How to increase headroom and gain on a pedal

[spargo]

I have a question about headroom and gain in an overdrive pedal circuit specifically.  I've heard people say they've modded a pedal to add "3x the headroom, 4x the gain" etc.

My question is, how do you increase headroom and gain on a circuit, and how do you know by what amount it has increased?

Thanks for any help!

[CynicalMan]

Headroom refers to the input level a circuit can handle without clipping. In an overdrive pedal, which is meant to clip, the headroom would be very low. Gain is the output level divided by the input level, whether it's voltage, power, etc. Basically it's how loud a circuit makes the input. Gain often, although incorrectly, refers to the amount of distortion in a pedal or amp.

The easiest ways to increase headroom in a circuit are to increase the supply voltage (unless there's diode clipping), reduce the gain of a gain stage, or reduce the input voltage within the circuit. Otherwise, you can design the circuit itself to clip later, for example using rail-to-rail op amps. To increase gain, you need to increase the gain of a gain stage in the circuit, add more stages, or remove voltage dividers and filters in the signal path.

You don't really want to increase the headroom of an overdrive pedal unless you want less overdrive. Which pedal or pedals were you looking at?

[spargo]

Thanks for the reply!  I'm referring to an "improvement" on the Marshall Bluesbreaker circuit, where they claimed to have modded it to have 2x the available gain and 5x the headroom.  I was wondering how that was measured and how you would go about doing so in the circuit.  From my understanding, the bluesbreaker is not a very heavy overdrive in the first place.

[jacobyjd]

Before we can help, we need to establish exactly what 'headroom' and 'gain' both mean to you.

It sounds like 'gain' is the amount of distortion you can achieve...

...and 'headroom' is the amount of volume you have available once you set the gain.

Is that correct?

[CynicalMan]

I assume you're talking about this:
www.jhspedals.com/JHS_Pedals/JHS_Morning_Glory_Overdrive.html

It seems he's claiming to have increased the clean headroom by 5 but it also seems he's saying he increased the maximum distortion by 2.

[spargo]

Well, possibly.  I'm fairly new to the term "headroom", and am most familiar with gain in microphone applications or knobs on guitar pedals (which will typically add crunch or distortion).  Part of the difficulty is I don't know what headroom and gain are referring to in this instance.

So I'm not sure what they're referring to in this specific instance, but they said there was increased gain and increased headroom on a marshall bluesbreaker circuit, and I'm wondering what that may mean and how you would possibly do that in the circuit.

[Mark Hammer]

There is dynamic range, and there is headroom.  Sometimes those two are the same, and sometimes they aren't.  IN this instance, they aren't.

"Headroom" usually refers to the amplitude the device can tolerate, given what comes in and how much gain is applied, before that signal simply runs out of room.  If you are 6'8" and the ceiling is 7'2", there is only so high that you can jump before you run out of "headroom".  On the other hand, someone who is 5"2" has a LOT more headroom in the same place.

When diodes are used for clipping, they impose a fixed maximum amplitude.  That is in fact how they DO their clipping.  However, even though the ceiling is fixed for any given diode, it can be altered by changing to a different type of diode, or different numbers of diodes.  Ge diodes generally set a maximum of around 250mv on how much signal can pass before clipping, but if you string 3 of them in series, the 250mv adds up and the ceiling is now 750mv.

LEDs are often used for clipping circuits, and have a clipping point several times higher than a regular silicon diode.  That has a good side and a bad side.  The good side is that, with the clipping ceiling set much higher, there is more output because the limit on output is set higher.  That means you can pummel the amp harder.  And with a much higher ceiling, there is more room for producing audible dynamic differences in volume via your picking.  This is undoubtedly what the Morning Glory folks are referring to when they talk about headroom.  The bad side is that when the ceiling is lifted the portion of your signal that the diodes can clip becomes very limited.  And if you want to make it clip like before, you have to increase the gain considerably to reach the same relationship with the clipping threshold.  Think of it like putting Gene Simmons' boots on the guy who's 5"2".

Now here is where it gets sticky.  In principle, if the diode-based clipping threshold is set higher, you have more "headroom".  But since the intention of such a box is to produce clipping, then you need to apply gain to get the signal up to the amplitude where it will clip.  Many op-amps typically found in such pedals are unable to swing and wider than within 1v or so of the "rails".  With a 9v supply, the "rails" will be 0v and 9v, so do not expect the chip to be able to produce a clean signal that swings from 3.5v below the reference point (4.5vdc) to 3.5v above it.  Indeed, in many cases, asking for +/-3v is asking a lot.

Remember that while you will see declarations of pickups being able to put out 1V p-2-p, that is only the initial attack transient.  The signal drops considerably right after that initial surge.  So let us say that a person has used a pair of LEDs for clipping, yielding a threshold of +/-1.5v that must be met for any clipping to occur.  Let us also say that your Strat puts out 400mv peaks when you slam it, but that simmers down to a 25mv signal while you work your finger-vibrato magic on the B string up on the 12th fret.  Now, to get the 25mv steady-state signal to anywhere within spitting distance of the clipping point, and assuming a maximum +/-3.5v swing, you'll need a gain of at least 140. 

Of course, when you apply a gain of 140 to a 400mv transient, and you're using an op-amp powered by 9v, it's pretty clear you can't get a 16v swing with a 9v battery.  So, even if you had no diodes of any type within a mile of the pedal, asking it to provide a gain of 140 with a 9v battery will most certainly get you a signal that has very clearly run out of headroom.  Add diodes into the equation and the pedal will not only produce clipping because of the diodes, but will also produce clipping because of the total lack of sufficient headroom to accommodate that amount of gain, given the input signal.  The tail end of the note may be modest and just fine, like our 5"2" person above, but that initial transient pick attack will be like asking the 6"8" person to jump on a springboard.  Quite frankly, he's not gonna get much lift.

So to summarize.  Can you increase the clipping ceiling beyond where it already is in a Bluesbreaker?  Yes.  Can you do it "cleanly"?  To a very limited extent.  Ultimately, you will either sacrifice intensity of clipping, or will require the sort of increase in gain that the chip itself will clip.  I might also caution that having lots of gain in a pedal increases its susceptibility to hiss.