a quiestion for the amp makers

[dschwartz]

Hi all, i´m planning to design a tube preamp, you know, they all look kind of the same.. but i want to improve a part that has allways bother me..
the tonestack...

i know tonestacks are pretty standard and they give sort of a "trademark" response, but in every tonestack i´ve built (specially with tube preamps) i found the controls to be too subtle, and less effective than i would like.. so..i´m thinking to use some kind of an active 3-band tonestack, using opamps (yes, opamps), in order to get a very responsive tonestack, with cut/boost at least 6dB per band..at the end, an equalizer...

my questions are:
am i crazy?
did someone tried something similar? how were the results?
do i need the tube buffer if i use an active SS equalizer?

a penny for your thoughts...

[armstrom]

For a simple single knob tone control try the Big Muff tone control. I've found it to give a wide range of tone control. You can hear it in a clip of my amp. http://www.mr2-power.com/hurricane%20test%201.mp3
the clip is farily long (2:26) At about 50 seconds in to the clip the tone control is turned all the way down toward the bass side. Then at 1:04 you can hear the amp with the tone all the way up toward the high side and then with the tone in the middle.
If you want more control any standard fender or marshall 3-knob stack will do the job.

I'm not sure what kind of tube preamp you're planning to build. If it's fairly high gain using high plate voltages I would suggest placing the tone controls between the two stages. Most preamp designs have a volume/gain control in there already. The second stage serves as a recovery stage to boost the signal back up after the losses created by the tone controls. If you're using one of the low voltage preamp designs you may find they don't have enough gain to allow you to use a complex tone control. Most of them only have a simple adjustable filter to help bleed off the low frequencies. Depending on where the control is and how its configured it may be highly interactive with the volume control also so that's something to watch for.

[jaysg]

Have you built a Baxandall type?  Ampeg used that on a lot of 70's amps and there's a lot of control when compared to a typical Fender/Marshall stack.

[stm]

Hi all, i´m planning to design a tube preamp, you know, they all look kind of the same.. but i want to improve a part that has allways bother me..
the tonestack...

i know tonestacks are pretty standard and they give sort of a "trademark" response, but in every tonestack i´ve built (specially with tube preamps) i found the controls to be too subtle, and less effective than i would like.. so..i´m thinking to use some kind of an active 3-band tonestack, using opamps (yes, opamps), in order to get a very responsive tonestack, with cut/boost at least 6dB per band..at the end, an equalizer...

*Lecture Mode ON*

Marshall tonestacks are very shy, to say the least.  8dB range for bass around 100Hz, around 6-8 dB variation for the mids, and 10dB range for the highs at 10kHz.

Fender and Vox tonestacks offer a much broader range of variation.  I recommend you experiment with Duncan Amp's Tone Stack Calculator (aka TSC) and see for yourself.  It is also a pretty good tool to tweak an existing design to your heart's content.

IMHO, the MID control in both Marshall and Fender tonestacks is a bit odd and not very effective.  In case of the Marshall, the MID control offers a rather narrow range of adjustment, and when BASS and TREBLE are in opposite extremes, the MID control interacts heavily with bass or high frequencies, depending on the case.  On the other hand, the MID control in Fender tonestacks behaves in a rather lame way: when both BASS and HIGH are low, the MID control works as a VOLUME control.  When BASS and TREBLE are high, the MID control works more like a real MID control.  When BASS and TREBLE and in an intermediate position, the MID control varies both mid frequencies and volume at the same time.  In this respect I like better some two-knob variations of the Fender tonestack where the MID control pot is simply replaced by a fixed 3k3, 4k7 or 6k8 resistor.  Much simpler and better defined behaviour!

The VOX tone control is another story.  It offers pretty wide range of adjustment for BASS (around 15dB) and TREBLE (over 20dB).  Both BASS and TREBLE controls interact with the middle frequencies, but in a clever way.  As BASS and/or TREBLE controls are raised, middle frequencies are slightly reduced, which improves the apparent effect of the BASS and/or TREBLE control knobs.  Also, both controls at minimum produce flat gain.  A winner in my opinion... But we are not in the land of milk and honey yet!  The VOX tonestacks as we know them have a flaw in the wiring that has survived to these days, probably for historical reasons!  This "error" produces and odd behaviour in the mid and high frequencies when the BASS pot is maxed out.  In Duncan's TSC you can see that around the last 10% of BASS pot rotation the mids drop heavily, and if the HIGHS are cut, they drop too.  Nevertheless this is simply cured by cutting the BASS pot connection to GND (the lower leg in the schem) and leaving it open.  Unfortunately this mod cannot be simulated in Duncan's TSC.  There was a page somewhere in the internet where this was explianed in more detail.

There are other topologies available, which again, IMHO, are not very useful, except for some particular cases.  For instance, one-knob tonestacks placed together with the volume/gain knob exhibit a crazy interaction.  Also, I have yet to see a well-behaved *passive* Baxandall tonestack.  Their active counterparts do behave properly, but they still need a huge separation between low and treble frequencies, so in the end you affect heavily very low and very high frequencies like 20Hz and 20 KHz, respectively, which are not very useful in guitar playing.  You can find this tonestack under the name of "James" in Duncan's TSC.

Finally, the BMP Tonestack is a very versatile building block, especially if a MID control is added.  This leaves you with a HI/LO control that's accompanied by a MID control.  Very wide range of tonal possibilities.  Useful modification as a guitar amp tonestack would include doubling the caps to lower the mid notch down to 500Hz or even a bit less.

So, what would be my favourite tonestack?  Something like a Fender tonestack with fixed mids, or a modified VOX tonestack without the BASS pot wiring glitch and with the mid notch scaled down to 400-500 Hz.

One last word of caution when using Duncan's TSC: there are two types of log pots in the software: A and B.  "A" type considers that the pot is divided in 30/70 proportion at mid rotation.  I think this kind of pot is extinct today.  Type "B" corresponds to 10/90 proportion at mid rotation, which is more in accordance to the typical log, audio or A type you can but these days.  I tell you this because if you want to tweak your own tonestack, just make sure you use a pot taper you can actually find.

Now addressing the subject of an active tonestack in a guitar amp, personally I avoid them since they are usually way much noisier (especially the treble controls) than their passive counterparts, especially if gain stages follow.  It is true that you need to compensate for their insertion loss, but in the end the result is much less noise anyway if you just place a flat gain volume recovery stage.  It is true also that their response can be affected by the output impedance of the preceding stage, as well as the input impedance of the following stage, but IMHO the lower noise overcomes the above.

*Lecture Mode OFF*

[dschwartz]

the thing is i dont really trust passive tonestacks, i have used duncan´s TSC  a whole lot, and what looks really nice at the screen, doesn´t really to my ears.. for example, bass controls almost never works, you may find some difference only if you turn that knob end to end... mid control , as STM say, are too dependable of the others controls, and treble is the only one that i can note a change..

STM, about that noise you say that active filters add..is really that bad?, i´m planning to use it in place of the tonmestack, after all the gain stages, and i think there´s no difference between that and putting an eq through the fx loop..it sure adds some noise, but that much to be anoyying? or something than cannot be trimmed out with some pF caps?

i plan to use a 12v AC for heaters, and rectify and regulate those 12V to 12Vdc for the active filters.. it should be pretty quiet..

[stm]

An active tonestack AFTER the gain stages is OK.  Thor is a working example.

An active tonestack BEFORE gain stages can be troublesome, since any hiss or noise added by the opamp will be amplified 100x, 1000x or 10000x, depending on how much gain you have after that.

In communications there is a simple principle regarding noise:  the noise introduced by the first amplifier stage in a chain is the one that really matters.  In other words, if your frontend is noisy, you will live with noise, no matter if all the following amplifier stages are noise-free.

[drewl]

Well you know about Duncans tonestack calculator....there goes my tip.....

It seems to me the less the stack alters the response the better!
Anyway, I've built many different types of amps, try some of the matchless preamps for a wide variance in tone, or I've been digging orange amps lately.

[DougH]

The vox top boost tone stack is one of my favorites. It is quirky, but in a very good way, as Sebastian pointed out. Here's a very good article about its history, and how it was a "happy accident": http://www.geocities.com/vintage325/topboost.html

One of the things to think about is that a passive tone stack will have more effect if it is positioned after the distortion producing part of the circuit. "Pre-distortion" tone stacks a la Fender et al will have less effect once the amp is turned up to the point of distortion.

A very smart implementation of passive tone controls in a guitar amp can be found in the Melissa amp design: http://vtg-amps.com/Melissa.html In this design, treble, mid, and bass controls are spread out over various stages of the amp and placed where they are most effective. (For example, the treble control is post-PI, where most of the distortion is generated, allowing you to smooth out the highs in high-gain operation, etc.) This amp was developed by an ax84 member and you can find a lot of discussion on it and how it was developed in the forum archives at ax84 ax84.com.

[snap]

the thing is i dont really trust passive tonestacks, i have used duncan´s TSC  a whole lot, and what looks really nice at the screen, doesn´t really to my ears..

the possibility (=need) to adjust the source impedance in the Duncan TSC should never be neglected!

[stm]

The vox top boost tone stack is one of my favorites. It is quirky, but in a very good way, as Sebastian pointed out. Here's a very good article about its history, and how it was a "happy accident": http://www.geocities.com/vintage325/topboost.html

This was exactly the article I was referring to, but I couldn't find it!

Now that you mention it, now I'm in doubt if cutting the extra bass pot wiring to GND will remove completely the mid interaction, or if it will just avoid the abrupt change near the end of the bass pot rotation.  So far, in TSC you can move the Bass pot from 0 to 90% and mid interaction is totally desirable and well under control.  Maybe another option to avoid the end-of-rotation jump simply consists in placing a stopper resistor in series with the pot.

[DougH]

I kind of like the mid interaction. If you turn up the bass, and adjust the treble appropriately it can give a unique "scoop" sound in high gain.

Check out that Melissa schematic if you get a chance.

[DSV]

In the Infinity (universal guitar preamp, here for the thread and schems: http://www.diystompboxes.com/smfforum/index.php?topic=71425.0 ) I used this:
- an active Baxandall tone control
- a mid scoop (which can be configured also for boost), like an equalizer band with a slightly lower quality factor.
- a switch, for selecting the scoop frequency (subtle, but works for my needs)

There's a pretty good range of tones available!
(Maybe the 4th order low pass filter I added for cabinet simulation has also some important influence though)

EDIT: I think that for high voltages you can use an eq band like the one in Mesa amps (with a true inductor).

[Jered]

the thing is i dont really trust passive tonestacks, i have used duncan´s TSC  a whole lot, and what looks really nice at the screen, doesn´t really to my ears..

the possibility (=need) to adjust the source impedance in the Duncan TSC should never be neglected!

  D Schwartz, what snap said. I had the same problem as you until Doug H pointed out that I needed to adjust the source impedance on the TSC.

[sean k]

I've been thinking about a tube preamp and the Melissa looks great.
I'm glad too that the big nanos like the .47 aren't in the high voltage 'cause a cap like that in film for 630V is way huge.
Also the 100 nano on the mid control could be a 100V as theres no DC on it except the grid bias of the next stage which is somewhat dropped through the various voltage dividers.

Instead of going straight out there you could set up the PI as a parallel tube with the tone control between the plate and the cathode then volume and out.

[dschwartz]

the thing is i dont really trust passive tonestacks, i have used duncan´s TSC  a whole lot, and what looks really nice at the screen, doesn´t really to my ears..

the possibility (=need) to adjust the source impedance in the Duncan TSC should never be neglected!

yes i know..but i don´t know how to measure de input and output impedance..

i like bmp a lot as a tone control, but i like to have all three controls. i just want them to be more effective..

that melissa preamp is very smart, although i still dont understand that bass control..why the diode??, and no source resistor?

[DougH]

that melissa preamp is very smart, although i still dont understand that bass control..why the diode??, and no source resistor?

I didn't see it in the writeup on the Melissa site. A few searches for "Melissa" at the ax84 forum would probably yield the answer

But if you think about it, you partially answered one of your own questions. The diode is a sub for the cathode (source) resistor as the fwd dc voltage of the diode is pretty constant. So it will enforce, say, a .7v drop from cathode to ground. The designer, Carl Berger, also came up with the idea of using a zener diode in place of or addition to a cathode resistor for providing fixed bias in an output stage (or mixing fixed/cathode biasing). So this may be a variation on that idea, although this type of diode won't provide the regulation of a zener, so it's not the whole story.

[stm]

I kind of like the mid interaction. If you turn up the bass, and adjust the treble appropriately it can give a unique "scoop" sound in high gain.

If you mean with BASS at maximum, then this "feature" should not be modified.

Check out that Melissa schematic if you get a chance.

Thanks for pointing that.  It is an interesting circuit and the toneshaping distribution makes a lot of sense.

that melissa preamp is very smart, although i still dont understand that bass control..why the diode??, and no source resistor?

I didn't see it in the writeup on the Melissa site. A few searches for "Melissa" at the ax84 forum would probably yield the answer

But if you think about it, you partially answered one of your own questions. The diode is a sub for the cathode (source) resistor as the fwd dc voltage of the diode is pretty constant. So it will enforce, say, a .7v drop from cathode to ground. The designer, Carl Berger, also came up with the idea of using a zener diode in place of or addition to a cathode resistor for providing fixed bias in an output stage (or mixing fixed/cathode biasing). So this may be a variation on that idea, although this type of diode won't provide the regulation of a zener, so it's not the whole story.

Watch out!  This is no ordinary silicon or schottky diode!  It is a CONSTANT CURRENT (CC) diode drawn as two overlapping circles with a bar at the cathode in the original schem.  It is marked as a 0.75mA CC diode, and it could likely be replaced by a J201 with its gate connected to source.  This special diode set the DC bias for the triode.  By definition, a current source has a nearly infinite impedance, thus audio signals do not pass here.

The 22uF capacitor is there just  to block DC current.  Then the 22n cap and bass pot enter into the game.  If I understand this part correctly, when the pot is shorted, frequency response is nearly flat and gain is higher for all frequencies.  When the pot is at maximum resistance, high frequencies retain their gain, while lower frequencies become attenuated as there is less bypass through the 22n cap.  In other words, this forms a sort of bass cut control.

[dschwartz]

i see., that diode is a CCS..

I ran and simulated a fet stage biased with zeners, it worked pretty nice, but at the end is the same thing as using trimpots..although zeners are cheaper..maybe those magic diodes may be used to bias fets..

[DougH]

I kind of like the mid interaction. If you turn up the bass, and adjust the treble appropriately it can give a unique "scoop" sound in high gain.

If you mean with BASS at maximum, then this "feature" should not be modified.

Yes, with the bass near maximum you will start hear it. Then adjust the treble appropriately because with such a deep, narrow scoop the treble will sound "spikey".

[DougH]

i see., that diode is a CCS..

I ran and simulated a fet stage biased with zeners, it worked pretty nice, but at the end is the same thing as using trimpots..although zeners are cheaper..maybe those magic diodes may be used to bias fets..

No magic bullet with that diode as far as biasing is concerned. I suspect it has to do with making the bass control more effective. You could always wire up a normal triode voltage amp with Rk and use a pot as a variable resistor to dial in/out the effect of a "big" capacitor for bass control. Don't know if it works as good as with the diode though.