A different approach to TS midrange honk

[Mark Hammer]

The Tube Screamer, and all of its various cousins, grandchildren, great-grandchildren, etc., have used an approach to clipping in which the driving signal is equalized, pre-clip, to result in roughly equivalent distortion for lower notes/strings and higher ones, as well as for chords (which will have lower strings in there) and single notes.  That attenuation of the bass end of things, pre-clip, is complemented by an attenuation of the treble end of things, post-clip, such that the resulting sound has more body than stridency.

But the outcome of that strategy has also been to produce what many describe as a "midrange honk" that countless alternative design and mods have been directed at eliminating.  As much as they try, it is difficult for them to escape the reality that playing anything on a wound string yields a hotter signal and pushes the signal much closer to clipping than an unwound string.  That doesn't mean it sounds bad, but it does mean that the harder you try to keep the bandwidth, the more the goal of smooth string-to-string equivalence in clipping eludes one's grasp, and the harder you work towards that second goal, the more you have to sacrifice i the way of bandwidth.  Just one of those perpetual dialectics.

So here's what I'm wonderig.  Have any of you folk ever experimented with post-clip mid-scoops, intended to restore tonal balance?  That is, everything about the TS formula is followed to achieve smooth clipping, but another active or passive stage is inserted to provide a broad shallow-ish attenuation of those gronky mids, so as to yield a less midrangey sound.

[Paul Marossy]

Recently I stuck the TS tonestack on the output of a circuit I have been playing around with. It's kind of an interesting tone control. It goes from a mid scoop sound at the most CCW position into sort of a varying mid boost and then it kind of all evens out towards the most CW position.

How it reacts as a tone control is also dependent on whether there are clipping diodes involved or not. I surmise that is due to the amount of harmonics you get between clipping diodes vs. no clipping diodes.

[Mark Hammer]

I suspect you're correct.  The clipping obviously adds harmonics, but the 1k/.22uf network shaves the top off such that primarily lower-order harmonics remain, which is part of what gets the midrange honk.

It,s not the filter itself, so much as what there is to filter.

[Paul Marossy]

I kind of like it. In my circuit, I've actually been working on taming the high end even though I have that thing stuck on the end of it. A lot of the end result depends on what kinds of tone shaping is done before it gets to the tone control.

[DougH]

I wonder how much feedback diode clipping influences it?

I've generally had a lot better luck with an "open" (i.e. unmiddy or "unmuddy") freq response -and- smooth distortion etc using an op amp to drive another stage into clipping, rather than using diodes in the feedback loop.

[Paul Marossy]

I wonder how much feedback diode clipping influences it?

I've generally had a lot better luck with an "open" (i.e. unmiddy or "unmuddy") freq response -and- smooth distortion etc using an op amp to drive another stage into clipping, rather than using diodes in the feedback loop.

That also results in what I consider to be a more amp-like response to my playing.

[Steben]

I wonder how much feedback diode clipping influences it?

I've generally had a lot better luck with an "open" (i.e. unmiddy or "unmuddy") freq response -and- smooth distortion etc using an op amp to drive another stage into clipping, rather than using diodes in the feedback loop.

another stage = active clipping (transistor) or passive (diodes)?

I'm thinking experimenting with tone clippers (caps in series with clipping diodes) instead of pre-filtering can eliminatie midhonk.

[WGTP]

http://www.aronnelson.com/gallery/main.php/v/WGTP/BassBoostNotch.jpg.html?g2_imageViewsIndex=1   

[Mark Hammer]

Thank you Mr. Thick.  That's pretty much what I was expecting.  What I wonder, though, is whether anyone has actually tried it and how it sounds.

[petemoore]

  Adding a pot then the I eyed this, it looks like the kind of somethings that can go right acoross 2 pots, if the 1k - 100k is a pot and it's next to the volume control, it could be solidly woven across the relevant lugs.
 Tube Reamer just happened to be opened up here..for a resetting of my 'waveover' boost control parameters.
 It is quite nice circuit but the stringent is a bit strong right now, this works out fine with mine if preboost or postdistorted, and is a sound all it's own so to speak.
 The midrange honk can be 'worked with' and is a cool 'frequency field' to work with, but getting there for me requires re-do of other controls and tends to sound 'bottlenecked' when I engage...not surprizing still but not exactly listener/user friendly from bypass or other to TS midhump.
 Going 100kL shopping in 'basket', see if I can dig up some 10k~22k's and a couple caps.
 Should be interesting and I'll let ye know !
   This modification works good enough that I decided to change my signature.
   Much easier to work with excellent range of control [10k/22k is what I found here], [a ''104'' on top and a .039uf were found], small 100kL fit right in !
  So does the tone control I have now, I left the preferred treble-cut control where it was, next to the volume pot, adding in the 'bass boost' easily and by far the best 20 minutes I've ever spent playing around with TS type circuits modifications, absolutely necessary in my case, as the effect approaches 1/2 the way to 'normal', the stringent-offender frequencies replace the smooth-fill tone, this knob Will stay at a about 1/2, no more.
  Thank you all for sharing such a useful effect converter ! !
  As soon as I collect the correct very small handful of components, another one will be placed in the chain.
 

[DougH]

I wonder how much feedback diode clipping influences it?

I've generally had a lot better luck with an "open" (i.e. unmiddy or "unmuddy") freq response -and- smooth distortion etc using an op amp to drive another stage into clipping, rather than using diodes in the feedback loop.

another stage = active clipping (transistor) or passive (diodes)?

I'm thinking experimenting with tone clippers (caps in series with clipping diodes) instead of pre-filtering can eliminatie midhonk.

Yes, an active stage is what I was referring to. Reference Aron's "Shaka" overdrive series for examples. Those sound much less "honky" than tubescreamers to me and achieve the same overall goals (and some much more).

[ayayay!]

But the outcome of that strategy has also been to produce what many describe as a "midrange honk" that countless alternative design and mods have been directed at eliminating.  As much as they try, it is difficult for them to escape the reality that playing anything on a wound string yields a hotter signal and pushes the signal much closer to clipping than an unwound string.

Yes but, isn't this all a result of the design?  The culmination of the connecting + input and the - input w/ tone control, being blended upon output?  I've seen (and tried) a couple pedals where I just tie off the - input to the output, insert tone control somewhere, and use something like your SWTC.  + input gets input signal (no - input connection.)  Honkiness solved.  

...heh, there's a white joke in there somewhere...

Edit:  Found it.  Check out the R.F. Drive.  I think that's basically what I used in some version of some something circuit something.  

[jacobyjd]

...heh, there's a white joke in there somewhere...

I FINALLY HAVE THE FUNK!

That said, your solution is nice and simple. I may check that out sometime...then again...getting rid of the mid hump in a tubescreamer almost...you know...defeats the purpose of it...?

[ayayay!]

I just edited it a little.  Be sure to double check it.  

then again...getting rid of the mid hump in a tubescreamer almost...you know...defeats the purpose of it...?

Sure!  That's what makes a TS a TS, right?  Meeself, I like to play around with the RC on the tone control and manip it a little bit to be more of a high rolloff that doesn't boost the mids, when rotating CW.  Rather you find a mid freq you can live with and just rolloff the highs.  I think the LPE might be a good example off the top of my head.  

[petemoore]

then again...getting rid of the mid hump in a tubescreamer almost...you know...defeats the purpose of it...?
 Changes the function, that's for sure. perhaps: Transforms the purpose of it ?
  I submit the purpose of the design started with part of the ad copy claims...somthing to the tune of "Makes your amp sound like cranked tube amp"...
 Since 'it' has no idea what or where it is going to be a small part of some complex equation, in this theory, the design figured narrowing the frequency band to take a load off the downwind equipment will make a bass friendly amp sound mid hump, an amp easily swamped by bass will have mid-hump instead of mud-humps.
 Kind of like '1 size chair fits all'...make the seat short and a tall guy won't have to climb up and can fit in [albeit plunked angle-like on buttbone with knees pointing diagonally upward...], everyone fits easily into it.
 A taller seat height is greatly preferred, as are taller table heights for this 6'2'' guy, I got tired of leaning over and having air under my upper legs from lack of seating support, so I build a box and put a riser on seats as first order of business when working or getting seated 'correctly'...
 Same thing with the TS...fits any amp = narrower frequency band = less complex and demanding signal for the any amp it will be plugged into.  
 Since I like having the amp pushed across the frequencies I like to have it pushed by, not the ones convetion decided would be goodnuffer any, this is the ticket to TS type tone I can switch to without the gnark-tone killin' the otherwise very usable vibrations it now makes the speaker produce.
 Sounds of TS are all-in, but the added bass/cut mid-hi's what have you transformed it into 'taller work and comfort platforms' for the amps that accept and delivers a much wider spectrum that what the TS lets through.
  It took all the supressed thoughts I had about TS voicing such as 'is this a joke or isn't it working right?' and made them valid, while allowing the bypass switch to be hit without the 'Dirt-Dump-Hill' eq shape require amazing listener tolerance and knob dialing which only decreased the intolerance slightly.

[MikeH]

That said, your solution is nice and simple. I may check that out sometime...then again...getting rid of the mid hump in a tubescreamer almost...you know...defeats the purpose of it...?

Totally- just build something else right?   

[WGTP]

I use the above reverenced notch filter all the time.  My bedroom/lab amp is a Valve Jr. with a 12" sealed enclosure.  The Valve Jr. has no tone controls, so the notch filter is the main "voicing" control.  Since I primarily use breadboard distortions into the VF, constant tweaking of the frequency response occurs, primarily in the form of capacitor value changes.  As Mark said, most distortions have the lows and highs rolled off for various reasons and I'm done with the mid heavy '80's sound.  You can get anywhere from mild fendery/marshaly mid reductions to a big old metal notch.  I usually settle for something in between.  By adjusting the resistor prior to the notch filter, you can sort of balance the bass side and treble side.

Currently I'm using a 4 stage CMOS distortion with a notch in between stages I got from the ROG EQ.  It consist of 2-220k resistors and a 330p and 6.8n cap.  Also when you stick a notch filter in the FBL of the inverters, it becomes a gyrator.

An easy way to model it, is to use the James circuit in the Duncan Tonestack Calculator.  By zeroing out some of the values, you can get down to the input impedance 2 resistors and 2 caps.  I'll give the values in a following post. 

[WGTP]

Duncan Tonestack Calculator - James Circuit  I used 1k because that is the size of the resistor after the op amp in the Tube Sceamer.  6db cut a 1000hz.  Halving the resistors raises the notch an octave to 2khz.  Doubling the resistors lowers it an octave to 500hz.  The notch filter actually consists of R1, R4 and C2,  C3.  Next change C2 to 52n and C3 to 5n for a 15db notch.  Now tweak till your groove is fully on. 

Bass and Treble - 100% make sure or it won't work right
Zsrc - 1k
R1 - 10k
R2 - 10M
R3 - 1
R4 - 10k
R5 - 1M
R6 - 10M
C1 - .1p
C2 - 26n
C3 - 10n
C4 - .1p

[mac]

http://www.aronnelson.com/gallery/main.php/v/WGTP/BassBoostNotch.jpg.html?g2_imageViewsIndex=1   

It looks like Sekova or Shin-Ei Fuzz tone control.

mac

[Mark Hammer]

Well, although they didn't "invent" it, yes that's pretty much what it is, a passive midrange scoop.

The region of the dip is set by the value of the upper cap, which allows the treble to pass, and the joint action of the cap to ground and fixed resistor that precedes it, on the bottom.  That part forms a lowpass filter.

I suppose another approach is to use something like the Big Muff type tone circuit, which pans and adjusts the balance between the outputs of a simple highpass and simple lowpass filter.  jack Orman has a nice discussion about tailoring one's Big Muff tone control for more "presence", but the forumula can be easily played with to provide a wide, shallow dip that acts in complementary fashion to the mid-peak, and restores (what you feel is) just the right balance of top and bottom end: http://www.muzique.com/lab/tone3.htm  That can be implemented as either a chassis-mount control, or a simple trimpot, or one of those compromise controls like Full-Tone and Way Huge use these days: knobless long-shaft trimmers that are accessible from the outside.

Of course, in the case of the BMP, that circuit eats up enough signal, that EHX felt it imperative to add a gain recovery stage after it, to bring the level up high enough to produce adequate volume.