Tone control in negative feedback loop

[iainpunk]

Hey everyone. I was wondering if it was possible to modify this tone stack in to an active tone control by putting it in the negative feedback loop of an opamp.

I have done this before with the big muff tone stack but it has DC continuity while this one doesn't. Any tips or ideas?

Thanks in advance, iain

[pinkjimiphoton]

you're getting dc cuz it would need a blocking cap i would think bro, try about an .047uF cap right where it says input. you may find it works better on the output, but i've seen dc weld and destroy pot wipers. smokin pots aren't good for you.
did i just really write that?
i seem to remember doing this on a thing i called a .... shit! can't remember. too many years ago. i built a 4069 variant of a redlama, and used some of the unused parts stacked up with a cap somehow or other for a tone control. it definitely worked, i'd have to look up the project on my other puter, which is down at the moment.

but yeah, i would think it would work. you may need to play with the coupling cap value. 47n should cover a guitar ok i'd think, but too much bass will likely make it unstable.

look forward to seeing how ya make out bro. peace!

[iainpunk]

Hey man, thanks for the reply but i think i didn't state the question clearly enough. The thing is that the feedback loop of the opamp has to let dc through so that the bias at the inverting input is the same as the non-inverting input thus letting the opamp act as a comparator and hitting the output straight onto the V+ rail. So to let the pedal not do that, i believe i need an opamp that has ground-level inputs and the output referenced at V+/2 or something like that.

I think i just answered my own question though, the lm386 has that, and it also has enough power to drive the power amp's low impedance.

iain

[diffeq]

To my limited knowledge (not an EE or anything), opamp main rule is that they try to match voltage on two inputs, so if you put a tone control in a negative feedback, an opamp will try to compensate all the treble/mids/bass cuts you do. Just like opamp rectifier compensates for a diode loss. I think a proper filter is more applicable for this kind of stuff. There are many active EQs out there. Then there is a buffered-out approach.

[antonis]

I was wondering if it was possible to modify this tone stack in to an active tone control by putting it in the negative feedback loop of an opamp.

Try to connect R_T middle lug to op-amp inverting input and R_M grounded leg to op-amp out..
(and bias non-inverting input to V_CC/2 in case of single supply, of course..)

[Ben N]

Slightly OT, but Jimi, I think with that TMB stack you don't really need a blocking cap to block DC, since there is no path from the input to the output that does not pass through a cap with a corresponding shunt to ground. There is really no way for DC to get from one side to the other. Which I think is the OP's point.

[Elektrojänis]

I think it should work if you add it to standard inverting opamp configuration... The tone control network would just be connected in parallel with the feedback resistor.

If i've undestood the feedback tone controls right, one thing you need to consider is that in the feedback loop the effects of the tone shaping are reversed... Boost becomes cut and cut becomes boost. (Technically more of the latter, as passive tonestack is only capable of more or less cut.) Standard Marshall (and Fender) style tone controls like that always have a mid dip and are incapable of "mid boost". If you place that in feedback loop you end up with a tone control that always has a mid hump and is incapable of mid cut. That might suit your aplication just fine, but I think you should just be aware of that.

Other thing that migh be interesting is to use the tonestack between the negative input and ground of the opamp and wire it in the non inverting configuration... Think Tube Screamer gain stage, but throw away 0,047µF cap and 4k7 ohm resistor and put the tonestack there instead.

[amptramp]

There is a 3-band Baxandall tone control here:

http://www.interfacebus.com/tone-control-3-band-tlo82-opamp-design.html

This gives you boost and cut for bass, mid and treble.  A tube version is Figure 18 here:

http://www.angelfire.com/electronic/funwithtubes/Amp-Tone-A.html

[pinkjimiphoton]

Slightly OT, but Jimi, I think with that TMB stack you don't really need a blocking cap to block DC, since there is no path from the input to the output that does not pass through a cap with a corresponding shunt to ground. There is really no way for DC to get from one side to the other. Which I think is the OP's point.

hi ben,
my thing was i was concerened about the dc on the pots, i've smoked a lotta pots lately

[PRR]

Try 10Meg from Out to -In, in shunt with your tonestack. That will probably let the opamp bias itself, without much effect on tone range.

The LM386 may not be the best horse for this because of limited gain, and it may go unstable when Treb is full-up (see LM386 minimum gain considerations).

As said, in the NEGative feedback loop, the tone curve is REVersed. Instead of a mid-dip you get a mid-Peak. The Fender stack, stock, is musically useful on e-guitar. Reversed I suspect it will be quite un-useful.

Why not use it the way Fender intended? Between one stage and the next?

[R.G.]

Thinking for a moment about opamps:
An opamp circuit with a small impedance in its feedback loop has low gain; an opamp with a high impedance in its feedback loop has higher gain. This is because it tries to maintain the inverting input at a virtual ground, and so the output is forced to a voltage that keeps the inverting input at virtually ground - or at least, what it thinks virtual ground is - that being the voltage on the + input.

Things change when you put a "T" network in the feedback path. With a T network having a path to ground, you have to consider that what the feedback path is really doing is attenuating the output signal before giving it back to the inverting input. If the feedback path attenuates a lot, the output has to be BIG to get enough feedback back to the inverting input to force it to virtual ground. If the feedback path attenuates very little, the output doesn't have to be very big to get the inverting input to virtual ground, and the gain is low.

Works for attenuation which is variable with frequency (i.e. a filter) too. If the attenuation at a given frequency is low, the gain at that frequency is low. If the attenuation at a frequency is high, the gain is high at that frequency. So putting a filter in the negative feedback path "inverts" the filter. Where the filter has dips, the opamp output has peaks, and vice versa.

PRR posted about the frequency curve being reversed just as I hit "post". Simul again.

For this case, putting the tone/volume stack of a typical guitar amp into the feedback path of an opamp will give you the inverse of what you expect.

I'd have to think about it some more, but what might work is making the input of the filter be the input of the feedback tone control, connecting the output of the filter to the output of the opamp, and connecting the "ground" for the tone/volume stack to the inverting input of the opamp. That is the transformation the feedback version of the Baxandall controls did, in effect. I'd have to do some analysis and modelling to tell if the opamp would really be stable with this kind of feedback or if it works at all, but that's my first guess at how to do this.

[pinkjimiphoton]

As said, in the NEGative feedback loop, the tone curve is REVersed. Instead of a mid-dip you get a mid-Peak. The Fender stack, stock, is musically useful on e-guitar. Reversed I suspect it will be quite un-useful.

oh man, you just gave me an idea to try maybe a new stupid simple gizmo coming,,, sorry for the ot for a min, but what if ya put two different tone stacks into the feedback loops of a dual oa and used an lfo of some kind to sweep between them?

brain death struck again, sorry.... lol

[PRR]

> connecting the "ground" for the tone/volume stack to the inverting input of the opamp.

I dimly recall that Fender did do that, patent and product. Brain-twist at first but it resolves into the same thing. With hi-Z in and lo-Z out, at the cost of an amplifier. So how is that different from passive stack and recovery amp? (OK, the NFB form adds hi-Z in.)

If a wheel works, why turn it inside out? Monowheel

[Danich_ivanov]

Spice simulation for the win 

[teemuk]

Here's how Vox engineers did it back in the day with a tone control circuit of great similarity:
http://www.korguk.com/voxcircuits/circuits/vlead50.jpg
Basically their "V.A.E." (as Vox titled the circuit) follows similar principle as James-to-Baxandall circuit conversion regarding which circuit terminals go where. Notice also how the DC feedback path is established with a high-ish resistance - in parallel - to tone control circuit's impedance. Effectively impedance of the DC feedback path is swamped by the tone control circuit at AC frequencies.

As noted, the active circuit configuration will behave somewhat differently than the passive version. Vox basically converts the tone control from a shunting attenuator to an active impedance bridge so controls will work more sensitively with a linear taper instead of logarithmic. Simultaneously, this difference calls out other tweaks regarding overall impedances of the associated RC filters. Basically all this is not any different than conversion from James (passive) to Baxandall (active), which also requires some principal modifications although the filter circuit arrangement (at quick glance) remains seemingly "similar".
Similarly as converting from James to Baxandall, the new active configuration also adds gain (read: lesser insertion losses) so signal-to-noise ratio is (at least theoretically) improved in comparison to passive scheme with huge insertion losses.

So yes, the conversion definitely works, but with other tone control types than Baxandall its just not that popular arrangement.