What makes a Box of Rock clean up so incredibly well?

[pedalbuild]

Built a Box of Rock quite recently. What struck me is that it cleans up insanely well with your guitar volume knob. It's almost like a switch from high gain to a very mild breakup. I have quite a few drives, but nothing else comes close.

But what makes it clean up so well? What in the circuit design causes or influences this? Any explanations much appreciated.

Schematic I found online:

[KarenColumbo]

(I'm glad this came up. Did not want to spam the forum with my never-ending supply of questions    )

[stallik]

I'd like to know as well. Best pedal I've found for doing this is Deadastro's Spitfire. Much better than my BOR so I'm now wondering if I've got something wrong with that build.

Simplistically, it seems logical that any circuit that derives it's distortion by overdriving an early stage would do this better but I'm finding that this is far from always the case. Some circuits appear to distort higher up the frequency range so the output of the guitar has a bearing on the effect as the impedance? changes and high frequencies roll off. Also, the output level of the guitar has a bearing just because there's more overdrive to start with

[alexradium]

major part of the overdrive distortion pedals works with clipping diodes,or opamps,they work good in one setting with max volume on guitar,this circuit mimics what happens in a tube preamp,where each stage contribute in part to the final sound and the distortion happens to be much more gradually,although jfets are even better than mosfets

[ElectricDruid]

But what makes it clean up so well? What in the circuit design causes or influences this? Any explanations much appreciated.

Like AlexRadium said, I'd guess this is down to the gain being staged progressively through the pedal, rather than all in one big hit. If you have several gain stages all with lower gains, you get lots of "approach to clipping" as first one and then another and then another stage start to clip. If you have one stage with a gain of x400, it's much more "BANG!" and you're in.

[Rob Strand]

There's a couple of features that stand-out to me:

The second stage doesn't actually have that much gain.  The 1M feedback resistor + 470k input resistor sets a low gain.  So if you back-off the guitar's volume there's not that much signal in the first two stages

The second stage is actually acting as a treble booster above 720Hz.  The 470k & 470pF + 1M's looks deceptively like an amp stage which normally has a HF small shelf boost.  However the second stage operates with a 1M feedback resistor which completely changes the behaviour compared to the amp style RC network.   The HF boost extends from 720Hz upto 4kHz or so, and hence has a lot of HF boost.

The post EQ undoes the HF boost and you get the common overall EQ which peak arounds 1kHz to 2kHz (actually 2.5kHz), followed by HF roll-off.   

Interestingly enough the overall response and EQ concept isn't enormously different to the Boss OD-3, except for the fact the OD-3 cuts the lows a lot more with the typical 720Hz high-pass (like a TS9).

So the pre/post EQ gives a transparent sound.

[amptramp]

For small signal levels, you would expect a clean boost.  All stages have unbypassed source resistors and gate biasing is directly from the collector to stabilize the operating point so each stage has relatively low gain.  The two-section lowpass filter after the tone control ensures the distortion products (which are all higher harmonics) do not dominate the sound even though you have a shelving highpass going into the second FET.

There are no diodes or nonlinear circuit elements in the signal path other than the amplifier stages themselves and since the bias comes directly from the drains, each stage is more or less centred within the linear operating range.  When the input AC voltage is less than a certain amount, this acts like a linear amplifier.

[antonis]

Strange design, though, 'cause 5k1 resistors come handy on E24 series..
(no need for 3k9 + 1k2..)

[bool]

...
So the pre/post EQ gives a transparent sound.
...

This and the slightly glassy/scooped "character" of the mosfets; imho if you subbed the BS170 for say 2N7000's, you would get more mid-forward overall "sound". (at least this happens in other type circuits).

[Vivek]

On another thread, someone mentioned that one of the reasons that the Fuzz Face changes so much with guitar volume is because the gain of the first stage depends upon the impedance of the guitar, which changes from inductive at high volume control settings to resistive at low volume control setting

So besides lower signal due to low volume control setting, there is also the additional effect of lower gain in the first stage, leading to greater cleaning up effect.

Does this Box of Rock also capitalise on the same idea ?


Any other pedals where gain of first stage depends on volume control setting on guitar ?

[Steben]

On another thread, someone mentioned that one of the reasons that the Fuzz Face changes so much with guitar volume is because the gain of the first stage depends upon the impedance of the guitar, which changes from inductive at high volume control settings to resistive at low volume control setting

So besides lower signal due to low volume control setting, there is also the additional effect of lower gain in the first stage, leading to greater cleaning up effect.

Does this Box of Rock also capitalise on the same idea ?


Any other pedals where gain of first stage depends on volume control setting on guitar ?

Only low input impedance circuits. Another paradox: guitar circuits should have high input impedance, unless when you want a low one 

[ThermionicScott]

On another thread, someone mentioned that one of the reasons that the Fuzz Face changes so much with guitar volume is because the gain of the first stage depends upon the impedance of the guitar, which changes from inductive at high volume control settings to resistive at low volume control setting

So besides lower signal due to low volume control setting, there is also the additional effect of lower gain in the first stage, leading to greater cleaning up effect.

Any other pedals where gain of first stage depends on volume control setting on guitar ?

The Blackstone Mosfet Overdrive does that trick.  Any series resistance from your volume knob adds to the Rin of the op-amp input stage, adjusting the gain downward.  My tedious and boring 4049 thread had me working through a bunch of classic inverter overdrives, and the 100k input resistor on the TSF/Red Llama hurts the cleanup effect in comparison to circuits that minimize (or eliminate) any fixed series input resistance.

I still think a dual-stage overdrive using an inverting op-amp into an inverter stage could be a real Tubescreamer-killer, with way better cleanup from the guitar and smoother overdrive from the inverter.  But people just want Tubescreamers.   

[Vivek]

Mad idea 1

Can we have a circuit with high input impedance

But we send very little DC from pedal to guitar

Or 100khz sine wave

Just to sense the setting of volume knob

Then we adjust some internal gain stages based on our reading of the volume control setting of the guitar ?


Mad idea 2

Rockman autoclean setting reads peak voltage of guitar signal and makes some conclusions from that.

Doesn't work perfectly with each guitar

But can that be used to guess volume control position

If peak is 100mv, assume volume control is low,if peak is 1.2V assume volume control is high ?

Mad idea 3
Can a reverse compander increase the difference of gain between high volume setting and low volume setting ?


Mad idea 4
There was an odd schematic with diodes being used from non inverting pin to ground. It needed high input voltages. Mostly for synth I suppose. Anyone remembers it. It seemed to expand signals that it clipped

Clarification:
Does it mean that all these "clean up so well with volume control" circuits need to be first in the chain to be able to do their magic ?

[Steben]

Clarification:
Does it mean that all these "clean up so well with volume control" circuits need to be first in the chain to be able to do their magic ?

Not necessarily...
A rather hard clipping device (think of a small region of onset compared to the output swing) will get very clean once you get the signal low enough. With a little volume change at the input you have large amount of harmonic change. This can be achieved at the end of a chain as well.

The thing about low impedance is twofold:
1. exponential change of gain. With the low input impedance, every addition of resistance in the source makes for a non linear function of gain change.
2. change in LCR behaviour as said before, primarily loss of inductive influence which leads to low pass effect getting diminished with lower volume control at the input.

A Rangemaster treble booster style circuit has low input impedance as well, but with a deliberate low value input capacitor as well.

[POTL]

Any circuit will clean up well with the volume knob if you adjust the gain CORRECTLY. As we know, MOSFet, unlike JFet, amplifies the signal without going into clipping in the range of the power supplied to it. If the power is 9v, then the signal from the guitar can be amplified in the range just above 0 and just below 9v, something like 4vpp. As soon as the AC gain crosses this mark, the signal will start to clipping, and the sound will be distorted, anything below will sound clear.

[antonis]

If the power is 9v, then the signal from the guitar can be amplified in the range just above 0 and just below 9v, something like 4vpp.

Delete a "p" from "vpp" or make it 8Vpp..

[POTL]

If the power is 9v, then the signal from the guitar can be amplified in the range just above 0 and just below 9v, something like 4vpp.

Delete a "p" from "vpp" or make it 8Vpp..

Thanks for the amendment. the main thing is that you understand me.

[Steben]

Fets are not that defined in their onset of clipping.
Especially MOSfets


What I mean is some devices will need a "little turn" on the knob to play with clean to dirt, such as the Fuzz Face. Others will need a lot of turning and a lot of amplitude loss to get really clean.
It all depends on what you desire.

Always thought of MOSfet inverters as a great last stage, mimicking "soft" power amps (non feedback). Large signals mean better S/N ratio etc etc.
The main reason for the multiple stages is the fact inverter chips have them and they beg to be used. A discrete inverter is a solution, precede for example by jFET stages or opamp stages, but matching the fets is the problem then.

[POTL]

JFets enter clipping faster, they have less headroom, MOSFets are much more flexible in this regard, and supplying more power to the transistor can further increase headroom. BJTs are close to MOSFet in this regard and have a large margin. If we talk about operational amplifiers, then the whole problem is that their distortion is unpleasant and diodes / LEDs are used to improve the sound, which negatively affect the dynamics. I've tried many constraints and MOSFet and BJT are the most flexible.

Returning to the reaction to the volume knob. Make the transistor gain strong and the volume knob will clear closer to the minimum value. Make the transistor gain less and the volume knob will clear almost immediately. If you make several successive cascades, then everything will depend on how you set up each one.

[Steben]

JFets enter clipping faster, they have less headroom, MOSFets are much more flexible in this regard, and supplying more power to the transistor can further increase headroom. BJTs are close to MOSFet in this regard and have a large margin. If we talk about operational amplifiers, then the whole problem is that their distortion is unpleasant and diodes / LEDs are used to improve the sound, which negatively affect the dynamics. I've tried many constraints and MOSFet and BJT are the most flexible.

I think it is true and not true. It all depends on definition and desires.
As in other threads got discussed, "dynamics" is a word with an array of meaning. linear and hard hitting the rail - the ulitmate headroom - is to some the ideal dynamics since nothing gets compressed. Wjile others use "dynamics" to describe the typical compressing dynamic behaviour of a device.