Fun with Baxandall

[marcelomd]

Hi,
I simulated the interaction between the bass and treble potentiometers using different Baxandall topologies for each side.

Each plot was generated with the same circuit repeated 3 times. One sweeping both controls (green), one sweeping only the bass (blue) and one sweeping only the treble (red). Ideally the blue and red curves would overlap the green one. That apparently is not the case.

[Vivek]

I suppose that the distance between the corner frequencies plays a part.

Question : is a slight interaction actually a good thing ?

[marcelomd]

I suppose that the distance between the corner frequencies plays a part.

Question : is a slight interaction actually a good thing ?

I suppose so, but note there's a bit of resonance in the response of the parts with two capacitors. I started to think about Q values, filter topologies, etc.

Personally I don't mind a bit of interaction when _playing/making/creating_ sounds. Hifi and/or mixing consoles are a different matter.

[Vivek]

Sometimes when I input Baxandalls into SPICE, I got wrong results if I did not use DC block capacitor at input.

Maybe the Sine voltage source has zero resistance and it's possible to load the IC or make wrong voltages on inverting input due to DC short to ground via the Sine voltage source.

[marcelomd]

Nope. Added series resistance and a coupling capacitor to the signal source. No effect.

I guess it is what it is.

[marcelomd]

Lesson learned today: Use the search, Marcelo.

Rob did an amazing job talking about this in 2019. A few posts after that he does a better job showing the different responses too. Thanks Rob!

I guess what I was wondering as far as this circuit is more about the physical arrangement of the components, especially in the Bass knob circuit, and if that was related to the fact that it has a subtle affect throughout its range.

I did a lot of manual circuit analysis of tone controls about 25 years ago.  There's many finer points but I can give you a brief summary.

There's two bass variants and four treble variants, which makes for a lot of combinations.   Some of the variants are have similar performance, but they are not the *exactly* same.  We can reduce the options down to:
- Bass control with one cap two caps.
- Treble control with one cap or two caps.

All circuits have subtle differences on the finer scale.  The smaller differences are hard to capture with simple manual circuit analysis.  For a start the analysis separates the bass and treble but in the real circuit the bass and treble interact.  One reason is the circuit is an additive equalizer, instead of the cascade or  "multiplicative" equalizer we have pictured in our minds.  Through analysis it is however possible to show differences between the two bass controls.  In general these differences are best seen through circuit simulations.

The problem with the design procedures is they are concerned with the controls at maximum but the differences show-up when the controls are less than the extreme settings.  These cases are obviously more difficult and messy to analyse.

For the treble control.  The two cap variant produces less boost, or a dip, just where the frequency response starts to rise above 0dB.   That could be considered as a faster *initial* slope.   At the higher frequencies the general behaviour of the treble controls are pretty similar.   The general  behaviour at difference settings isn't too different beyond that.  The two cap treble circuits affect the bass more (in that they detune the bass frequencies).

For the bass control.  In the full boost/cut positions the two bass variants can be made to produce the same results.    The differences come in when the control is not set to the extremes.    At small boosts for example the two cap variant boosts the lower frequencies, as the control is increased the "take-off point" near 0dB moves up in frequency and the low frequency "leveling-off" point moves up in boost.    The single cap variant starts boosting around the middle of the bass frequency range, as the control is increased the "take-off" point near 0dB moves up in frequency but in this case the low frequency "leveling-off" point moves down in frequency and increases the boost.   So very roughly the two-cap variant keeps the lower frequency "leveling-off" point the same whereas the single cap variant keeps the middle of the bass frequency range the same.  The two sound quite different when swept over the full boost-cut range even if the maximum boost/cut behaviour is kept the same.   The two cap bass variant has more of a tendency to produce a dip in the middle of the response.

EDIT:  This pic might help.  The first one is more of an apple to apples comparison as the treble circuit is the same.

VT0 = one cap bass  (one cap treble), VT5 = two cap bass (one cap treble)


VT0 = one cap bass (one cap treble), VT6 = two cap bass (two cap treble)

[Rob Strand]

A few posts after that he does a better job showing the different responses too. Thanks Rob!

A condensed version of wasting many hours of my younger days  .

[11-90-an]

A few posts after that he does a better job showing the different responses too. Thanks Rob!

A condensed version of wasting many hours of my younger days  .

a *productive* waste of time, at least for us benefiting and learning from your work and knowledge...   

[marcelomd]

A few posts after that he does a better job showing the different responses too. Thanks Rob!

A condensed version of wasting many hours of my younger days  .

Sometimes I feel like I'm reinventing the wheel over and over again...

EDIT: Well, why the hell not?

Since there are real experts here =) Two questions about the different implementations I've seen.

1- Sometimes there's a 1M or so resistor on the feedback loop of a bax stage. Any reasons for that? Simulations show very little change.

2- Any practical difference in using larger or smaller sizes of resistors, besides availability of components?

Thanks, guys!

[Vivek]

Sometimes when you reinvent the wheel for yourself, you have a great feeling of discovery in your heart which urges you to discover more !

And you never forget the lessons learnt by yourself

And you get better at discovering other things, some of which could be new to mankind.

I too spent some time with LTSPICE and 2 variants of this tone control about a month ago.




I suppose that choice of values has effect on

Input impedance
Band interaction

[Vivek]

I saw a James circuit on the net called Mid boost where a switch is used to increase the value of one of the caps around the treble control. Need to see what that does.


https://monster.partyhat.co/article/amplifier-tone-stacks/images/tone-stack-mid-boost.gif

[Vivek]

Nope. Added series resistance and a coupling capacitor to the signal source. No effect.

I guess it is what it is.

What are the DC voltages on the IC pins ?

I see + tied to 4.5 and - tied to ground. The IC should saturate  under these conditions.

[Vivek]

What are the best pole and zero frequencies for Guitar ?

Last month I had entered the Pete Cornish CC1 into LTSPICE, I'll post those graphs soon.

[Vivek]

1- Sometimes there's a 1M or so resistor on the feedback loop of a bax stage. Any reasons for that? Simulations show very little change.

Guess : to set DC gain of opamp when crackly old pots lose contact with substrate for short instances while bring rotated.

[Rob Strand]

Sometimes I feel like I'm reinventing the wheel over and over again...

EDIT: Well, why the hell not?

When you think about stuff yourself I goes to a different place in the brain.   Somewhere where you don't for get it.   Also, the thought process you went through can spark new ideas.    If someone explains a traumatic even it's definitely not stored in the same place as the person who experienced it.

1- Sometimes there's a 1M or so resistor on the feedback loop of a bax stage. Any reasons for that? Simulations show very little change.

There's two motives one is to stop the DC through the bass pot causing scratchy noises the other is to deliberately cause the bass to become more like a peaking bass control.  You will also see two 1M resistors which kind of makes the controls like those simple Mid boost/cut EQ's

In a similar vein, a coupling cap on the input side will cause the bass control to become more peaking when at full boost.  When the control is centered the input resistance is high and the high-pass cut-off is low.  When you advance the bass control the impedance drops and the high-pass cut-off shifts up and  the response is more peaking.

The treble side has some quirks a well.   The feedback cap (output to opamp -input) has less effect when the control is centred and when it is advance the treble roll-off increases and the treble side become more peaking like.     A cap across the treble pot is like the those simple mid EQ in that it is peaking from the start.

2- Any practical difference in using larger or smaller sizes of resistors, besides availability of components?

Outright scaling of all the values just boils down to extra noise.  If you keep the pots to say 50k the extra noise is minimal.

The resistor between the bass and treble sections (I think I called that RBB) affects the interaction of between the controls, and affects the dips you see in my old post and the depth of the notch when both are boosted mid-way.  For the one bass-cap circuit a small resistor between the bass and treble cause a reduction in the treble boost when the *bass* control is maxed out; perhaps mentioned in my old post.

The only other scaling thing is messing with the size of the treble pot vs the bass pot, nothing is forcing us to use the same values.   What happens here is large pot value cause the boost/cut to be bunched up at the extreme pot settings without much going on near the middle.

[marcelomd]

Sometimes I feel like I'm reinventing the wheel over and over again...

EDIT: Well, why the hell not?

When you think about stuff yourself I goes to a different place in the brain.   Somewhere where you don't for get it.   Also, the thought process you went through can spark new ideas.    If someone explains a traumatic even it's definitely not stored in the same place as the person who experienced it.

You are 100% right. I do reinvent the wheel regularly when programming. 1- to learn the "why" of things. 2- to exercise some skill, for example when learning a new language. And 3- not so often, to make a simpler version of something that already exists.
This is more or less related to the rubber duck effect. When you find a solution to a problem while explaining it to someone.

1- Sometimes there's a 1M or so resistor on the feedback loop of a bax stage. Any reasons for that? Simulations show very little change.

There's two motives one is to stop the DC through the bass pot causing scratchy noises the other is to deliberately cause the bass to become more like a peaking bass control.  You will also see two 1M resistors which kind of makes the controls like those simple Mid boost/cut EQ's

This is the one I'm thinking about. Looks like it just decreases levels.

Simulation:

2- Any practical difference in using larger or smaller sizes of resistors, besides availability of components?

Outright scaling of all the values just boils down to extra noise.  If you keep the pots to say 50k the extra noise is minimal.

Nice. Keeping the pots to 50k I can play with resistors to avoid super large (physically big) or super small capacitors (hard to find values here)

EDIT: I asked in another post. But what is the advantage of having mid controls before bass+treble? I read somewhere GK does it like this to reduce noise.

Thanks!

[marcelomd]

Sometimes when you reinvent the wheel for yourself, you have a great feeling of discovery in your heart which urges you to discover more !
And you never forget the lessons learnt by yourself
And you get better at discovering other things, some of which could be new to mankind.

Which is super cool, don't get me wrong =)

Nope. Added series resistance and a coupling capacitor to the signal source. No effect.

I guess it is what it is.

What are the DC voltages on the IC pins ?
I see + tied to 4.5 and - tied to ground. The IC should saturate  under these conditions.

I'm just simulating right now. The input signal is a super quiet 0.1Vpp. Simulating the transient response, nothing is saturating.

I saw a James circuit on the net called Mid boost where a switch is used to increase the value of one of the caps around the treble control. Need to see what that does.
https://monster.partyhat.co/article/amplifier-tone-stacks/images/tone-stack-mid-boost.gif

Nice!

[Vivek]

EDIT: I asked in another post. But what is the advantage of having mid controls before bass+treble? I read somewhere GK does it like this to reduce noise.

Thanks!
[/quote


If the mid control Opamp also provides some overall gain, it could give a higher signal to the Bass and Treble section, thereby reducing noise.