Tubescreamer tone control - why?!?

[antonis]

Two caps in parallel equals just a bigger cap, so the cutoff point gets lower. And that resistor between one of the caps and ground probably just makes the rolloff even less steep at some frequencies.

Same stands for R9 in series with 4.5V cap..

P.S.
It's Electrosmash analysis (not mine..)

[Radical CJ]

The Maxon820 adopts the tube screamer's tone control but sticks an opamp buffer in the middle of it. This seems to be adding even more components without further functionality .

[Elektrojänis]

It's Electrosmash analysis (not mine..)

I noticed that and did even look it up on their page. It's kind of weird since I have usually thought that Electrosmash is above my level of understanding. Still it seems obvious to me that the filtering there is not second order and Tom's response plot seems to confirm this. Electrosmash plots on the Tube Screamer analysis do not visualize it very well since the frequency axis is linear.

The Maxon820 adopts the tube screamer's tone control but sticks an opamp buffer in the middle of it. This seems to be adding even more components without further functionality .

That will actually make the combined response from the pre tone control filter and the tone control on the bass side second order. And that buffer is inverting... Maybe Maxon engineers thought they need that for some reason.

[antonis]

I noticed that and did even look it up on their page. It's kind of weird since I have usually thought that Electrosmash is above my level of understanding.

I can assure you that this isn't their only flaw..

[ElectricDruid]

I've done some more experiments and I made a mistake before. Either that or LTSpice was playing up. I'm not sure, I can't get it to do it again.

Anyway, this graph is wrong:


It should look like this:


This is better because now we have the middle setting in the middle. Much improved! However, there's still one problem which is shared by all EQs of this design, which is that all the action happens at the extremes. The lines are every 10%, so you can see that virtual all the action is from 0-10% and from 90-100%. In the middle 80% of the travel, not a lot happens. We loiter about between -3dB and +3dB. This is presumably why the Tubescreamer used a weird s-taper pot for the tone control. It compensates for this defect.
In conclusion, I still think a baxandall-style control (treble only, in this case) would be a better choice. The frequency response can be made identical, and it gives a much more even response with a basic linear pot.

[antonis]

I still think a baxandall-style control (treble only, in this case) would be a better choice. The frequency response can be made identical, and it gives a much more even response with a basic linear pot.

Agree but Baxandall-style tone controls action also concetrates at the ends..

I'd suggest a "tilt" control (like the Ambler type below)..
Configuration similar to Baxandall but with the remarkable attribute of pot swapping its ends over as the frequency goes up..



P.S.1
+/- 7.5dB might seem inadequate but they actually are the "double" because of the way the control works. Any dB Bass boost is accompanied by the same amount of Treble cut - and vice versa..

P.S.2
Of course, items values can (and must) be scaled up, especially for a low current capability preceding stage..
(as they are, input impedance for full Treble boost is only 620Ω..)

[ElectricDruid]

Agree but Baxandall-style tone controls action also concetrates at the ends..

Here's the same "every 10%) graph for the bax treble. As you can see it's a *lot* more even:


I call that an improvement.

I'd suggest a "tilt" control (like the Ambler type below)..

Ah, now, the thing is, (1) I'm trying to tweak the Tubescreamer, not completely alter it, so keeping the response similar/identical but removing the need for a strange pot is good. (2) I've already played with Tilt controls and dirt pedals! - https://electricdruid.net/designing-the-hard-bargain-distortion-pedal/

[antonis]

I've already played with Tilt controls and dirt pedals! - https://electricdruid.net/designing-the-hard-bargain-distortion-pedal/

I know that you know that I know what you know, Tom..

[Fancy Lime]

Let me needlessly throw a spanner in the gears here and ask: are you sure you want to do what you are doing? Trying to add a (proper) boost functionality to the tone control on a tube screamer reminds me of a famous quote from a movie I've never seen:



I have difficulties imagining a situation or setup where an actual treble boost after the mid and treble boost of the clipping stage and before the (usually also treble boosting) input stage of an amp would ever be useful. I would much prefer either a Rat filter type of variable frequency low pass or a shelving treble cut like e.g. in the OCD. The second opamp half can the be used as a full range boost stage to push the amp harder if desired. But that's just me and I'm not usually the best at anticipating othe people's needs and wants, so: grain of salt etc.

Cheers,
Andy

[ElectricDruid]

Andy,

I'm not actually *adding* any treble boost, although I can see it might look like that from the graphs above. Believe it or not, those graphs are what the tone control in the Tubescreamer actually does if you remove the brutal 720Hz LPF from in front of it. The graphs are *just* the tone control part, not the LPF as well. Swapping the tone control stage from a graphic-EQ-style to a bax-style just helps the pot response is all.

As far as *changes* go, I'd like to try and open the sound up a bit, maybe get away from that very midrange-forward tubescreamer sound. Or at least make it a bit more optional! I'm thinking that perhaps a switch that changes some key cap values might be the way to go here, so you can choose between a "classic" and a "new" voicing.

Here's the original circuit in the sim, including the lowpass filter at 720Hz (R4/C1):


And here's the response (note the very uneven spacing of the lines, requiring the weird pot to try and corrrect it):


Here's a bax-style replacement, again with the 720Hz lowpass in front of it:


And here's the response - it's basically the same, except now the lines are much more evenly spaced, so no weird pot is required:

[m4268588]

Code Select Expand

Version 4
SHEET 1 800 600
WIRE 384 -368 336 -368
WIRE 400 -368 384 -368
WIRE 496 -368 480 -368
WIRE 512 -368 496 -368
WIRE 336 -352 336 -368
WIRE 496 -352 496 -368
WIRE 48 -336 32 -336
WIRE 144 -336 128 -336
WIRE 240 -336 144 -336
WIRE 288 -336 240 -336
WIRE 240 -320 240 -336
WIRE 288 -288 272 -288
WIRE 144 -272 144 -336
WIRE 336 -256 336 -272
WIRE 240 -224 240 -240
WIRE 240 -224 176 -224
WIRE 240 -208 240 -224
WIRE 176 -176 176 -224
WIRE 240 -112 240 -128
WIRE 272 -112 272 -288
WIRE 272 -112 240 -112
WIRE 288 -112 272 -112
WIRE 384 -112 384 -368
WIRE 384 -112 368 -112
WIRE 32 -96 32 -336
WIRE 240 -96 240 -112
WIRE 272 -96 272 -112
WIRE 176 -80 176 -96
WIRE 32 0 32 -16
WIRE 144 0 144 -192
WIRE 144 0 32 0
WIRE 176 0 176 -16
WIRE 176 0 144 0
WIRE 240 0 240 -16
WIRE 240 0 176 0
WIRE 272 0 272 -16
WIRE 272 0 240 0
WIRE 496 0 496 -288
WIRE 496 0 272 0
WIRE 32 16 32 0
FLAG 32 16 0
FLAG 32 -336 Sig
FLAG 336 -256 0
FLAG 512 -368 OUT
IOPIN 512 -368 Out
SYMBOL signal 32 -112 R0
WINDOW 123 2 104 Left 2
WINDOW 3 2 120 Invisible 2
SYMATTR SpiceLine Rser=0
SYMATTR InstName V1
SYMATTR Value ""
SYMBOL e 336 -368 R0
SYMATTR InstName E1
SYMATTR Value 100k
SYMBOL res 32 -320 R270
WINDOW 0 32 62 VLeft 2
WINDOW 3 0 50 VRight 2
SYMATTR InstName R4
SYMATTR Value 1k
SYMBOL cap 480 -352 R0
SYMATTR InstName C1
SYMATTR Value 220n
SYMBOL res 128 -288 R0
WINDOW 0 6 40 Right 2
WINDOW 3 6 56 Right 2
SYMATTR InstName R1
SYMATTR Value 10k
SYMBOL res 224 -336 R0
WINDOW 0 14 24 Right 2
WINDOW 3 6 40 Right 2
SYMATTR InstName Rvr_1
SYMATTR Value {VR_1*1k}
SYMBOL res 224 -224 R0
WINDOW 0 50 16 Left 2
WINDOW 3 50 32 Left 2
SYMATTR InstName Rvr_3
SYMATTR Value {VR_3*1k}
SYMBOL res 160 -192 R0
SYMATTR InstName R6
SYMATTR Value 220
SYMBOL cap 160 -80 R0
WINDOW 0 18 48 Left 2
WINDOW 3 18 64 Left 2
SYMATTR InstName C6
SYMATTR Value 220n
SYMBOL res 384 -128 R90
WINDOW 0 8 56 VBottom 2
WINDOW 3 24 56 VTop 2
SYMATTR InstName R7
SYMATTR Value 2k2
SYMBOL res 224 -112 R0
WINDOW 0 14 15 Right 2
WINDOW 3 14 32 Right 2
SYMATTR InstName R10
SYMATTR Value 11k
SYMBOL res 256 -112 R0
SYMATTR InstName R11
SYMATTR Value 2k2
SYMBOL res 384 -352 R270
WINDOW 0 32 62 VLeft 2
WINDOW 3 0 50 VRight 2
SYMATTR InstName R12
SYMATTR Value 1k
TEXT 0 56 Left 2 !.AC Oct 10 10Hz 100kHz\n \n.Parama Tone=0.5\n+ VR_1=Tone*(1-1e-6)+1e-6\n+ VR_3=(1-Tone)*(1-1e-6)+1e-6\n.STEP Param Tone 0.0 1.0 0.1
LINE Normal 320 -184 320 -168 2
LINE Normal 320 -168 256 -168 2
LINE Normal 256 -168 264 -172 2
LINE Normal 256 -168 264 -164 2
Do you have any additional requests?

[Elektrojänis]

Then again, if you just remove that 720 Hz lowpass and replace the tone control with the rat filter with part values selected to provide similar range, how different would it be compared to the original?

I quess there would not be that kind of very slight boost around 1 - 2 kHz at the brightest setting, but looks like it is only maybe 1 dB anyway.

[Fancy Lime]

Tom,

I see. I misunderstood and thought you wanted "true boost". I always applaud efforts to improve the feel and response of controls. It's an aspect where many classic pedals are sorely lacking and your approach looks very promising.

I kinda sorta concur with Petri in that the Rat filter is a good option too. But it also has the uneven response problem and needs a rev log pot to work the "normal" way round. I find that the Rat filter's feel is much improved by changing the cap to 6.8n, changing the series resistor to 3.3k and adding a 100k resistor in parallel with the 100k variable resistor (rev) log pot. The total available range may however be excessive for a TS type of thing. A lin 10k pot (without the parallel resistor), 3.3k series and 6.8n cap may well be sufficient and should give a good feel.

Andy

[Elektrojänis]

I kinda sorta concur with Petri in that the Rat filter is a good option too. But it also has the uneven response problem and needs a rev log pot to work the "normal" way round. I find that the Rat filter's feel is much improved by changing the cap to 6.8n, changing the series resistor to 3.3k and adding a 100k resistor in parallel with the 100k variable resistor (rev) log pot. The total available range may however be excessive for a TS type of thing. A lin 10k pot (without the parallel resistor), 3.3k series and 6.8n cap may well be sufficient and should give a good feel.

For a moment, I forgot the taper thing on the rat control. For me regular log taper and wiring it backwards works ok, but tone control that works "backwards" seems to throw some people off.

[marcelomd]

The Klon also uses a Baxandall treble control as the tone. I'd start there.

[Fancy Lime]

I forgot to say two more things: if you are sticking with the Baxandall, I think you can probably get rid of R1 and C2 and instead increase C1 for much the same result. I haven't simmed or breadboarded that specifically, just a hunch but may be worth trying.

Second: do consider adding at least the option for a bass control. It's just too easy to do with a bax already there and it's mad useful in a tube screamer. Trust me, I tried (different topology but same principle) and it is awesome.
https://www.diystompboxes.com/smfforum/index.php?topic=127739.20#msg1245690

Cheers and rock on,
Andy

[ElectricDruid]

1) Ok, I'll give it a try

and

2) Thanks! I will!

T.

[marcelomd]

Hmmm. Tube screamer+ bass and treble bBaxandall == Xotic BB preamp.