3-band filter with Baxandall

[Im]

Hi,

I'm a french student in first year in electrical engineering and I'm working on a project of design for a hifi audio amplifier.

There are 5 main parts in my projects :
- input (thevenin equivalent with variation of +/-0.5V and impedance of 100Ohm)
- filtering (topic here)
- pre-amp (voltage amplifier, with potentiometer allowing a gain of 0 to 10)
- power amplifier AB (push-pull)
- output (speaker modelised by a 8ohm resistor)

The problem is in the filtering part. As you can see on the title, I have to create a 3-band filter WITHOUT any op amp. I thought about the baxandall structure, but it's only for 2 band. Then I thought about adding in parallel another baxandall structure, but a bit modified. It could be a pass-band, with cutoff and gain in treble/bass already fixed, but with a common emittor with potentiometer to control the gain of the band pass.

I'm not good in calculus of circuits and it takes me a lot of time, without a good method too. Moreover, my deadline next thursday (I have an oral presentation) and I worked the whole semester on finding solutions, but I had also a lot of work so I'm actually quite in a bad situation. What I need before monday (I also have exams from monday to friday), is to assure that my idea will work, make the calculus, and also implement the impedance adapations for the branches, because I need to have the same impedance on both branches.

By the way, do you think that using the potentiometer on a branch will modify the impedance ? If yes, is there a possibility to follow the impedance of one branch for the other one ?
What could be the problems for the phase ? (a standard baxandall doesn't have any problem in 2-band, because the low pass and high pass just cancel each other)

Thank you, if needed I can make some simulations on LT Spice, but actually, I don't want to simulate and lose time if it's just impossible to use this solution. I will use it to find the best parameters faster

Thanks for helping if you can,

Paul

[Benta]

I'm curious about your signal chain. Normally, filtering would be done between the pre-amp and the power amp. Why do you want to do passive filtering right at the input?

[jonpaul]

Bonjour Monsieur Im:

BRAVO pour l'effort et votre apprentissage électronique! Je suis ingénieur électronique depuis 1968, et j'ai habité et visité à Paris depuis 1970.

The passive tone controls of audio amps are normally just bass and treble, the midrange is left as is.

The Baxandall is a particular type of band sliding control, there are several other popular types.

a HP/LP or BP filter can be RLC, active etc.

99% of all hifi are just simple R-C filters placed after preamp and before power amp.

Beware that active filters can have a "noise gain" that can add unexpected noise.

Excellent classical amps and preamps are available in manuals PDF on line,

The designs of our old friend Tomlinson HOLMAN (THX) for his APT-1 preamp and amp circa 1975 are fine examples.

https://www.kenrockwell.com/audio/apt/holman-preamplifier.htm

Bon chance avec votre projects,

Jon

[Im]

Hi,

@Benta Thanks for your answer. I also asked myself when doing the chain analysis, and I thought that it would be better to filter small signals, then I don't lose a lot of energy in joule's effect. Is there any problem about doing it right after the input ?

@jonpaul I'm glad to see that there are still engineers in electronics and that they were not all replaced by robots Yeah for the bass/treble control.. but I put this in my requirement specification, and I have to stick to it :/ My problem is not on "how do I do a band pass ?" but how do I deal with the control between all the filters. So the baxandall was the only one I found that fixed my problem for the balance of HP/LP. Now, I need to add a BP filter, without influing on the rest.

I had an idea, that was to put 3 classical filters in parallel and put variable resistor on each branch, doing something like a millman. However, it didn't work on simulation. And the last idea is to put the three filters in series, making a buffer between every filter stage for the impedance adaptation. However, I still don't know if modifying the impedance in a bloc modifies the output or input impedance (I would need to make the calculus to see where each component act on the signal, so I couldn't see if it's neglectable in the input/output impedance, or not)

Have a good evening

Paul (Im is just a username haha)

[The Electrician]

Hi,

I'm a french student in first year in electrical engineering and I'm working on a project of design for a hifi audio amplifier.

There are 5 main parts in my projects :
- input (thevenin equivalent with variation of +/-0.5V and impedance of 100Ohm)
- filtering (topic here)
- pre-amp (voltage amplifier, with potentiometer allowing a gain of 0 to 10)
- power amplifier AB (push-pull)
- output (speaker modelised by a 8ohm resistor)

The problem is in the filtering part. As you can see on the title, I have to create a 3-band filter WITHOUT any op amp. I thought about the baxandall structure, but it's only for 2 band. Then I thought about adding in parallel another baxandall structure, but a bit modified. It could be a pass-band, with cutoff and gain in treble/bass already fixed, but with a common emittor with potentiometer to control the gain of the band pass.

I'm not good in calculus of circuits and it takes me a lot of time, without a good method too. Moreover, my deadline next thursday (I have an oral presentation) and I worked the whole semester on finding solutions, but I had also a lot of work so I'm actually quite in a bad situation. What I need before monday (I also have exams from monday to friday), is to assure that my idea will work, make the calculus, and also implement the impedance adapations for the branches, because I need to have the same impedance on both branches.

By the way, do you think that using the potentiometer on a branch will modify the impedance ? If yes, is there a possibility to follow the impedance of one branch for the other one ?
What could be the problems for the phase ? (a standard baxandall doesn't have any problem in 2-band, because the low pass and high pass just cancel each other)

Thank you, if needed I can make some simulations on LT Spice, but actually, I don't want to simulate and lose time if it's just impossible to use this solution. I will use it to find the best parameters faster

Thanks for helping if you can,

Paul

Do a search for "passive 3 band tone control".  I got a number of circuits mixed in with the active tone circuits.

[jonpaul]

Monsieur Paul, Rebonjour!

You cannot paralle the 3 filters! They must be in series, eg

LPF>>BPF>>HPF

Passive eg RLC, RC must have buffers in-between sections due to loading and filter impedances.

Beware of the noise if opamps are used.

Bon Chance,


Jon