Tone stacks only attenuating frequencies, or?

[larstore]

As far as i can see, most amps uses some kind of a passive RC tonestack. In other words, it will remove frequencies from the original signal.
In my mind, an ideal tonestack should therefore let the original signal through if everything is turned to 10 right?

I am asking because i would like to disable the tonestack when tuning the different sections. This way i am working with all my available frequencies.

Am i correct, or are there other factors in the tonestack i am not aware of?

[shooter]

Not a wiz here, but most TS's are RC networks with specific rolloff, or cuts, you can see this with tone-stack software, I used a 1meg and switch to "elevate" or bypass my tone-stack on one build, can't find the schematic to verify the R value but you can search "tone-stack bypass" n find it.  There is years of whys n how's if you really get geeked about TS theory!!

[larstore]

I used a 1meg and switch to "elevate" or bypass my tone-stack on one build, can't find the schematic to verify the R value but you can search "tone-stack bypass" n find it.

I have also disconnected the tonestack from ground, and bypassed it entirely using a jumper, however; my question is:
Is this equal to a TS where everything is turned to 10? (apart from signal loss, which is all frequencies?), or are there other factors i don't know about?

[Willabe]

In my mind, an ideal tonestack should therefore let the original signal through if everything is turned to 10 right?

And what is the original signal? The signal from the guitars PUP's?

No PUP was developed in a vacuum, have to have an amp to run it through to test it and tweak it. Leo developed his PUP's through his amps and vice versa.

Kind of a catch 22 situation?


                        Brad      

[larstore]

And what is the original signal? The signal from the guitars PUP's?
No PUP was developed in a vacuum, have to have an amp to run it through to test it and tweak it. Leo developed his PUP's through his amps and vice versa.

hehe, i was of course thinking of the signal from the previous stage (preamp, CF or whatever)
I am new to this, but my idea was to build/tweak amps without any TS first, and shape the sound (the low frequencies in particular) to whatever "maximum" i want, and then add a suitable TS to limit these to other situations.

LT

[shooter]

There are many ways to "skin a cat", the whole point, hopefully, you learn what works for you, you gain knowledge and understanding, best of all you wind up having fun and an amp tailored by you for you.
Disclaimer....It CAN become addictive!

[SILVERGUN]

A lot of guys use and have used this tonestack calculator for determining tone stack values and their effect on signal loss.

It'll at least give you a visual for what you're trying to picture.

Double click on any value to change the value, and mess around and ask more questions.

http://www.duncanamps.com/tsc/

[AZJimC]

A passive tone stack only attenuates frequency ranges, but, I'm not sure I'd want to tune a circuit without it. For instance, if I tuned a circuit to my pleasing with the tone all on 10, then the only way to go from there is down. I'd prefer to set everything centered, and tune the amp to a pleasing tone, then when I wish I can add a bit of treble or take it away, from a centered baseline. It's not adding treble, but it seems like it is.

On one amp with a TB stack, no mid, I used the calculator to get very close to flat EQ at 1-2 on bass and treble. When I crank them up, the eq becomes mid dipped, but I still have that near flat setting available

[MakerDP]

A passive tone stack only attenuates frequency ranges, but, I'm not sure I'd want to tune a circuit without it. For instance, if I tuned a circuit to my pleasing with the tone all on 10, then the only way to go from there is down. I'd prefer to set everything centered, and tune the amp to a pleasing tone, then when I wish I can add a bit of treble or take it away, from a centered baseline. It's not adding treble, but it seems like it is.

On one amp with a TB stack, no mid, I used the calculator to get very close to flat EQ at 1-2 on bass and treble. When I crank them up, the eq becomes mid dipped, but I still have that near flat setting available

I like this approach.

Another consideration is that when you bypass/lift your tone stack, it comes with a gain-boost because nothing is being attenuated anymore. So, what you are tweaking is not necessarily what you are going to experience when you put the TS back in play gain-wise.

[alerich]

I have also disconnected the tonestack from ground, and bypassed it entirely using a jumper, however; my question is:
Is this equal to a TS where everything is turned to 10? (apart from signal loss, which is all frequencies?), or are there other factors i don't know about?

No, those examples aren't the same. If you rig up a switch to lift the common tone stack ground at the mid pot and then dime the controls and A/B it with the switch open and closed you can clearly hear the difference. The pots in the tone stack create a load in the circuit (since they are essentially a resistor to ground) that bleeds some signal off to ground no matter where the pot is set. Opening that switch to lift the ground and then bypassing the entire stack with a jumper should produce an audible difference, too since the tone stack components are still in the signal path even with the ground connection lifted.

There is a lot of signal (gain) shunted to ground in the average tone stack and for good reason. If you think about 12AX7 gain stages they are capable of a gain factor of 100. Put a couple in series (100 x 100 x 100) and things add up pretty fast. this doesn't include the other various voltage dividers and R/C tone shaping networks found in the design. An old timer once told me "Building a tube amplifier is easy. Building a tube amplifier that sounds good is another matter entirely".

[Ritchie200]

An old timer once told me "Building a tube amplifier is easy. Building a tube amplifier that sounds good is another matter entirely".

Amen brother.  If I am not mistaken, if we were to analyze all of PRR's posts, most have included the comment, "use an existing design/layout/values and tweak from there as it has been done to death over the years" - or something quite similar.  After all these years, I dont think there are any new mousetraps.  We are only slightly off a well travelled path with designs from time to time only due to personal preference or fad.


It has been interesting to see Tubenit's tone journey over the years as I don't think anyone here has embarked on such a wide and varied practical adventure.  There is much to be learned and there is plenty of written material out there now that will save you hours of experimenting and you will learn the theory at the same time.


Jim

[jjasilli]

+1 to Reply #9 by alerich


+1 to Ritchie200's compliment to Tubenit, Reply #10.


BTW: if jumpering a tonestack, please be aware that the tube plate feeding the tonestack may not have a blocking (coupling cap) following it, because it's relying on the cap(s) in the tonestack to block hi voltage B+ from the next tube's input grid.  The solution is to add a large cap in series with the jumper path, say .22uF, which will block DC yet still pass the full audio frequency spectrum of signal.

[HotBluePlates]

... In my mind, an ideal tonestack should therefore let the original signal through if everything is turned to 10 right?

In that case, don't use a tone stack from any Fender, Marshal or Vox amp. Use a James/Baxandall circuit, and set the controls to halfway, which yields un-boosted/cut performance. Really, it is a cut, but to all frequencies equally.

A lot of guys use and have used this tonestack calculator for determining tone stack values and their effect on signal loss.

It'll at least give you a visual for what you're trying to picture.

+1. Your mental image of what's going on prbably deviates from reality.

Play a black/silver-face Fender, then play a tweed Champ or a tweed Deluxe with the tone control at half-way. If the later Fender is your tone benchmark, you'll call the tweed amp midrange-heavy. But when you look at the Fender in the tonestack calculator, you'll see you were wrong; the later amps have large midrange cuts.

I am asking because i would like to disable the tonestack when tuning the different sections. This way i am working with all my available frequencies.

I recommend tweaking the amp in the form in which you'll use it. If the end product will have tone controls, make sure they're in place before modifying anything else. Make sure you're listening to the amp through whatever speaker you'll ultimately use as well.

You see, you're always working with all available frequencies. The speaker will only go so high and so low, and won't output equal power for every frequency in between. The output transformer may impose additional high/low frequency limits. And of course, coupling and bypass caps, tube capacitance and the tone circuit will all have some degree of input. So to some extent, all frequencies are there, but they are all at different relative levels.

All common guitar amp tone circuits cut gain to some degree (there are active tone circuits, but they are very much the exception rather than the rule). The circuit cut all frequencies (some more than others), and then allow you to "cut less" at some frequency range to give an appearance of a boost. Guitar amp tone circuits are also pretty "low-Q;" each control tends to boost/cut a pretty wide band of frequencies. You'll need something like a good outboard EQ device or a pedal such as the Boss GE-7 or Empress ParaEq if you want more precision.

[jjasilli]

+1 to Willabe Reply #3.  For electric guitar the amp is part of the tone production process.  That's different from hi-fi, or acoustic guitar, where the amp is (usually) only for RE-production of tone.  The el guitar signal is not sacred or "fully formed"; it needs the amp & speaker to come to life.  The guitar amp, tonestack & speaker do not "corrupt" the signal.  They are essential to its original creation in actual soundwaves in the air that people can hear.


Also you state, Post #1: "I am asking because i would like to disable the tonestack when tuning the different sections. This way i am working with all my available frequencies. . . Am i correct, or are there other factors in the tonestack i am not aware of?"

The tonestack must be located at some specific location in the preamp.  There may be gain stages before, and after it.  Exactly where the tonestack is located among the gain stages has a major effect on tone, especially for overdrive tone.  It is not possible to tune sections of a guitar amp by purposely disregarding the important design feature of tonestack location.  Recommended reading: Dave Funk's "Tube Amp Workbook, Vol 1", Chapter 14 which covers this topic.

It is also not possible to tune different sections of the tonestack, by somehow rendering the other sections inoperative.  The different tonestack controls (Bass, Mid & Treb) are interactive in complex ways (except maybe for parametric or Baxandall/James tone control circuits).  See:  the AX84 Amp & Tone Theory Doc


Can anyone find the AX84 Amp & Tone Theory Doc on the Web?  If not I have it as PDF File & can post it or email it.

[MakerDP]

Is this the one you mean?

http://ax84.com/static/p1x/p1-ex-theory.pdf

[jjasilli]

Thanks, MakerDP. I have Rev B.  You seem to have an updated version.  I think it has basically the same tonestack info.

[MakerDP]

Stephen Keller is currently in the process of writing a major revision to it as well. It will cover not just the P1eX but all of its derivatives that have been developed over the years as well.

I learned a LOT from that document. I really like how it relates directly to an actual working schematic to explain how things work in a tube amp.

Another good resource for tone stacks (and everything else tube preamps) is Merlin Blencowe's book. He even shows an "adjustable tone stack defeat" control circuit that  looks interesting.

[PRR]

> let the original signal through if everything is turned to 10 right?

No.

Look at the Bass/Treble tone control found on older hi-fi.

The 700Hz-1KHz band stays the same level at all knob settings, only the ends of the band move up and down.

In passive systems if you want "boost", what you do is start by cutting everything, cut/un-cut as desired, and then find make-up gain elsewhere so un-cut sounds like boost.

So the bass-treble tone control cuts everything down 10:1. 50hz and 15KHz may then be dialed up or down, but 1KHz stays near 10:1.

If B and T are turned to 10, you get big boom, big sizzle, and weak in between.

The Fender stack is quite different. It only gives the curves that are "useful" for guitar. And the magic is in the extreme interaction of B and T and M. You can get a mid-dip MUCH sharper than any non-interactive network. Which means you can not tune one section at a time.

Don't beat head on wall. Find a happy plan and steal it.

[Jack_Hester]

What about an active tone control like the paraphase found on page 665 of the Radiotron Handbook.  I was looking in the 4th edition.  I don't know if previous editions show it.  I have a 3rd, but didn't look.

Jack

[MakerDP]

Don't beat head on wall. Find a happy plan and steal it.

I do not agree with this statement.

While it MAY be what he should do, we need to remember that innovation comes from experimentation. Sometimes it comes from the MISTAKES we make in our experimentation.

If the builder's goal is to get an "industry standard sound" built as quickly as possible, then YES, by all means, "Don't beat head on wall. Find a happy plan and steal it."

BUT if the builder is driven by curiosity and enjoys the process of exploring the question "what if I..." then he should try whatever he wants. He may end up with something that sounds like garbage but then again he may become the next Dumble and 25 years from now his amps will be selling for $25,000 and their super-secret mojo will be the discussion of Internet forums internationally. Or, he may just end up with something unique that is pleasing to his own ear because of his "outside the box" thinking, which is just as cool.

[shooter]

+1 reply

[jjasilli]

"If the builder's goal is to get an "industry standard sound" built as quickly as possible, then YES, by all means, "Don't beat head on wall. Find a happy plan and steal it."

BUT if the builder is driven by curiosity and enjoys the process of exploring the question "what if I..." then he should try whatever he wants. . . Or, he may just end up with something unique that is pleasing to his own ear because of his "outside the box" thinking, which is just as cool."



This is a false dichotomy.  Plagiarizing and adapting known amp circuits need not result in standard tone.  E.g., Marshal & Mesa Boogie largely copied Fender, but did not produce Fender tone.


Amp design is the science / technology of research & development.  Science is not magic or attitude (even though the tone that results may seem magical to the listener).  Shooting in the dark while affecting a cool attitude is unlikely to produce a good result. 


Most amp builders lack the technical foundation to engage in original design.  Even professional designers work off-of known, successful, pre-existing designs.

[terminalgs]

Am i correct, or are there other factors in the tonestack i am not aware of?

There is phase shift that comes from C's and R's in parallel or in series networks, or from even from a single coupling cap between the plate and the TS (like you find in many Baxendalls). With a single coupling cap, you get freq. dependent phase shift occurring before the frequencies attempt to find their way though the TS. That was the chief design purpose of the slope resistor, to eliminate that extra unnecessary phase shift.

I am asking because i would like to disable the tonestack when tuning the different sections. This way i am working with all my available frequencies.

Then you need a piece of straight wire where the TS is.  and you need a big coupling cap (.1uf) between the plate and your 1M volume control so the low frequency roll off is well below what your guitar will produce.

when you bypass/lift your tone stack, it comes with a gain-boost because nothing is being attenuated anymore.

Only if you completely by-pass it. lifting the ground on the TS doesn't disable the freq. dependent impedances nor the freq. dependent phase shift presented by the TS's arrangement of caps and resistors in parallel and in series.  everything will be louder, but different frequencies will still see the different "paths" as having different impedances.  Those impedances, and the impedance of the volume pot to ground (1M) will set up different dividers for different frequencies.

[MakerDP]

This is a false dichotomy.  Plagiarizing and adapting known amp circuits need not result in standard tone.  E.g., Marshal & Mesa Boogie largely copied Fender, but did not produce Fender tone.

They did not make EXACT copies though. Obviously, there are parameters that "box you in." A tube can only do so much and you have to do certain specific things to it to make it do anything at all. But, Jim did not EXACTLY copy Leo... he changed the tubes... he then discovered he could tweak this and that. He worked with what he had on his workbench, with what he could source locally without paying a ton of money for imported stuff - like different tubes and speakers. And I think most importantly... he didn't have people on the Internet to tell him he MUST do it like this or that or it will sound like sh!t.

Amp design is the science / technology of research & development.  Science is not magic or attitude (even though the tone that results may seem magical to the listener).  Shooting in the dark while affecting a cool attitude is unlikely to produce a good result. 

Most amp builders lack the technical foundation to engage in original design.  Even professional designers work off-of known, successful, pre-existing designs.

Which is probably why there are so few truly distinctive "tones" out there... too many people taking the safe road. Pretty much everything in the mainstream sounds either like a Fender, Marshall or VOX.

Go back in time, tell that to Edison and convince him to give up on his science / technology of research & development. Now, imagine  where we'd be today. You do realize he was "shooting in the dark" when he tried several HUNDRED designs of the lightbulb over several years of R&D, right?

SOMEBODY had to come up with this stuff on their own. And now somebody in the future will step-up and take it to the next level... I hope...

[jjasilli]

Go back in time, tell that to Edison and convince him to give up on his science / technology
I am not suggesting that anyone give up on science; rather that it be employed instead of flailing around.


SOMEBODY had to come up with this stuff on their own. And now somebody in the future will step-up and take it to the next level
All that, including the future, has already happened, in the past.  It is unlikely that a better incandescent lightbulb will be built.  If it is, it will likely be done by professional researches.  No longer by a diy guy tinkering in his basement.



Tube amp technology is a 1920's or '30's thing that basically peaked in the 1940's or '50's, and reached fulfillment by the early 1960's.  Since then, anything that can be done by mere tinkering has already been done.  If you don't study the history at best you'll be re-inventing the wheel while selling short the great technical & business pioneers who developed the wonderful amps that we take for granted today.

[HotBluePlates]

Don't beat head on wall. Find a happy plan and steal it.

I do not agree with this statement.

Then modify PRR's advice to: "Find a happy plan, steal it, then fiddle it until it sounds best to you."

Am i correct, or are there other factors in the tonestack i am not aware of?

There is phase shift that comes from C's and R's in parallel or in series networks, or from even from a single coupling cap between the plate and the TS ... That was the chief design purpose of the slope resistor, to eliminate that extra unnecessary phase shift.

This is incorrect; the human ear is unable to detect absolute phase in a mono sound source, and so no amount of phase shift of a mono guitar signal is audible. You can only hear phase shift when you also have a non-shifted (or differently-shifted) signal to hear at the same time.

The slope resistor has nothing to do with phase shift, per se. What is really going on is that a Fender tone stack is basically taking a bridged-T filter, keeping the left R fixed (which is now your slope resistor), and making the right R variable (S in the figure below; your Treble control in a Fender stack). The upper cap (D in the figure) is your 250pF treble cap, and the lower cap (C in the figure) is the composite value of your mid and bass caps. The overall value of the lower cap, in Fender's incarnation, is controlled by the setting of the Bass pot (wired as a rheostat between the 2 caps). A resistance between this lower cap and ground makes the natural mid-notch of the circuit less-steep if the resistance to ground is increased (Which is your Mid control).



Now there are phase shifts in the circuit, but you can know everything about how the circuit works and never know what any of the phase shifts are.

I am asking because i would like to disable the tonestack when tuning the different sections. ....

Then you need a piece of straight wire where the TS is.

when you bypass/lift your tone stack, it comes with a gain-boost because nothing is being attenuated anymore.

Only if you completely by-pass it. lifting the ground on the TS doesn't disable the freq. dependent impedances ...

We don't know what tone circuit he will use. If he's using the common 60's or later Fender, or Marshall tone stack, breaking the connection from mid-pot to ground will effectively take the tone circuit out of operation.

A bridged-T filter (and the Fender/Marshall tonestack is just a bridged-T with variable components) has a response as shown below (solid green line). Overall, there is a low-pass section, where there is maximum response for bass and then a cutoff moving towards infinity. There is also a high-pass section, which passes some high frequency with no cut, but as frequency moves lower there is a cutoff extending to infinity. The component values are sized relative to each other such that the portion of roll-off of each sections don't overlap at/near zero roll-off; the net effect is to cut out some amount of midrange.



If you increase resistance to ground, the midrange cut becomes less-deep. If you increase resistance to infinity ohms (disconnect the mid-pot from ground), there is no mid-cut. Neither the low-pass or high-pass sections can work because there's no ground reference to cut against, and the top-most cap acts just like a coupling cap.

Don't take my word (though I think I'm steering you straight); try disconnecting the mid pot for yourself. See if the mids and overall gain are boosted, and also try to change the tone of the amp with the tone controls.

[jjasilli]

HotBluePlates:  If you increase resistance to ground, the midrange cut becomes less-deep. If you increase resistance to infinity ohms (disconnect the mid-pot from ground), there is no mid-cut. Neither the low-pass or high-pass sections can work because there's no ground reference to cut against, and the top-most cap acts just like a coupling cap.Don't take my word (though I think I'm steering you straight); try disconnecting the mid pot for yourself. See if the mids and overall gain are boosted, and also try to change the tone of the amp with the tone controls.
This is confirmed by Merlin (See MakerDP Reply #16) and the Duncan Tonestack Calculator -- to the surprise of me & Willabe who assumed a ground-lifted FMV tonestack would still attenuate signal and filter frequencies.  Merlin shows a fairly typical FMV tonestack but with a 470K mid pot  (basically, a RAW control to my mind).  At full mid pot rotation, with the other controls at mid position, signal is flat with 0dB loss.  The Duncan Calculator is nearly the same: flat signal with 3 - 4dB loss. 

[HotBluePlates]

To be fair to Terminalgs, yes, the treble cap and pot still have a little effect, but it is so small in comparison with the original range of adjustment as to be negligible.

See the plot below from Duncan's Tone Stack Calculator. I took the stock Fender tone stack and changed the Mid pot to a 1MΩ control. The curve with the midrange notch is stock Bass and Treble controls at half-up with the Mid control at 0. The flat-ish line at the top of the plot has all the same settings, but the 1MΩ Mid pot is full-up.

With the Mid full-up (1MΩ to ground), varying the Bass pot does nothing (and it shouldn't, because the Mid pot isolates it from ground, and therefore any effect). The Varying the Treble control to full-off puts a ruler-flat line across all frequencies with a loss of 2dB. Turning the Treble control full-up increases bass cut to -4dB, with a smooth transition to very little Treble loss at -1dB. The Tone Stack Calculator already assumes the 1MΩ Volume pot will follow the tone circuit, so these results match what will happen in the amp.

Increasing the Mid pot again to 10MΩ moves all results up 1dB. The only change with the full 10MΩ Mid pot resistance is that at maximum treble, the Bass is 2dB below the Treble level.

So yes, there are couple-dB difference between simply lifting the Mid pot ground and a pure bypass where nothing is in the circuit. But in the sim and audibly, the changes are extremely slight. Compare that with the resulting boosts: Mids are up 31dB, Treble is up 13dB and Bass is up almost 10dB.

[terminalgs]

This is incorrect; the human ear is unable to detect absolute phase in a mono sound source, and so no amount of phase shift of a mono guitar signal is audible. You can only hear phase shift when you also have a non-shifted (or differently-shifted) signal to hear at the same time.

certainly true with a single frequency audio signal, but being that guitars can produce multiple frequencies simulataneously, Is it possible that the phase shift for different frequencies might be detectable?  a 250pf||250kohm pair would provide 100Hz, 1000Hz, and 10KHz with phase shifts of -2deg, -21deg, -75deg respectively. Perhaps even -75deg still isn't detectable to the human ear.  I know 250pf||250Kohm is simple example, perhaps the network of R's and C's in the fender TS, while more complex that C||R, might not present as much phase shift as C||R.  I'm not an expert, so if I'm wrong, I'm happy to admit it! (sound the gong, or the price-is-right loser horn...)

The slope resistor has nothing to do with phase shift, per se. What is really going on is that a Fender tone stack is basically taking a bridged-T filter, keeping the left R fixed (which is now your slope resistor), and making the right R variable (S in the figure below; your Treble control in a Fender stack).

I gathered that role for the slope resistor from reading US3147447 A "Tone Control Circuit" patent by Clarence L. Fender:

"A further object is to provide a tone control circuit in which phase-shift effects are much less pronounced and damaging than in prior-art tone control circuits. A further object is to provide a tone control circuit in which the need for a coupling capacitor, with attendant phase-shift effects, is eliminated."




thanks for the clarification and the illustration of the bridged-T filter. I had never noticed the fender TS is a variant of that bridged-t filter.
 

[terminalgs]

To be fair to Terminalgs, yes, the treble cap and pot still have a little effect, but it is so small in comparison with the original range of adjustment as to be negligible.

This is how I viewed the fender TS with the 6800R ground end lifted AND with if the volume pot is @ "10" and  the treble pot is swept down all the way:


the bass pot, the .047 and the 6800 become non-players.  high frequencies that see the CR 250pf/1M as low impedance would pass unattenuated (> 600Hz ?) and lower frequencies would go the .1uf route and find divider with R1=250K, and R2=1M. Is this accurate?





compared to the shaping provided with the 6800R grounded, this is (as you pointed out) nearly nothing.


(thanks for the all the great dialogue , BTW!!)

[jjasilli]

http://ecee.colorado.edu/~mathys/ecen1400/pdf/FrequencyResponseOfRCcircuitsHandout.pdf
If I'm reading the above webpage correctly, phase shift does in fact vary with frequency.  However, in a mono system I think Hotblue is correct that we could not detect this with the human ear. 


Also, for el guitar amp tone production, such phase shift within the signal spectrum may appear as a feature of the total process of tone production,  rather than a bug to be avoided or cured.  Just the opposite for hi-fi sound re-production.  Here phase shift messes with the integrity of the musical program; and way wreak havoc with the soundwaves in the air produced by multiple drivers, woofer, mid & tweeter of hi-fi speakers.

[HotBluePlates]

This is incorrect; the human ear is unable to detect absolute phase in a mono sound source, and so no amount of phase shift of a mono guitar signal is audible. You can only hear phase shift when you also have a non-shifted (or differently-shifted) signal to hear at the same time.

certainly true with a single frequency audio signal, but being that guitars can produce multiple frequencies simulataneously, Is it possible that the phase shift for different frequencies might be detectable?  ...

I did a poor job of stating what I meant. 

In a mono sound source, static phase shifts are undetectable. See A Question of Phase. RDH4 (on page 618; downloadable from the Library of Information, linked at the bottom of this page) is largely non-specific on the possible effects of "phase distortion" of mono signals.

Each frequency is delayed a different amount, but the delay of any single frequency is constant because the C and R are constant. If C or R were varying, you might hear the changing phase, but you'd always hear the effect of phase shift if you could hear the delayed signal and the non-delayed signal simultaneously (which is the basis of phase, chorus and vibrato effects; you hear the changing delay layered over a bit of the original signal to hear a composite phased/chorus'd/vibrato'd signal).

I gathered that role for the slope resistor from reading US3147447 A "Tone Control Circuit" patent by Clarence L. Fender:

"A further object is to provide a tone control circuit in which phase-shift effects are much less pronounced and damaging than in prior-art tone control circuits. A further object is to provide a tone control circuit in which the need for a coupling capacitor, with attendant phase-shift effects, is eliminated."

Well, that was Leo's claim. You can claim anything in a  patent... And he was trying to patent an embodiment of a tone control circuit, and probably needed a way to distinguish his circuit from all the other tone controls already in existence (and probably patented).

Bottom-line, any frequency-dependent element in a circuit has different amounts of phase shift/delay for different frequencies. Phase shift can be expressed in degrees of phase-shift, or in terms of time delay (a 1kHz signal has a complete cycle every 1 milli-second, and a 45-degree phase shift is 1/8th of a cycle, or 0.125mS, or 125nS of delay).

If this is critical to you, you have to do away with all coupling caps. And maybe filter caps... And of course cathode bypass caps. And transformers (which are made of coils with inductance)... There are a very few audio amps which use direct-coupling throughout to eliminate coupling caps. To really get rid of all reactive components, you'd have to use only a battery power supply and have a circuit entirely composed of resistors and tubes... But then the tubes have some inter-electrode capacitance, and all wiring has at least minute inductance and capacitance between wire and to ground, so... In the end, it really requires perspective on whether or not the issue is relevant.

It's a bigger issue inside a feedback loop: Negative feedback is 180-degrees out of phase with the original signal, but reactive components cause phase shifting. You can express the degrees of phase shift in terms of dB of feedback, because phase angle is proportional to the amount of cut of an R-C network in dB. So having reactive components inside the feedback loop, enough to turn absolute phase of the feedback from 180-degrees to 360-degrees (or 180 degrees of extra shift) result in positive feedback. Approaching 180 degrees of added shift, the loop will ring. So really, designers care about how much phase shift is inside a loop very much more than overall phase shifting when there is no feedback loop.

[terminalgs]

thanks HotBluePlates, I appreciate the continued explanation. all excellent.

[HotBluePlates]

This is how I viewed the fender TS with the 6800R ground end lifted AND with if the volume pot is @ "10" and  the treble pot is swept down all the way:


the bass pot, the .047 and the 6800 become non-players.  high frequencies that see the CR 250pf/1M as low impedance would pass unattenuated (> 600Hz ?) and lower frequencies would go the .1uf route and find divider with R1=250K, and R2=1M. Is this accurate?

Swap the positions of the 250pF cap and the series 100kΩ & 0.1uF cap so that they match a schematic. Now it should be easier to follow what happens...

If the Treble pot is swept all the way up, then there is direct connection from the 250pF cap to the Volume pot (as shown in the diagram above). The Treble pot is also a voltage divider between 2 paths, so its position matters. The Bass pot is a rheostat between the ends of a 0.1uF cap and a 0.047uF cap which are not connected to the 100kΩ slope resistor. Looking carefully at the layout, you'll see the 2 caps have 0Ω between them at the minimum-Bass setting. So your drawing reflects a state when the Bass pot is set to a minimum, and the cap should be 0.147uF total (both caps in parallel). Both caps and the Bass pot are still in play...

The connection to the tube grid is immaterial, because the Volume pot will just divide the voltage down from the maximum Volume setting; either way, sweeping the Volume pot doesn't change the resistance-to-ground seen by the tone circuit.

Check out PRR's Reactance Reminder thread for information on how to calculate cap reactance for various frequencies. We'll use that information to replace the caps with their equivalent resistance (equal to their reactance) at 2 representative frequencies: 80Hz for the guitar's low E and 5kHz for harmonics/overtones of the guitar's treble range (it's very likely your speaker will have audible falling response above 6kHz).

80Hz
250pF cap has a reactance of 159/(0.08kHz * 0.00025uF) = ~8MΩ
0.147uF cap has a reactance of 159/(0.08kHz * 0.147uF) = ~292Ω

5kHz
250pF cap has a reactance of 159/(5kHz * 0.00025uF) = ~0Ω
0.147uF cap has a reactance of 159/(5kHz * 0.147uF) = ~0Ω

Read the thread and do the math to verify for yourself. Now replace each cap with a "resistor" for the specific case for each frequency.

At 80Hz, the previous stage's plate sees a path through the 250pF cap of 8MΩ to the top of the 1MΩ volume pot. There is a 2nd parallel path through the 100kΩ in series with ~292Ω, for a total of 1,000,292Ω. The total parallel resistance of the 2 paths will be 8MΩ || 1.000292MΩ = ~890kΩ. This forms a divider with the 1MΩ Volume pot to cut 80Hz to 1MΩ/(1MΩ+890kΩ) = ~53% of the 80Hz output of the previous stage plate. This is 20 log 0.53 = ~-5.5dB.

At 5kHz, the previous stage's plate sees a path through the 250pF cap of 0Ω to the top of the 1MΩ volume pot. The 2nd parallel path through the 100kΩ has substantially 0Ω in series, for a total of 100kΩ. The total parallel resistance of the 2 paths will be 0Ω because the 250pF cap's low reactance shunts the path through the 100kΩ resistor. There is essentially 0dB of loss through the tone network at 5kHz.

~5-6dB of difference between the 2 paths is sensibly close to what the plots showed for Mid pot disconnected from ground, and Treble control at maximum. I had the Tone Stack Calculator set for Bass halfway up, which accounts for the ~1dB less cut.
Now do the same thing for the Treble control at minimum and Bass at minimum. The previous stage plate sees a path through the 250pF cap, then 250kΩ resistance from the Treble pot, then the top of the Volume control. The other parallel path is 100kΩ in series with 0.147uF then the top of the Volume pot.

At 5kHz, both caps are short-circuits, so there is 250kΩ in parallel with 100kΩ (~71kΩ), then the top of the Volume pot for almost zero loss. At 80Hz, there is a path of 8.25MΩ through the 250pF cap in parallel with 0.100,292MΩ through the 0.147uF composite cap. Total resistance looks like ~100kΩ. Previous stage plate output is cut to ~91% its original value, or less than 1dB. So overall frequency repsonse is substantially flat; this again jives with the Tone Stack Calculator result.
You can break down any tone circuit into an equivalent circuit in this manner, though I'd recommend picking extreme settings of the tone controls in some cases to simplify the equivalent circuit you're working with.

[terminalgs]

Drawing the TS flip-flopped was probably a poor choice. 

The Bass pot is a rheostat between the ends of a 0.1uF cap and a 0.047uF cap which are not connected to the 100kΩ slope resistor.

I'm confused by this, the .1 and .047 are connected to the plate via the slope. 


thanks for the explanation, it is very useful.


And double-Thanks for the pointer to the reactance thread, I missed it entirely.

[HotBluePlates]

The Bass pot is a rheostat between the ends of a 0.1uF cap and a 0.047uF cap which are not connected to the 100kΩ slope resistor.

I'm confused by this, the .1 and .047 are connected to the plate via the slope.

Probably poorly-written on my part. Yes, each cap has one end connected to the slope resistor. The other end (which is not connected to the slope resistor) connects to either end of the Bass pot; one extreme setting of the Bass pot places 1MΩ between that end of each cap, while the other extreme setting of the pot connects the caps together so they are in parallel.

[jjasilli]

Go back in time, tell that to Edison and convince him to give up on his science / technology
I am not suggesting that anyone give up on science; rather that it be employed instead of flailing around.

SOMEBODY had to come up with this stuff on their own. And now somebody in the future will step-up and take it to the next level
All that, including the future, has already happened, in the past.  It is unlikely that a better incandescent lightbulb will be built.  If it is, it will likely be done by professional researches.  No longer by a diy guy tinkering in his basement.

Tube amp technology is a 1920's or '30's thing that basically peaked in the 1940's or '50's, and reached fulfillment by the early 1960's.  Since then, anything that can be done by mere tinkering has already been done.  If you don't study the history at best you'll be re-inventing the wheel while selling short the great technical & business pioneers who developed the wonderful amps that we take for granted today.

OTOH:  http://el34world.com/Forum/index.php?topic=18119.0

[Ritchie200]

Yeah all that effort to get to the same result!!  Another that comes to mind is when I was a kid the old Estes model rockets catalog had a page devoted to the kids who insisted on stuffing match heads into an empty CO2 cartridge for a home-made rocket engine.  Every catalog had a picture of a different kid who lost a finger/hand/eye, etc., etc when it blew up - and not by lighting it....  I'm sure they all thought they were breaking "new" ground.

As you so very eloquently described, we are at our best not ignoring the basic groundwork - but building on it.

JMHO
Jim

[PRR]

> reading US3147447 A "Tone Control Circuit" patent by Clarence L. Fender:
"A further object is to provide a tone control circuit in which phase-shift effects are much less pronounced and damaging than in prior-art tone control circuits. A further object is to provide a tone control circuit in which the need for a coupling capacitor, with attendant phase-shift effects, is eliminated."

This is not-so-true.

A good patent examiner would have asked for clarification or computations. However the in-box is piled high, and it is 'only' a guitar-amp patent. The topology does not seem to be the same as any known prior art. "GRANTED" and move to some more important patent, maybe a nice moon-rocket or H-bomb idea (this was 1961).


In most simple R-L-C networks, phase is absolutely related to the slope of the amplitude response. If the response is flat, phase is zero. If the response tilts, phase is shifted, in that area, in proportion to the amount of tilt.

Phase shift is perhaps the least audible thing you can do to audio. In guitar music production you do not even know what the original phase is (in re-production you can compare the same signal two paths). Phase is often massively tweaked (by non-simple networks) in loudness-optimization processors for radio stations without phase-related ill-effect. Leo was just dressing-up his Claims.


> fender TS is a variant of that bridged-t filter.

The more common Fender Bass/Mid network is another bridge, though not a twin or bridge Tee. That is why it can give a massive dip.

> try disconnecting the mid pot

This should be obvious. This network feeds a naked grid. The naked grid is nearly an Infinite resistance. The lifted-bottom network is just a series element. Any amount of series-crap to an infinite resistance is zero effect. With non-infinite real-world values, the 270pFd treb cap is low compared to the 100pFd grid capacitance, and the 500K pot resistance is very-low compared to the >100Meg grid resistance.

[PRR]

I sit by my advice. Steal plans. Crawl before you climb Everest.

If you aspire to novel things, study what you steal and try variations.

Ignoring the work of the past dooms you to repeat most past failures and sucesses.

When you steal and study enough stuff, and have it all digested into your brain, one day you will reach-in for a common solution and pull-out something that actually has not been done before.

Edison *did* review the very-long prior art of the electric-light idea.

Edison perspired, was not inspired. He went though thousands of bad ideas before finding a few that didn't suck. I suspect Leo also spent whole days with an iron and a box of caps and bland or foul tones. We rarely find such determined searches today. (Except when we can trick a computer into doing the perspiring... bio-tech affinity screenings.)

Edison's lights were NOT that good--- all we use today (well, until 1999) were developed by German researchers and quickly replaced Edison's lamps.

(Edison did "invent" the electric bill business, albeit closely modeled on the Gas Lamp Company business. Edison was the first to collect payments from large numbers of electric users for lighting.)

His path to the phonograph and movies was more direct. He made one phono-sketch, he bought one roll of Kodak film. The cylinder phono didn't last real long, but his movie format was mainstream most of a century.

[jojokeo]

feeds a naked grid. The naked grid is nearly an Infinite resistance. The lifted-bottom network is just a series element. Any amount of series-crap to an infinite resistance is zero effect.

Okay naked grids and lifted bottoms just got my attention...
Aside from all the great reading I'm sure many have experimented plenty with tone stacks, tone stack values, and lifting them up like a plastic surgeon on Meg Ryan's (or any other Hollywood starlet's face or "golden globes"). Adjusting the tone controls on a tone stack that's been lifted does NOTHING to the sound, tone, volume, etc.. etc... end of story. Rather than hypothesize just do it and find out for yourselves if wondering?

On bypassing the tone stack OR simply not having one just remember two things, you need a coupling cap and a grid leak resistor. The grid leak affects the gain of the previous stage just as a tone stack does by it being in parallel with the preceding tube's plate.

A note of caution on the coupling cap: since the increase in gain and fullness of signal is now being experienced, a large cap is NOT wanted in most cases. That is with any moderate to high volume of said amp being used. You WILL experience flubby/farty bass and unpleasing "Mississippi Mud-like" distortion, especially if using the neck pickup. Start with a .002uF and work from there.
I think it was Alerich? that said earlier a good plan is to have your amp and/or design already in a good place along with speaker and/or guitar first and then set the controls in the middle and adjust from there? This is what I do and tweak for the correct ranges for each control that takes you to your happy "sweet spot" place. When doing this you want to test it with other gain controls and/or switches being utilized too.
 
Lastly, I totally agree on tone stack placement. This is a key to getting something sounding like something of a blues amp developing a bit a sag even vs something more of a higher rock tight gain feel of a Marshall - all else being equal.
 

[poirson99]

Hi, will this old topic wake up?

I am attaching a file of my PPEL84 guitar project.

Why ?

Because there is a tone stack and I would like to bypass it as mentioned above ! SW3 and C20'
But I don't like the R25 1M gate leak resistor. It is in parallel with the 3 potentiometers P2/3/4.

What will happen ?

Look !

[HotBluePlates]

... there is a tone stack and I would like to bypass it ...
But I don't like the R25 1M gate leak resistor. It is in parallel with the 3 potentiometers P2/3/4.

Remove R25.  It is not needed.

Now break the connection from P4 to Ground.  Add a switch between P4 and Ground.

   - Close the Switch:  tonestack is active
   - Open the Switch: tonestack is bypassed

[jojokeo]

Here's the way I see it done better simply switching in/out the 4n7 cap w/ a 1M "anti-pop" resistor. This way you'll get a much fuller signal as if it was a complete bypass. But you won't get as much of a jump/increase in the signal being much louder. Also, you'll still have a little Tone control - Control... A complete bypass of the signal can often require a voltage divider along with it so that the signal jump/increase isn't so great between the two - especially helpful and needed if the switch was a footswitch control... Another thing w/ a complete bypass - the bass and low mid frequencies jump along too which is another undesirable outcome at mid to high Vol levels... Things to consider...

[poirson99]

Awesome ! thank you for the feedback.
I had read discussions about a switch under P4 or also the possibility of increasing the value of P4 to 100k or even 250k. So the potentiometer would block the path to the ground.
But this solution forces the signal to find its way through the tone stack whereas I would prefer a bypass.

Jojokeo's solution is hybrid compared to a pure bypass but avoids the problem linked to the 10dB absorbed by the tone stack.
I like this track.

[HotBluePlates]

I had read discussions about a switch under P4 or also the possibility of increasing the value of P4 to 100k or even 250k. So the potentiometer would block the path to the ground.
But this solution forces the signal to find its way through the tone stack whereas I would prefer a bypass.

Disconnecting pot P4 from ground (or adding a high resistance of 220k-1MΩ) is bypassing the tone stack.

I understand it appears that "signal goes through tone stack components."  However, with no connection to ground, the tone stack components cannot "reduce signal level" so they are effectively bypassed.  And the tone stack components simply "reduces signal level" by differing amounts at different frequencies.