tone stack question

[jcm-jmp]

hey guys

this tone stack is confusing me. it looks like the plate of the 12ax7 feeding the stack and the plate of the 12ax7 right after the stack are connected togather through the tone stack network. can you do this?

also it looks like the grid of post stack 12ax7 is tied to it's own plate through the tone stack network as well.
is this some kinda way to shape the tone through negative feedback?

im real lost here.
can some one explain how this tone stack works to me?

thanks
robert

[kagliostro]

Sorry don't able to answer you but curious

can you post the name of that amp

or better post a link to the complete schematic

thanks

may be that is a way to bypass a preamp stage with the exclusion of V3B

Kagliostro

[tubenit]

Looks to me to simply by a hotswitch, switching out a gain stage & grounding the grid of that 12AX7 triode.

with respect, Tubenit

[jcm-jmp]

ok sorry guys

the switch is tied to another identical tone stack.

it's the crunch and lead channel's tone stacks from a triple x peavey.

here is a link to the full schematic
 http://schematicheaven.com/newamps/peavey_xxx.pdf

switching out a gain stage & grounding the grid of that 12AX7 triode.

the switch just switches between the 2 stacks depending on which channel is selected.

[JayB]

It's an attempt to make the tone stacks active using negative feedback. Subtracting a frequency at the grid will make it louder because the circuit will not have that frequency to add to the negative feedback from the output.

[jcm-jmp]

i think its an active tone stack boosted by negative feedback.

the pots are just mixing resistors controling the attenuation or boost of the frequency that passes through the stack and and added back to the signal at the grid. this gives either a boost or a cut in that controls (bass, mid , treb)frequency range at the gainstage following the tone stack.

the gainstage feeding the tone stack simpley provides gain to feed the tone stack which leads to another question, but i'll will wait to ask it after i get this tone stack figured out.

  JayB you posted as i was typeing

[jcm-jmp]

so my next question is

a cathode follower is better for feeding a passive tone stack because of the empeadence.

is plate driven better for an active tone stack? or is the empeadence the same? and using  plate driven just adds a little more overall gain at the tone stack?

[jjasilli]

It's an attempt to make the tone stacks active using negative feedback. Subtracting a frequency at the grid will make it louder because the circuit will not have that frequency to add to the negative feedback from the output.

I suspect that this is wrong.  I'm not sure, but I think this NFB will decrease signal strength.  This circuit is very complicated not to mention the cap to ground before the stack; and all the resistors and pots in the stack.  All that resistances to add up to the equivalent of one huge slope resistor networked with a 100pF treble cap.  That is, this tonestack should be very shrill, EXCEPT perhaps for that cap to ground and maybe the NFB.  If so, this seems like a very complicated way to solve an excessive treble problem that never had to exist in the first place!

Also I think Tubenit & Kagliostro are on to something with that 2nd triode stage being switched in & out for a lead boost.  If so the NFB may be killing treble to keep the lead boost from getting too shrill.  Actually it may also be attenuating all frequencies to keep the lead from being too muddy also.

[JayB]

so my next question is

a cathode follower is better for feeding a passive tone stack because of the empeadence.

is plate driven better for an active tone stack? or is the empeadence the same? and using  plate driven just adds a little more overall gain at the tone stack?

Cathode follower is better at driving any preceding circuit because of the impedance, about 1k. Perfect for driving long cable runs like one would see using an effects loop. Less loss of highs as well. Plate fed tone stack because of the impedance, 38k, will not drive a tone stack as well. You could still half the impedance, 19k, by running both halves of a 12ax7 in parallel for less loss. Which one is better? Well, personal taste I think. I think the plate follower allows more options when it comes to an active tone stack. For driving long cable runs with little lose? Cathode follower wins the game. There is also the virtual earth mixer which is something peavy has used in a few of their amps. It uses a plate and negative feedback too.

[JayB]

It's an attempt to make the tone stacks active using negative feedback. Subtracting a frequency at the grid will make it louder because the circuit will not have that frequency to add to the negative feedback from the output.

I suspect that this is wrong.  I'm not sure, but I think this NFB will decrease signal strength.  This circuit is very complicated not to mention the cap to ground before the stack; and all the resistors and pots in the stack.  All that resistances to add up to the equivalent of one huge slope resistor networked with a 100pF treble cap.  That is, this tonestack should be very shrill, EXCEPT perhaps for that cap to ground and maybe the NFB.  If so, this seems like a very complicated way to solve an excessive treble problem that never had to exist in the first place!

Also I think Tubenit & Kagliostro are on to something with that 2nd triode stage being switched in & out for a lead boost.  If so the NFB may be killing treble to keep the lead boost from getting too shrill.  Actually it may also be attenuating all frequencies to keep the lead from being too muddy also.

May be wrong in the detail, it's still using NFB to make that tone stack active by selecting which frequencies to feed back into the grid. Decreasing signal strength for those frequencies control by those pots. The grid is connected to one side of those pots, the output of that same triode is connected to the other side of those pots. So, your just selecting what frequencies get cut or boosted by the negative feed back.  That last stage is always in circuit for the lead and crunch channels, it's not used for the clean. The relay before it, just switches tone stacks depending on which channel being selected. Take a closer look, it threw my off at first too.

EDIT: Btw, that .02 cap you pointed out works just the same as the .047 cap for the drive in a tube screamer does except that it is fixed in this case. You could replace that 22k resistor for a 500k pot for a drive control.

http://www.schematicheaven.com/effects/ggg_ibanez_ts-808_ORIGINAL.pdf

[DummyLoad]

i'm probably way off base, but to me, it looks like simply two XOR switched tone stacks (identical values) in the FB loop of a plate follower configuration with one exception; in the "lead" mode the lead stack has a high-pass filter (c9 and r22) with a -3dB cutoff set to around 100Hz.

[bnwitt]

i'm probably way off base, but to me, it looks like simply two XOR switched tone stacks (identical values) in the FB loop of a plate follower configuration with one exception; in the "lead" mode the lead stack has a high-pass filter (c9 and r22) with a -3dB cutoff set to around 100Hz.

Yeah, what he said

[PRR]

> can you do this?

Nobody puts you in jail.

Ignore the caps and extra pots. What is the gain of this scheme when you turn the knob?

[DummyLoad]

> Nobody puts you in jail.

since you're offering up a get out of jail card...

from 1 to ~~25?

[jcm-jmp]

im thinking 0 because everything you put in is returned through neg feedback canselling each other out.

but i have seen gain control circuits simular to this with opamps in place of the 12ax7?

so on this particular drawing PRR I'm going to stick with gain of 0 until a cap is put in place then it forms an RC network with the pot's resistance giving a variable tone control. i.e which is how our little tone stacks are functioning in the triple x.

i think, maybe, i hope!

 

[DummyLoad]

im thinking 0 because everything you put in is returned through neg feedback canselling each other out.


a gain of 0 is irrational, less than 1 is attenuation. perhaps you mean unity gain?

[jcm-jmp]

  a gain of 0 is irrational, less than 1 is attenuation. perhaps you mean unity gain? 

yeah what he said!

yes unity gain.

check this out if you put a cap from the plate to the grid of the same gain stage you form  parralled capacitance with the tubes internal capacitance allowing you to form a low pass filter.

some high gain amps use this added cap to lower treble responce of that stage and smooth out the overall treble sound.

so this is a simple form of what we are doing with our tone stack. we could even put a pot in series with our cathode bypass cap and call it presence, or put the pot in place and increase the cap value and have a full frequency range gain control.

but

this are all things that can be done on the cathode side of the tube. the plate side of the tube can be used in the same manner using RC filters to shape tone characteristics through negative feedback.
and this is the xxx's choice of tone stack which explains why it's compressed and lacking some high frequency responce!
yeah, no, maybe?

[jjasilli]

a gain of 0 is irrational, less than 1 is attenuation. perhaps you mean unity gain?

Well, it's not irrational, it's just -0-.  The square root of 2 is irrational. See:  http://en.wikipedia.org/wiki/Irrational_number I thought the ) gain comment was just a joke, meaning that the NFB in this circuit would entirely kill the signal. 

Perhaps I'm just being too literal (again). 

[DummyLoad]

Well, it's not irrational

dictionary definition of irrational:

inconsistent with reason or logic; illogical; absurd

yes, it is... word context literal. why have an amplifier with a gain of 0? any number multiplied by zero is 0.

[PRR]

Gain of zero is quite rational and useful.

Yes, an "amplifier" which ONLY did zero would be a useless thing. But next time you turn your amp to "11", look at the other end of the knob. Don't it say zero? Could you use it when changing cords, so the pop and buzz didn't scare the dogs? Zero gain IS a useful position some times.

What if you put the pot in the middle? So that it is 125K to each side? Well, 135K on one side, but that's almost the same. Nobody knows this one? It is a plain UNITY gain amplifier. (Or a bit less due to the 10K and the non-infinite gain of one 12AX7.) An amp which only did unity may not be very exciting, but at least signal passes.

The third setting, wiper left, ISO got right. 250K/10K is 25. The gain can't be more than 25.

As Robert says, there's cancellation. If the signal completely cancelled, there would be no output to do the cancelling. So it does not quite cancel unless the gain is infinite. 12AX7 has gain of say 50, which is more than 25, but not a lot more than 25. The actual gain will be nearer 16. That's an obvious boost.

So at wiper right, center, left we have three gains:
0 === silent
0.9 == normal
16 == boosted

Just like that, it might be a useful gain-changer.

But as Robert says, there's a bunch of caps (and resistors!). We know that caps do one thing for low frequencies and another thing for high frequencies. A coupling cap cuts bass. A brite cap bleeds treble around a volume pot.

So I presume that R-C mess does different things for bass, middle, and treble.

And in each frequenncy band, you can dial from silence to normal to boosted.

Which is a nice flexible tone control.

[HotBluePlates]

Looks like I got around to this late...

I redrew the circuit last night to figure what was going on. I simplified to have 2 conditions: all pots max, all pots min.

With everything at maximum, the parallel resistances of the pots looked to me like they'd simplify to about 84k of resistance between the feedback and the connection with the 10k series input resistor. To me, it looked like a common-cathode gain stage used as an op-amp, with a 10k and 84k resistor, making gain 10/94 of the normal stage gain (call it 0.1*A). If the 12AX7 would normally be a gain of 60, this arrangement reduces the total stage gain to 60*0.1 = 6, with all controls at maximum.

I didn't want to untangle what the resistance would be with all controls at minimum, because it results in a Delta arrangement, but it looked to me like you could guess it to be nearly 10k between the feedback wire to the 10k series input resistor. That would be unity gain with all controls at a minimum.

I made a bunch of approximations on the way to deciding this was what the circuit looked like. So assumptions I made are probably wrong, but I feel the overall idea is sound.

So it looks to me like there's a 6:1 range of adjustment in this tone circuit (with dB of voltage being 20 log a/b), that's ~15dB of adjustment, which is very likely enough to do whatever you need if your sound doesn't completely suck going into the tone circuit. This stage is only for tone control and not really for signal amplification.

And it should be pointed out that there are already 3 stages of gain before this tone stage, and the last one probably cannot have high gain due to the load that follows. There is also significant attenuation going into the effects loop circuit, even if the loop is not used, and there is another gain stage after that which is all about getting a usable signal after knocking it down so far before and after the tone control stage.

All that before the phase inverter....

Basic design philosophy is not the same as for a vintage amp: (For distorted channels,) develop a signal in the preamp, use multiple methods to squash/tear it into a distorted sound, shape that sound with a fairly clean op-amp style tone circuit, send the shaped sound to external effects (likely at/near guitar-level) and recover the effected sound to feed to the phase inverter/output stage for relatively clean (reproduction-minded) power output.

[DummyLoad]

Yes, an "amplifier" which ONLY did zero would be a useless thing.

that was my point. 

[jcm-jmp]

But as Robert says, there's a bunch of caps (and resistors!). We know that caps do one thing for low frequencies and another thing for high frequencies. A coupling cap cuts bass. A brite cap bleeds treble around a volume pot.

So I presume that R-C mess does different things for bass, middle, and treble.

And in each frequenncy band, you can dial from silence to normal to boosted.

Which is a nice flexible tone control.

PRR,

thank you for the quiz. i really had to dig into triode gain stage design theory again to even get a close understanding of how the schematic you posted would work, which made a nice refresher for me.

thanks
robert  

[jcm-jmp]

What if you put the pot in the middle? So that it is 125K to each side? Well, 135K on one side, but that's almost the same. Nobody knows this one? It is a plain UNITY gain amplifier. (Or a bit less due to the 10K and the non-infinite gain of one 12AX7.) An amp which only did unity may not be very exciting, but at least signal passes.


this is the point that made me really use that 3rd and final brain cell i have. unity gain can easily be confused for zero gain. in context putting the pot in the center position provides no amplification i.e. zero gain.,
but
since there is a certain amount of signal being fed into the pot/stage you have a starting point that can be added to or taken away from. so this point is called unity (same output as the input)

any pot position counter clockwise from center is going to antennuate signal down to zero gain.
any pot position clockwise from center is going to amplifythe signal up to a gain factor of 25.
that 3rd brain cell is really smoking at this point  

[tubesornothing]

yeah I think its semantics now:  does "zero gain" me zero output signal, or does it mean the signal does not gain or lose anything (i.e. unity gain).  I am all for "zero gain" meaning "unity gain" (i.e. the input signal is not reduced nor increased.  So what do you call it when your output signal is zero?  How about "attenuated to zero"? or "no output signal".

Nothing like a few words to keep us away from the true meaning of christmas:  time to work on amps!

[jjasilli]

gain is a factor, i.e. a multiple.  Why fight it?  See http://en.wikipedia.org/wiki/Gain, especailly the last sentence:  "A gain of factor 1 (equivalent to 0 dB) where both input and output are at the same voltage level and impedance is also known as unity gain."

The language is already there to be used.  Deliberately conflating gain factors with dB is not helpful, and can only cause unnecessary confusion.  So there is a Zero associated with unity gain; it is Zero dB. 

[tubesornothing]

sounds good to me.  Although it would be just great to waste more bandwidth on esoteric discussions.

[plexi50]

I got a XXX last week. And i am glad i found this old post. What an ice picky shrill EQ.  Is there a way to just do away with the rc network and build a normal tone stack? The more i look at it the more confused i get. The XXX schematic does not label the pots as to what each one is.... But the JSX schematic does which looks to be the same as the XXX. Also isnt V2 D1 & D2 clipping the signal? I like some distortion but this is rediculious. Major almost uncontrollable feedback

[plexi50]

Clipped one side of V2 / D1 & D2. Amp sounds much better and smoother. The over kill fuzz pedal buzz is gone. The signal now is going straight through to R3 1Meg. Jumpered R3 1 Meg but needs to be there for some attenuation of the signal. Negligable channel switching pop. Now whats up with the excessive feedback in the EQ?

I also think that C19 or C20 cathode caps may be removed for a little less gain. This is a monster on Red Bull