Proposed Baxandall Circuit with Third Midrange Pot - Review

[frets]

Hi Guys,

I hope you all are having a good start to the week.

This is not guitar related, but I'm hoping you all will help me out.  I've built a little Class D amp so I can listen to music in stereo in my work room (instead of on a tinny radio).  I've got the little amp built and it works great but I want to add some simple tone controls do it.  Somethin passive that will fit in a project box.  I've come up with a modification to the Baxandall that i'm hoping you'll look over and edit.  What I've done is put in a proposed Midrange pot.  Let's put it this way, a pot somewhere between the bass and treble in terms of response - a "Mid-Pot".  Each channel would have these three pots.

I did some calculations with respect to the circuit and came up with the following:




So,

1.  Does the Circuit look okay?
2.  If not, how can I remedy it?
3.  Do the values I've selected for the Mid pot look good; or, are there values that would be more appropriate to my intent?  4.  Such as doubling the values to 10nF and 100nF respectively?

As always, Thank you guys....

[thetragichero]

pot values would be helpful for the folks with better spice chops than i

and i know it's nitpicky but unless it's in a negative feedback loop it's a James tone stack. this terminology mix up is about as common in the guitar community as the tremolo/vibrato nonsense mr leo saddled us with

[frets]

Trag,
The pot values are 100k.

I just finished building it and trying it out. It’s a volume killer. The pots sound okay, it was a bit hard to decipher on my little test speaker.

Is there any way to fix the volume killing factor and still keep it passive?

[reTrEaD]

If I can steal a JohnH -ism, attempting to use a passive bax in a guitar can only end in "grief and tears before bedtime".

There's inherently a (roughly) 20dB insertion loss and it really should be driven by a fairly low impedance source. Although the series resistance of a magnetic pickup is fairly low, the series inductance causes significant high frequency loss when heavily loaded. So you should consider a magnetic pickup to be a medium-high impedance source. Scaling the resistances (and capacitive reactances) down, as you've done, makes the loading problem even worse.

Your proposed "midrange" control really isn't one. It's not a resonant filter. It's just a duplication of the treble control but with a lower corner frequency. Also, since there isn't anything to decouple the midrange control from the treble control, there's bound to be some ugly interaction between the two.

Sorry to be a Debbie Downer, but I can't imagine a way to do this without some active electronics.  Better if the tone controls are embedded in the feedback loop of some active gain stages, but at the very least, a low impedance stage to drive a passive bax and enough gain either before or after to compensate for the insertion loss.

Just my 2¢

[Yogi B]

Oct 19, 2022 11:54:46 GMT -5 frets said:

Is there any way to fix the volume killing factor and still keep it passive?

Not without limiting the range of adjustability, or swapping to something else entirely.

The fundamental idea of a Baxandall/James tone circuit is to get a flat frequency response when each control is in the centre by having the ratio of impedances equal for both resistance & capacitance. With the ratio of the two halves of a regular 10% log pot being 9:1, that ratio should be copied to the components either side (though most use the 10:1 because that fits better with standard component values). Thus in the centre position we have two (or, with mids, three) voltage dividers in parallel, each with the same 9:1 ratio and a resulting gain of 0.1 (−20dB).

For the active version this ratio is changed to 1:1 (with linear pots being employed), because that's the ratio of input impedance to negative feedback impedance that makes an inverting op-amp stage have a gain of one. However, I don't see a reason why not to adapt this ratio to a passive circuit instead — such that the gain at the centre is 0.5 (−6dB), rather than the usual 0.1 (−20dB). Though obviously this would give you less 'boost' available (since there's less difference between the flat output response & the input signal). For example, the following circuit gives a centre frequency of around 600Hz & roughly +5dB/−13dB at its extremes (relative to the flat response at −6dB):

(R_IN & R_LOAD are respectively the output & input impedances of the surrounding equipment, where the given values are loosely those at which they start to hinder performance.)

Oct 20, 2022 11:25:38 GMT -5 reTrEaD said:

Your proposed "midrange" control really isn't one. It's not a resonant filter. It's just a duplication of the treble control but with a lower corner frequency.

But maybe that's enough? Yes, ideally, it would control only mids, but compromise is already upon us with the limitation to a passive circuit.

Passively there's maybe two other options. First is to make a mids control as a combination of the the bass & treble controls, using both caps that: are in series with the pot (as in the treble control arrangement); and alternately bypass the pot (as in the bass control arrangement). The primary issue here is the resulting filter will be pretty low Q, thus shall either have too wide a bandwidth (such that it overshadows the other controls) or limited range of adjustment in comparison with its neighbours.

Second would be to introduce an inductor into the circuit, this can be better in terms of Q, but (still) presents the issue of interactivity — I don't see an obvious way to implement any third control that would be sufficiently decouped from both the bass & treble controls (perhaps from one or the other, but not both). (In an active solution the answer would be simple: add the mid control via a separate preceding/proceeding stage.)

[frets]

Thank you guys for your help and guidance with this issue.  I have learned from the entire exercise.  Thank you Yogi for going to the trouble of making a revised schematic.

I’ve decided to give up on the passive version.  I am building a simple 12 volt powered EQ based on the LA3600.  5 pots per channel.
Again, this is for a little 100 watt stereo I built for my workroom.  I figure it will get the job done.  And I can get it to fit in a project box. Here is the circuit for those interested.