A better bass screamer

[Fancy Lime]

Hi all!

I recently found some isolated bass tracks from Metallica's Master of Puppets album. Although I am not a exactly a fan of Metallica or Cliff Burton's bass playing style, he was rocking some righteous overdrive sounds on that one. Metallica being praised for their sound, that's got to be a first. Anyway, seems like Cliff used a Tube Screamer, of all things, on that album for his main overdrive sound. I did not verify or research this properly, just found the internet rumor and it sounds plausible enough, at least to serve as an inspiration for my own breadboarding adventures. But of course, there is some massive bass boost after the TS in the recording or else the album would sound like ...And Justice for All It certainly sounds nothing like the polished parallel processed bass sounds popular in contemporary metal but is much more reminiscent of a good 90s crust punk bass sound.

So I finally had to try something that I just always assumed was a good idea but had never actually built or breadboarded and have never seen done quite this way for this purpose anywhere, despite my best efforts to find it. Namely, using a tube screamer style clipping stage with all its high-mid and treble boosting to distort the high frequencies much more than the low ones, and then just tacking a variable bass boost stage on to compensate for the boosted highs. This avoids all the weird phase problems you tend to get with parallel processing (which are not necessarily a problem if you keep the phase cancellations out of the fundamental range and like the slightly disjointed or "broadened" sound but I like my bass sound compact like it was chiseled from a solid piece of rock, thank you very much). It is also really easy to do if we use the second opamp half as an active bass booster stage. To avoid clipping the second opamp half, it seemed wise to move the volume control in front of that stage. The overall topology is quite similar to what DOD did on a lot of pedals in the 80s and 90s (I think), for example the FX50B. These things have a non-inverting opamp clipping stage, then volume control, then another non-inverting opamp stage as an active balance kind of control, that lets you boost highs if you turn the por right or lows if you turn it left. Since more treble boosting seemed pointless for the task at hand, I just made it a 12 db bass boost. For the clipping stage, I went asymmetrical, which to me sounds better on bass than symmetrical clipping. So it is really more of a descendant of a Boss OD-1 than a tube screamer. One nifty feature of clipping stages of the OD-1/TS variety, is that the gain pot and the cap parallel with it (C6 in the schematic) form a low pass, whose corner frequency shifts down as gain goes up. This means that with a carefully chosen cap in this position, you get a fairly consistent sound across gain settings, effectively eliminating the need for a separate treble control. I recommend making this cap switchable for more flexibility and adaptability to different amp sounds. 500p, 1n, and 2.2n worked great for me but YMMV. I don't miss the tube-screamer treble control because boosting treble is pointless for the intended purpose anyway and the feedback cap, which is much bigger than in a TS, takes good care of treble cutting.



So there you have it. A simple, "noiseless" (meaning less noisy than most overdrives), good sounding (says who? says I!) bass overdrive with some righteous crunch (but not high gain unless you increase the gain pot to 500k or 1M) and more than enough thunderous low end even on a five string. The pick or finger attack gets distorted quite audibly while the fundamental frequencies are only mildly clipped (depending on the gain setting of course). This avoids muddiness and retains the punch while adding a good amount of aggressiveness to the attack without sounding clanky. It responds very well to playing dynamics in that the sound changes a lot depending on how you play. At the same time, it of course compresses the output dynamics like all clipping overdrives without (excessive) clean blend, so you get a nice consistent volume without loosing your expressiveness (as can be the case with too much distortion or compression). For best results, serve with guitars through a dimed HM-2 or FZ-2, blast beat drums and vocals that make the neighbors wonder if you are strangling a giant parrot with Tourette's. Or do your own thing, I'm not the boss of you.

If someone who is more talented at it than I am wants to design a layout for this, that'd be great of course. If not, I will do my best to design a perf layout but I am super slow with that and quite busy at the moment, so it may be a while.

Please let me know if you find any errors (Antonis, you there?) that need addressing before making a layout. Since this one is probably mostly interesting for beginners, I should add: please do not hesitate to ask if any parts of the circuit need explaining or if you have mod requests. Oh and because it is going to come up: yes, you can use a 4558 in this but if you do, I suggest changing R2 from 1M to 470k. Or better yet, just use a TL072, as in the schem. The chip was chosen for a reason.

Cheers,
Andy

[Mark Hammer]

Thoughtfully developed.
My own sense of the thought behind the TS was that the intent was to make the degree of clipping more consistent across the fingerboard.  All other things being equal, lower notes have greater amplitude, so attenuating the bass prior to clipping, or rather applying less gain to the lower end at the clipping stage, results in not as much clipping of lower notes as there might be otherwise.  Clipping diodes in the feedback loop produces a degree of compression that further evens out notes..  At least in theory.

But should there be a DC-blocking cap somewhere between U1A and U1B, given that you're also feeding U1B a bias voltage?

[aron]

Mark, is there a way to put a capacitor in series with the feedback loop diodes in the tube screamer (like they were mentioning in the big muff) to do the same thing with the clipping? I haven't tried it yet.

[Mark Hammer]

Mark, is there a way to put a capacitor in series with the feedback loop diodes in the tube screamer (like they were mentioning in the big muff) to do the same thing with the clipping? I haven't tried it yet.

Worth considering.  The BMP can be shrill, but it is rarely fizzy, even at max treble.  That arrangement with the cap and diodes might be why.  I recall seeing drawings where different cap values were used in conjunction with the diodes.

[Fancy Lime]

Thoughtfully developed.
My own sense of the thought behind the TS was that the intent was to make the degree of clipping more consistent across the fingerboard.  All other things being equal, lower notes have greater amplitude, so attenuating the bass prior to clipping, or rather applying less gain to the lower end at the clipping stage, results in not as much clipping of lower notes as there might be otherwise.  Clipping diodes in the feedback loop produces a degree of compression that further evens out notes..  At least in theory.

But should there be a DC-blocking cap somewhere between U1A and U1B, given that you're also feeding U1B a bias voltage?

Hi Mark,

Thanks! The consistency across the fretboard thing is quite true for guitar and may well have been the original reason behind the 720Hz high pass ground leg. However, with the different frequency spectrum and higher string tension of a bass and all strings being fat and wound changes the situation somewhat. A bass is more consistent across the fretboard amplitude-wise but has more of a dynamic difference between the attack, especially when using a pick, and the decaying part of the note. This is the effect I am utilizing here. So, same technique as on guitar, quite different outcome.

Blocking cap: we could put one in there but because U1A has 0db of DC gain, we don't really need it. There is only the few mV offset voltage across the volume pot, which should not cause problems. The TS doesn't have a blocking cap there either and also has a resistor to V/2, although I don't know why, that resistor in the TS seems completely redundant to me. The Boss SD-1 omits that resistor.

Mark, is there a way to put a capacitor in series with the feedback loop diodes in the tube screamer (like they were mentioning in the big muff) to do the same thing with the clipping? I haven't tried it yet.

Hi Aron,
Yes, that is indeed the other underappreciated way of making a TS-style overdrive better suited for bass, imo. However, it is a bit more finicky to design because of the interaction between the gain pot and the "bass clipping avoidance cap". In my experience this approach works better with the gain pot on the ground leg like in the DOD 250. It can also get quite boomy at higher gain, which can be counteracted by putting an additional (variable) resistor and/or another pair of diodes in parallel with the cap that is in series with the original diodes. Still a good way to design a bass overdrive but slightly more difficult to get right. I have done extensive experimentation on this technique but no real publication-worthy design. I will try to cobble together some schematics to illustrate the important points.

Cheers,
Andy

[snk]

Interesting design, thank you Fancy Lime!

[Vivek]

Thanks for sharing the schematic and design logic.

Since the clipping is asymmetrical, does it make any sense to change Vbias so that it can handle larger swings ?

Would a small cap be needed in parallel to C7 ?

[niektb]

The value of C10, is it meant to be 4u7?

[Mark Hammer]

You know, you wouldn't think that after all these years, and literally hundreds and hundreds of Tube Screamer-related threads, there'd be anything more to learn.  But I guess you'd be wrong.

Thanks, Andy.  Much appreciated.

[Vivek]

Some of Barber's circuits have mid boosted clipper stage + Bass recovery second stage (what he calls as the RIAA style recovery)

[duck_arse]

I like the way the screamer 220nF tone cap is now in the feedback loop instead.

[Fancy Lime]

Thanks for sharing the schematic and design logic.

Since the clipping is asymmetrical, does it make any sense to change Vbias so that it can handle larger swings ?

Would a small cap be needed in parallel to C7 ?

Hi Vivek,
the clipping diodes are oriented the way they are on purpose (one on negative, two on positive swing) instead of the other way round with the asymmetry of the TL072 in mind. But the output asymmetry is pretty small and the input protection diodes D2 and D3 clamp input spikes down hard enough so that we don't run into a situation where the input symmetry needs to be considered with a 9V supply.
A small cap || C7: necessary? Probably not unless you use a very low quality C7 with lots of series resistance. But adding a 100nF ceramic cap here is cheap and probably best practice. Unless you are in a very RF-infested environment and don't have the thing in a metal box, it probably won't make a difference, though, since the bandwidth of both active stages is very restricted.

The value of C10, is it meant to be 4u7?

Hi niektb,
oops, my bad. Good catch, thanks! Fixed it now. If you happen to have a 4.7kF cap, there are probably better uses for it than a guitar pedal. And it's probably going to be the size of a fridge.

Some of Barber's circuits have mid boosted clipper stage + Bass recovery second stage (what he calls as the RIAA style recovery)

Hi Vivek (again ),
interesting, do you know which ones and/or have a schematic? Always curious about parallel evolution.

I like the way the screamer 220nF tone cap is now in the feedback loop instead.

Hi Duck,
that is actually a coincidence. I picked the 10k pot because that's a reasonably common value and 220n just gave the right frequency response with that. However, this indeed makes modding the tone stack of an actual TS9 into something resembling this here a sweet little bit neater. I'll write a little How to Mod post for that as soon as I get around to it.



And finally, as promised, here is Aron's idea from earlier with a few bells and whistles, implemented in a single opamp stage. You could leave out the treble pot and add a TS tone stage for more flexibility. It is not really worked out or tested, just drawn from memory after some experimentation on the breadboard. I find this design a bit difficult because there are a few interactions that need to be considered. First, it does not allow (very) asymmetric clipping because C4 will end up in the middle of the diode drops of D3 and D4 after a few cycles (depending on the Bass setting), no matter if D3 and D4 are the same or different. But that is the same for almost all clippers to ground and it doesn't seem to faze people there either. Also, the gain control in the ground leg makes it difficult to add a treble boost for additional clarity (the "Rat Trick" with two ground legs), which can be done easily if the gain pot is in the feedback loop. But with gain in the feedback loop, it messes with the frequency response. On the other hand, maybe someone more clever and/or patient than me can use that wisely to their advantage. D1 and D2 are optional and can be asymmetrical. I recommend them if you want to avoid opamp clipping. But if you use an LM308 or 301 for this thing... Rats, now I want to build one. See what you did Aron



Cheers and thanks for the kind words and input,
Andy

[Vivek]

Barber Electronics LTD Silver :
Mid boosted Clipper
Muff style tone
Bass recovery stage with internal trimmers




Bass fixed at 10K, different values of Mid



Mid fixed at 10k, different values of Bass

[marcelomd]

Love the idea. Makes total sense, since that's how a bass amp distorts. The holy Ampeg Growl. The thing with parallel paths tries to emulate a biamp setup, which sounds great too, in a different way.
Check out the Exotic BB Preamp. It's a TS with a fixed tone control into a Baxandall tone stack. Works great on bass.
IIRC Cliff used a Morley Power Wah Boost. I started this whole pedal thing because of him. That said, I'm not a fan of his tone.

[Fancy Lime]

Huh, interesting. The bass recovery in the LTD Silver is really quite similar, right down to the choice of corner frequency and maximum boost.

The Exotic BB is also interesting. I have seen a bunch of TS + Baxandall designs here and in other DIY forums but I was unaware that there was a commercial version.

Two more that should be mentioned for completeness' sake are the Digitech Bad Monkey and the Way Huge Green Rhino. They use a gyrator based bass control in addition to the normal TS treble control, so it's a bit more complex but the basic idea is similar.

Parallel distortion: it really depends on whether both paths are more or less full range (but with some tone shaping) or there is an actual crossover separating two frequency bands and treating them differently. The latter is way more complicated and I cannot remember ever having seen it in a pedal but some producers use that approach in their DAWs. It should cure the phase problems of the "two full range paths" solution.

Andy

[niektb]

In my TS clone I had a Bridged-T centered around 400Hz (and the original 720Hz HPF), works absolutely massive on bass. Have a friend that uses it all the time that way 
I can dig up a schematic if you're interested 

[Elektrojänis]

First, it does not allow (very) asymmetric clipping because C4 will end up in the middle of the diode drops of D3 and D4 after a few cycles (depending on the Bass setting), no matter if D3 and D4 are the same or different.

How about having a separate cap for each polarity of the diodes and not connecting them at the center? You would need to separate the D1 and D2... And duh... A dual gang pot to separate the bass control for them.

[marcelomd]

The Exotic BB is also interesting. I have seen a bunch of TS + Baxandall designs here and in other DIY forums but I was unaware that there was a commercial version.

Two more that should be mentioned for completeness' sake are the Digitech Bad Monkey and the Way Huge Green Rhino. They use a gyrator based bass control in addition to the normal TS treble control, so it's a bit more complex but the basic idea is similar.

Parallel distortion: it really depends on whether both paths are more or less full range (but with some tone shaping) or there is an actual crossover separating two frequency bands and treating them differently. The latter is way more complicated and I cannot remember ever having seen it in a pedal but some producers use that approach in their DAWs. It should cure the phase problems of the "two full range paths" solution.

Andy

I have a Mesa Boogie Flux Drive, which is an almost 1:1 BB Preamp with a charge pump. I think the Wampler Clarksdale is a TS+gyrators.

The Darkglass X series of pedals lets you control the frequencies of the parallel paths separately. I think it's a proper active crossover (and not simply a low pass/high pass). Low side has a compressor and the high side has the overdrive.

In my TS clone I had a Bridged-T centered around 400Hz (and the original 720Hz HPF), works absolutely massive on bass. Have a friend that uses it all the time that way 
I can dig up a schematic if you're interested 

Oooo I am! =)

[Fancy Lime]

In my TS clone I had a Bridged-T centered around 400Hz (and the original 720Hz HPF), works absolutely massive on bass. Have a friend that uses it all the time that way 
I can dig up a schematic if you're interested 

Yes, please! In fact, the very first pedal I ever built does something a bit similar. It was based very very loosely on the DOD FX91 (which itself is kind of a TS for bass), has a parallel clean path (but buffered and boosted, if I remember correctly, unlike the FX91), and a switchable bridged T at around 800 Hz after the mixer. I have been meaning to trace it for a while (back when I built it, I just built straight from idea to perfboard, no breadboarding, no schematics. I am still amazed that this thing works at all), maybe now is a good time for that.

First, it does not allow (very) asymmetric clipping because C4 will end up in the middle of the diode drops of D3 and D4 after a few cycles (depending on the Bass setting), no matter if D3 and D4 are the same or different.

How about having a separate cap for each polarity of the diodes and not connecting them at the center? You would need to separate the D1 and D2... And duh... A dual gang pot to separate the bass control for them.

No, that does not work because the caps just keep loading an loading on every cycle if only one diode direction is connected to them. That shifts the threshold up and up (or down and down). So this is not suitable for getting asymmetry. However, if you add a (variable) resistor between the two C-D junctions, the excess charge can discharge through the resistor. This makes a very nice "adaptive clipping threshold" circuit, great for a very natural dynamic overdrive. I'm sure there is a patent for this thing somewhere. I have used the same configuration for crossover distortion to simulate the increasing crossover distortion with increasing amplitude that some tube amps have. Sounds quite good. Always wondered if anyone else used that sort of thing, I have never seen it elsewhere. If you are interested, I can try and work up a bit of a detailed explanation and minimum working circuit.

Cheers,
Andy

[Vivek]

Please help me understand the differences in Bass and Guitar signals.

Suppose we assume that the Guitar pickup output is 1Vp for few cycles, what is the equivalent data for bass ?

Suppose we assume 720Hz is good place for mid hump for Guitar, what is the equivalent for Bass ?

Suppose we assume high pass filter of 60 Hz is good for Guitar, what is the equivalent for Bass

Suppose we assume 5000 Hz is upper limit of Guitar Speaker, what is the equivalent for Bass ?

Thanks