voltage.htm

I think 1V to 1.5V peak is a good figure of the worst case peak with hard plucks.

[Fancy Lime]

To expand a bit on what Tom said: consider the likely equipment of the target audience for the device you are building.
Are modern metal players with their eight string guitars and superhot Duncan Invader pickups likely to want to play a clean-ish to light overdrive preamp? Probably not. Or if they do, they will blast it with a distortion pedal anyway and hardly notice your intricate edge-of-breakup sound design.
50's Tele players with special underwound pickups may be a bit more relevant to consider in your case. But they almost certainly already own an overpriced vintage treble booster, which they never turn off (because these things didn't even have an off switch in ye olden days).
So what you really want to be designing for, is probably post-1970s Strat to slightly hotter than PAF with a bit of wiggle room on both sides. Considering your breadboard handicap, your best bet may be to look at the gains and clipping thresholds of preamps or pedals that you know to have about the right amount of distortion with "normal" Strats, Teles, or Les Pauls. YouTube pedal reviews are not the worst place to try and get an overview of the designs for which you may want to have a look at the schematic and figure out the gain/clipping structure.

Andy

[marcelomd]

Thanks a lot guys.

My idea is simply to have a range, or reference, to design around. I guess you all are right with the sweet spot being between a strat and a PAF. I'm making this for me and my friends anyway. The only extreme case there is my G&L bass that puts out an obscene 3Vpk.


I like the idea of trimpots. Maybe add the directly or add a few pads to solder them off board. No easy way out with capacitors, though.

This is what the latest project looks like. Note the row of extra components for mods:

[teemuk]

Is there any reference for signal levels that are useful for amp/effects design?

At least many guitar amps seem to expect about 20 - 40 mV nominal input.

Needless to say, a 40 mV "nominal" signal can be what you have when the initial 3Vpeak picking transient has faded down. 

Most input stages aiming towards transparency expect worst-case input of about 5Vpeak. AFAIK, there's very little complaining about distortion with such design. It could be that the very short-duration transient distortion is practically inaudible or that most pickups simply don't even peak above 5 volts.
Typical tube input stage has headroom of maybe 1.5Vpeak and its distortion is certainly not inaudible, yet it works just fine in applications complemented for their pristine cleans.

Let's say I want a clean-ish preamp that does some low gain overdrive. The first stage is a soft clipping opamp and I want the edge-of-breakup to be at +- 50% gain.

Given wide variety of pickups and input signal ranges you are asking too much. Tune things for distortion at 50% gain on weak pickup tele and it's way off when someone plugs in a humbucker or an acoustic with an active pre.  ...or basically just any random effect pedal.

At best you can settle to a compromise and ponder about options such as adding a "Pad" feature to tame the hottest input sources.

What is the resistor/potentiometer values I should use? The answer is different for a 10mV and a 200mV signal.

The answer is also totally dependable on where and in what circuit you are planning to use that potentiometer.

The lower the resistance the lower the noise, but that's just one piece of the puzzle.

In the same vein, how do I know if I have to attenuate some signal before an active tone control?

You can calculate some likely input levels and how much distortion is tolerable. At best, you audition.

By the way, you can integrate (soft) clipping to most active tone control circuits too. It's not "hi-fi" but might certainly help in a musical instrument preamp where moderate clipping is tolerable if not preferable.

Reason I'm asking is I won't have access to a breadboard for a while, so I'm simulating a lot and ordering assembled boards with a few spare resistor/capacitors to mod my first idea. It's kind of wasteful, I know, but it's what I can do for now.

You can probably simulate with audio waveform input matched to proper amplitude levels...?
(It will certainly provide better idea of how to adjust gains for real life signals and how much transient clipping you is needed to even hear typical "distortion" than simulations with generic sine inputs).

[marcelomd]

Is there any reference for signal levels that are useful for amp/effects design?

At least many guitar amps seem to expect about 20 - 40 mV nominal input.

Needless to say, a 40 mV "nominal" signal can be what you have when the initial 3Vpeak picking transient has faded down. 

Most input stages aiming towards transparency expect worst-case input of about 5Vpeak. AFAIK, there's very little complaining about distortion with such design. It could be that the very short-duration transient distortion is practically inaudible or that most pickups simply don't even peak above 5 volts.
Typical tube input stage has headroom of maybe 1.5Vpeak and its distortion is certainly not inaudible, yet it works just fine in applications complemented for their pristine cleans.

That's awesome. I believe this is only for the first 12AX7, right?

Let's say I want a clean-ish preamp that does some low gain overdrive. The first stage is a soft clipping opamp and I want the edge-of-breakup to be at +- 50% gain.

Given wide variety of pickups and input signal ranges you are asking too much. Tune things for distortion at 50% gain on weak pickup tele and it's way off when someone plugs in a humbucker or an acoustic with an active pre.  ...or basically just any random effect pedal.

This is exactly the question. It's impossible to handle both vintage blues and evil metal at the same time. What I want is to develop some intuition about gain scaling and how to make things work for the average case (which I have no idea what looks like).

By the way, you can integrate (soft) clipping to most active tone control circuits too. It's not "hi-fi" but might certainly help in a musical instrument preamp where moderate clipping is tolerable if not preferable.

I believe Tech 21's PSA1 does that. IIRC 1 or 2 controls are a Baxandal mid control with diodes in the feedback loop.

Reason I'm asking is I won't have access to a breadboard for a while, so I'm simulating a lot and ordering assembled boards with a few spare resistor/capacitors to mod my first idea. It's kind of wasteful, I know, but it's what I can do for now.

You can probably simulate with audio waveform input matched to proper amplitude levels...?
(It will certainly provide better idea of how to adjust gains for real life signals and how much transient clipping you is needed to even hear typical "distortion" than simulations with generic sine inputs).

Good idea. I certainly can do that as soon as I get back home to my gear. No idea when that will be, though. Half of my city is underwater and most roads there are blocked.