Adding filtered signals from the same source

[BetaTripp]

Hi, I've decided to try and design a simple tone control circuit for a guitar amp, i'v gone with 3 filters in parallel (high pass, low pass and band pass for high mid and low ranges) but im not sure how to add the signals afterwards to put them into a pre-amp... ill attach the schematic. i am planning on replacing resistors 12, 15 and 16 with pots to control each range. adding them simple by joining the outputs doesnt look right and doesnt really give me much of a signal.

cheers!

 

[davenn]

not surprised you don't have much of a signal. You have 2 of the outputs going straight to ground

C8, R16 and R15, C7

have you done any googling for what standard tone control circuits look like ?


Dave

 

[davenn]

here's one of dozens of examples ....
( it's a pretty standard layout)

even without the input and output amplifiers, its very different to your layout



Dave

 

[BetaTripp]

here's one of dozens of examples ....
( it's a pretty standard layout)

even without the input and output amplifiers, its very different to your layout



Dave

yeah ive looked at some. they just based around the same idea. filters in parallel..? thats the ones not based around an op-amp. The capacitors go to ground because they are low pass filters and the value of the capacitors dictate how much of the signal will go to ground...? the resistors are to protect the capacitors. Technicalities? when simulated the outputs give decent signals at the points indicated.

 

[davenn]

they just based around the same idea. filters in parallel..?

you really cant do it any other way ... you must separate the signal into the low, mid and hi pass sections so that each one can be adjusted separately then the results recombined

The capacitors go to ground because they are low pass filters and the value of the capacitors dictate how much of the signal will go to ground...? the resistors are to protect the capacitors

that's all incorrect

In your case ALL the signal from those 2 sections go to ground ... the caps don't need resistor protection

study some real circuits like the one I have shown you and get the feel for what is needed


What you need to remember is that any filter circuit like that is basically an attenuator
and that is why amplification before and after the filter section is needed



Dave

 

[hevans1944]

Technicalities? when simulated the outputs give decent signals at the points indicated.

Yeah, maybe, but of the "points indicated" only one (the junction of C5/R13) is associated with the output. The other two (junction of C8/R16 and R15/C7 don't connect to the output, so "measuring" at those points is meaningless as far as the output is concerned.

If you had actually built the simulated circuit with real components, attached a real signal generator or other audio source, and listened to the results at the points you indicated, using an audio amplifier connected to a loudspeaker, your mistaken notion of how tone controls work would have been immediately apparent and probably corrected by now. Simulations are useful for playing "what if" games if you understand the circuit being simulated. Simulations are of limited use compared to powering up and probing a circuit made with real components, which is hopefully what you want to end up with anyway.

It looks to me like you just "cut and pasted" a fragment of someone's circuit without any understanding of what the resistors and capacitors actually do to affect the output frequency response to the input signal. The resistors do not "protect" the capacitors. Where did you get that erroneous notion? The resistors and capacitors together form frequency sensitive voltage dividers. You need to add the outputs of the three RC dividers without them interacting. The easiest way to do this is with the op-amp circuit that @davenn showed you.

Have you cracked any books on electronics theory since you started posting here? Does the word reactance mean anything to you? How about Ohm's Law? Kirchhoff's Laws? These are just some of the "technicalities" you need to understand to create electronic designs that actually work. And, yes, you need to do math.

 

[davenn]

great additional info, Hop

 

[BetaTripp]

Haha, I didn't cut and paste anything. What I did was try to seperate the 3 frequecy bands, so they didn't interact, as I realised that if they were in series one filter could cut out some frequencies that a following filter might be designed for. I then calculated the components using f=1/2*pi*C*R, rearranging for C and inputting values for f and R. I centered the cut off frequencies around 600 Hz (HP and LP, BP centre being 600 Hz - from the frequency response of HP and LP which showed the mid ranges weren't being accounted for) as this is approx the mid frequency for the guitar range, being 80 - 1200 Hz. I then tried to connect the circuits so I could get an overall effect. I see that I have used the maths without knowledge of exactly what it means, and what other effects I should take into account, but this is why I thought writing posts on what I am doing would help as someone could give me some feedback. Thanks for covering the software as I generally just meant some technicalities would help me understand, as I didn't get why the results as they were. I know some theory yeah, but it's hard to apply it to designing (and reverse engineering) with just my view point on what it all means. Another view on things always helps.

Thanks for your feedback, I appreciate it.

 

[davenn]

it's all part of the learning process, at least you are giving things a go and that's great

Dave

 

[hevans1944]

@BetaTripp it's good to see you are trying to learn. You are on the right track using the filters in parallel, but without the op-amp to perform the summing function, the outputs will interact when you adjust the resistor values. Correct me if I am wrong, but I guess you are going at this from a musician's point of view rather than an engineer's point of view. That's okay. Trial and error is one of the best ways to learn if you take careful notes on what does and doesn't work. You can add in some theory later to help explain what you discover so you can design something similar further on down the road.

If you look at the theory some more, you will discover that an RC circuit alone will act as either a low-pass filter or a high-pass filter, depending on the location of R and C in the circuit, with about 6 db per octave of attenuation. The so-called cut-off frequency you calculated is the point where the signal has been attenuated by 3 db.

 

[AnalogKid]

What you are not seeing is that your circuit and the one in post #3 are functionally identical. By your description and schematic you start with one low output impedance voltage source and use it to drive three filter networks in parallel. That is what the first opamp and the three adjustable networks are. The filter topography is more mature, but that doesn't change the physics of the circuits. You are asking about how to "combine" the outputs of your three filter sections, which is exactly what the 2nd opamp is, an inverting summer. Because it is inverting, the first opamp buffer/driver also is inverting to preserve the signal phase. BTW, you cannot use a standard summing circuit with your design because it will load your filter networks, shifting or eliminating the corner frequencies. You have to add either a voltage follower or emitter follower to each one before summing.

"yeah ive looked at some. they just based around the same idea. filters in parallel.." a) so is yours; b) -

That's because after 100 *years* of development, that is what is known to work. The Baxandall circuit is 63 years old and in billions of products. Maybe there's a reason for that.

ak

 

[AnalogKid]

I just took another look at your schematic. It will not work the way you intend.

There is one single-pole 636 Hz highpass section and one single-pole 636 Hz lowpass section. When you combine these two the result will no change in tone whatsoever, but there sill be significant phase distortions. The bandpass section would be a very narrow band section, again with phase distortions, but the second section acts as a load on the first section (altering its corner frequency) and the first section is not a zero ohm source driving the second section (altering its corner frequency). This broadens out the center band peak response and increases the minimum attenuation through the filter.

Here is a simulation of the bandpass section. The second circuit isolates the two sections with a buffer, showing the effect of the stages loading each other in the first circuit.

ak