Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Bass boost circuit - need steeper slope

Status
Not open for further replies.

David Kosman

Member level 1
Member level 1
Joined
Aug 30, 2015
Messages
38
Helped
1
Reputation
2
Reaction score
1
Trophy points
8
Location
Czech republic
Activity points
356
sorry for my english.

Hi, I've built this bass boost circuit for headphones (found it on the web), but I would really like to have the curve more sharper/steeper.
I was playing with it in Tina quite some time (changing the all resistor and capacitor values), but this is the most steep curve I have got (RED curve). The blue curve I drew would be ideal.

Bez názvu.jpg

I was thinking that maybe I could do a peak boost, which would boost the 50 hz and not that much the 100+ Hz frequencies, but what I found were some weird circuits with inductors but they didn't say the value, and when I tried it in the program I had to put there like 1H, which are huge inductors.

I think I would get a good result by, for example adding a small peak with negative dB value at around 150Hz, so I would lower that mid-bass which makes the sound muddy for me.

Thanks for your help.
 

You won't get far with a passive tone control like that. To improve the slope you need to either cascade (one after another) several stages or use an active circuit. If you cascade circuits, you will have to provide an isolation stage between them to stop them interacting and also an amplifier to overcome the losses. Remember that a passive circuit like that doesn't 'boost' the bass, it reduces everything else so the bass is proportionately higher.

Brian.
 

You dont want too steep a filter because that results in group delay and audible phase errors.
Better to get a better speaker or choose a gyrator design for a graphic or another so-called parametric EQ. These emulate large L's with impedance inversion to C's

https://circuitdiagramcentre.blogspot.ca/2013/06/10-band-graphic-equalizer-circuit-for.html

- - - Updated - - -

Either Cut the mid-range or boost, in the end fidelity is more than a boom box but a flatter EQ with a mic in several locations. Speaker location can affect more than what you show and Bass prefers the room corner to add inphase with long wave reflections.

Logitech and RealTek have tools to use a flatten PC audio with a sweep tone or impulse response, but if this a portable, a SUb-woofer addon will do more than try to compensate a small speaker.
 

The decibel numbers on your graphs is wrong. It looks like the tone control circuit boosts the treble when the pot is turned up or it boosts the bass when the pot is turned down and is flat in the middle. It is simply a single RC that has a slope of 6dB per octave. The bass boost starts at about 320 Hz and increases bass frequencies at 6dB per octave then levels at 58Hz. Then the frequency range is 58 to 320 which is about 2.5 octaves which is about 15dB (your graph shows only 5dB) .

If you increase the steepness too much then bass will sound like "one note bass", a bongo drum or a "boom box".

My main speakers are sealed with 8" woofers. They produce bass that is -3dB at 60Hz then lower frequencies are reduced at 12dB per octave. I made a stereo bass boost circuit with a Sallen and Key second order (two RCs) highpass filter with too much positive feedback so it boosts at its cutoff frequency instead of a -3dB cut. Now the speakers produce good bass down to about 30Hz very well.
 

Attachments

  • Bass Extension circuit.PNG
    Bass Extension circuit.PNG
    65.4 KB · Views: 276

Ok, so thank you all for your responses. Anyway, I found some 3-band equalizer and used one part from it.**broken link removed**

That is very much like what I wanted, because 50Hz is gained to like 5dB and all above 100Hz is almost flat. There is some excess in the low-end, but thats ok for me. I've also put R and C on the end, to roll off the lowest freqencies a bit.
However, I've got some problems with it:

First, the potentiometer is somehow bad set (or placed/rotated), because setting it doesn't affect gain, but something different. On the image it is set to 10 percent, but 0 percent makes it like all the way to high freqencies, and above 20 percent it does pretty much nothing.

Second, as sunnyskyguy said, there is some phase shift, and group delay up to 3ms in the low frequency without the RC roll-off at the end. With it, there is up to 10ms group delay in the deepest bass. I was using high frequency roll-off in the previous circuit and the phase is like -60 there, but I didn't even realize it, it sounds ok for me, so the actual group delay is what I worry about.

Thanks very much for your help.
 

Sorry, your image attachment expired (apparently due to a quirk of the system interface).

First, the potentiometer is somehow bad set (or placed/rotated), because setting it doesn't affect gain, but something different. On the image it is set to 10 percent, but 0 percent makes it like all the way to high freqencies, and above 20 percent it does pretty much nothing.

Your potentiometer value may be too low or too high. The correct value will create a usable rolloff curve over the entire range of the dial.

A possible alternative is to try different C values.
 

Looks like this is what you wanted, at the expense of 9dB loss and impedance rise from 1.8k to 5.5M

- - - Updated - - -

Our hearing spans up to 10 octaves (fmin x2^10)

They make EQ panels by 2-Octave, Octave, half -octave and third octave, then tuneable Parametric EQ. with 1V line In/Out

They toss the cheaper versions of these in recycling stores.

Make or buy is your choice.
 

Your graph has dB numbers that are very odd and are not consistent.
The extremely simple circuit has nothing active to add positive or negative feedback so the slope is simply one resistor and one capacitor.
 

Looks like this is what you wanted, at the expense of 9dB loss and impedance rise from 1.8k to 5.5M

Is the impedance bad? In the previous circuit, they had paralel RC with 22k Ohm on the end and when I tried to remove it, there was a lot more power, so I left it like that.

- - - Updated - - -

Your graph has dB numbers that are very odd and are not consistent.
The extremely simple circuit has nothing active to add positive or negative feedback so the slope is simply one resistor and one capacitor.

Don't know what you mean exactly, but the circuit is all passive, and I have a Fiio E9 amplifier behind it, so still a lot of power.
The second circuit is actually more effective than the first one, the bass is closer to zero dB, so less loss.

- - - Updated - - -

Your potentiometer value may be too low or too high. The correct value will create a usable rolloff curve over the entire range of the dial.

A possible alternative is to try different C values.

Tried that but that doesn't work. Looks like the curve is moving just left and right as I'm changing it. I would be ok with the level in the graph, but I'm afraid that I wont even be able to set the exact 10 percent.
 

Looks like the curve is moving just left and right as I'm changing it.
That's plausible because you never won't get a steeper characteristic with a first order filter (or cascaded first order filters). You need an active filter circuit, like the bass boost suggested by Audioguru in post #4.
 

Bez názvu.jpg

The first frequency response is potentiometer set to 10 percent, the second FR is 0 percent. It will be hard to set the exact 10 percent for me. I tried putting resistors right next to the pot, but it is still too sensitive. The resistor on the right had no effect on it so I removed it.

I will be trying different values of everything, but it is really an annoying work.
 

It is a balance that is not suited for your requirements to control the slope as the shift in balance also shifts the breakpoint.

What is stopping you from using an active filter with independent control of gain or Q that does not shift the breakpoint such as guru's suggestion or an active EQ gyrator BP filter?

There is not much point trying to stretch the simple Baxendal filter to do more than it was intended nor capable of doing.

In the end when the output impedance is much higher than the input, you will need to restore it with a high impedance amplifier anyway. The advice given will also reduce your frustration with a better result.

If there was a theoretic solution, someone would have showed you by now. I admire your persistence, and hope you make the wise choice of an active filter.
 

when I tried it in the program I had to put there like 1H, which are huge inductors.

Was this a series LC bandpass type? There is a substitute for a large inductor, called a gyrator, or active inductor. It has an op amp and capacitor.

Add another capacitor and you have, in effect, a series LC bandpass.



The output is taken across R1 (first scope trace). You do not want to take output from the op amp, unless you need a high pass response.

The second schematic is an equivalent passive series LC bandpass. The inductor is 6H. As you can see, the gyrator creates the same effect from an inexpensive 300 nF capacitor.

I am not saying this is guaranteed to work better for you than Audioguru's active filter in post #4.
 

My bass boost circuit cuts very low frequencies because my speakers are sealed so their response drops at 12dB per octave below their resonant frequency (60Hz) so there is not much output below 30Hz anyway.
 

Thank you. Yeah, I will probably need an active filter anyway, because I also want a Butterworth filter for some high freqency roll-off.

Just a sideways question. I have desktop active speakers with their own amplifier, they actually sound very good. Their frequency response extends down to 20Hz, but in a normal listening position there's nothing below 50Hz, even when they are placed in the corner of the room. I heard that when you cut the low bass, the whole sound gets improved. Considering they have the built-in amp, it probably wouldn't be improved if I made that kind of circuit?
 

Active Sub woofers already have a gain pot.
Given the quality of the speakers, phase distortion becomes meaningless. Are they Yamaha? Bose?

If a small room nulls the bass, it isn't 20 Hz... more like 200.
Large displacement bass only affects midrange audio speakers without crossovers, which is unlikely to be your case if your 20Hz is accurate... which is suspect.
 

They are small 2.0, 2-way speakers, they're designed to play from 20Hz (of course just to sell it better), which I hear only if I put my ear right to the bass reflex. So I thought when I make them free of that, cut the low bass, the sound will be better (soundstage, midrange, etc.) as they say. But the thing is I could do it just to the signal coming to it's jack input, not set it on their actual amp. So if the sound would get improved by that.
 

A frequency response spec must state how much the output level peaks or is cut, and how much distortion. A frequency response down to 20Hz might produce such a low level response that you cannot hear it but the speaker is responding. Maybe plus or minus 3dB would be good.
When a speaker is fed 20Hz its output distortion might produce 40Hz and or 60Hz but it is responding. Maybe 1% distortion is good.

If your bass boost overloads the amplifier or speaker then of course the sound will be better if you cut the bass. +3dB is double the output power. +10dB is 10 times the output power.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top