Graphic EQ Individual Freq. Switching

[mth5044]

Hola.

I'm working on a graphic EQ that is GEOFEX's graphic EQ with a buffer added to the front of it. http://www.geofex.com/article_folders/eqs/paramet.htm ~halfway down the page. Values of  resistors and caps are scaled to reduce possible noise. Also modified for +/-9V operation The circuit is being sim'd in LTSpice with the schematic below.



What I'm trying to do is have each band switchable. The first chart below shows all bands in their 'off' (center position on potentiometer) position except 1k at max boost. I thought it would be easy enough to remove the wipers from the gyrators to disengage the EQ band, but that messes with the circuit as seen in graph two. Gain gets bigger and I imagine that is the reason for the boost in Q. I have tried other methods like switching wiper to vcc/2 (ground in this case), totally removing the potentiometer and cutting out various sections of the gyrator. No luck.




The only thing I can think of is to switch the pot out with two 5k resistors set up to emulate the pot at half way. That would work (I think), but would require a 3PDT toggle and two extra resistors per frequency. Anybody have any ideas on how to do this with a less involved switching schem?

Thanks!

[blackieNYC]

I'll bet you can find pots with a switching center detent.  The boss eqs do this with a linear fader, but you might find rotary pots. When the wiper gets to the center it hits a mechanical switch which is not truly part of the pot electrically. It would probably switch in the 2fixed resistors.
If you don't find one, there are of course the pots with a push/pull switch function.  That could be your bypass.

[mth5044]

Push/pull is a good idea, thank you! I never heard of the center detent being an actual switch, but I have a few with the center detent that just let you know when you're at center. I would prefer it like that, but this is for a synth friend who wants to toggle on/off the individual frequencies.

Ah well!

[PRR]

Your gyrator resistor values look very-very low to me. That's not a noise-issue; the 15K around the opamp set a hiss level higher than any 680 or 3k3 value, and is always there not just when a pot is min/maxed.

Change your gyrator resistors back to "normal scale" and see if your issues are reduced.

If you are sure you need the low-values, change ALL resistors throughout by the same ratio. (This may lead to pot values you can not buy.)

[mth5044]

Thanks for the comments, PRR. The scaling was more of a preventative design as I'm running the sim first. The scaling came from your recommendation in this thread, post 34, then a few posts later the builder confirms it does drop the noise a bit.

http://www.diystompboxes.com/smfforum/index.php?topic=84033.20

I tried it in the sim with the two 15k's at 150k and all the gyrators and pots at their values that RG has in his article, but get the same results compared to the scaled versions.

However, I'm not really worried about noise at this point, more about how to have each frequency band switchable without affecting the Q and boost/cut of the other frequencies. Seems like a 3PDT switch between two fixed resistors to fake the pot in the middle and the potentiometer is the only way to go at this point.

[Mark Hammer]

Another way would be to build each gyrator section around a unique device.  So, band X built around op-amp A, band B built around another op-amp, and so on.  This would let you use a toggle or small push-button to individually bypass or engage each section/band.

Some years back, Craig Anderton had a project in Electronic Musician called the Peakmaker.  It was a bank of individually selectable bandpass filters feeding a mixer, in tandem with a full-bandwidth signal to be blended in.  The full-bandwidth could be inverted or non-inverted, such that you could produce dips or peaks in the final output.  The circuit did not let you make any of the bands independently boost or cut, but it did provide for a lot of tonal control in a manner that let you ignore, or use, the bands you wanted.

[Mark Hammer]

Here's the Peakmaker circuit.  Probably not what you're looking for, but I figured I'd post it as reference material for folks.  The resistors do not HAVE to be 11k.  CRaig used that value because he knew it would be 1%.  I'll just say that we had fewer choices in where to get things 28 years ago, and less collective knowledge, so the 11k thing was a safeguard for builders, appropriate to its time.

[antonis]

I figured I'd post it as reference material for folks.

Folks thank you very much, Mark..!! 

(I suppose that filter "boxes" consist of passive RC filters )

[Mark Hammer]

No, the little boxes are what is shown in the generic bandpass filter sub-circuit.  The cap values to produce each passband are shown in the table.

Although the circuit is shown as simply a bunch of filters that can be selected in on/off fashion, there is no reason why the output mixer could not be modified such that each bandpass section was followed by a pot, so that the user could mix in the amount of each band they wanted, instead of all-or-nothing.

[antonis]

Thanks again, Mark...!!!

(you've just put me in trouble to build it as a stand alone EQ pedal..)

[mth5044]

Hey Mark, thanks for the project! That's pretty cool. That phase inverter is so simple!