How can I put a mid cut/boost into this circuit?

[vanessa]

I'm revisiting this e.q. schematic before I post a PCB layout for you all. My friend wants to get a mid cut/boost (or at the least a mid cut) tagged on to it before I build it for him. I'll set it up on the PCB to have it as an option. I'm not familiar with e.q.'s at all (but would enjoy the lesson). Help! 

[Dragonfly]

this might have some possibilities for you...

http://www.muzique.com/lab/tone3.htm

[vanessa]

That is interesting, but I'm not sure how a BMP tone stack would fit into this circuit. It has a separate treble and bass control. Also it would need to preserve the original tone of the circuit when bypassed...

[Mark Hammer]

One possibility is to re-situate the bass and treble turnover frequencies so that full treble and bass boost *sounds* like the desired mid-dip, though clearly that reduces utility of those controls AS treble and bass controls.

Another possibility, though a little problematic in its own way, is to replace R9 (currently 1k) with a mid-scoop circuit similar to that found on the Superfuzz and similar pedals.  I believe Jack O has a document on notch filters on AMZ, or perhaps on his AMZ-CD.  The trick would be deciding on component values that were able to produce a dip in the desired range without too much passive signal loss.

Just to refresh your memory, the Superfuzz mid-scoop (http://www.geofex.com/FX_images/univox1.gif) involves a 10k and 22k resistor in series, bypassed by a .001uf cap, with a .1uf cap to ground at the resistors' junction.

That 32k series resistance (10k+22k) is going to put a big dent in signal level.  *MAYBE* it can be easily offset by the volume pot, but maybe not.  If it were me, I'd certainly want to increase the value of that .001uf bypass cap if using it for bass.  A value of .0039 or higher might be appropriate.  The dip produced by the 10k/.1uf combo  (that's a lowpass at 160hz) might be about right, though.

Here's what I'm going to suggest, off the top of my head.  Make the 10k value 4k7, and the .1uf value .22uf.  Now make the 22k value 10k, and the .001uf value .0039uf.  That *should* result in mid-scoop that starts around the same point but brings in more uppermids and lower treble and does not provide as much passive signal loss as the Superfuzz circuit.  That way, you could flip back and forth (with a DPDT toggle) between the midscoop and a single fixed resistor similar to R9.  Volume level matching may be a bit off if you use 1k for R9 so maybe consider other values or even a 10k trimpot, tweaked to provide midscoop/no-scoop level matching.

[vanessa]

Mark, you and your Shin-ei circuits! LOL (Just kidding  ). This should work out perfect. I wanted to keep the integrity (tone) of the original circuit in place and this will do the trick.

One question... Is there a way to vary the scoop (say from no scoop to full) via a trimpot?

If so I'll incorporate that into the PCB layout. It would be very useful; the player could find their desired mid-cut they like, set it via trimpot. When they need it they can just flick the switch.

[Mark Hammer]

The simplest thing is to stick a variable resistor between the large cap (.1 in the original, .22 in my suggested list of values) and ground so that there is no bleed off.  Note that this would not have any bearing on the high end, though.  Imagine that you simply had a small-value cap bypassing the original 1k R9.  That cap would present a lower-impedance path to parts of the high end than the 1k resistor poses.  Similarly, if the cap to ground were effectively "lifted" by sticking a high resistance between it and ground, you'd have something that is essentially a .0039uf cap in parallel with a 4k7+10k resistance.  That would present a lower-impedance path for some high end than the 14k7 path provides.  Happily, you can always compensate by turning the treble down on the circuit you showed.

I would imagine that a 50k variable resistor between the cap and ground should be able to provide varying degrees of dip.  You may still want to use a bypass toggle for "no dip", though.

[vanessa]

That should be perfect. I think he only wants to be able to do a minor back-off from a full mid cut; so I don't think even a treble adjustment would even come into play for him, maybe for others?

[vanessa]

Here's the latest on the schemo. Does this look like it would work?

[vanessa]

I meant does the mid cut look like it would work.

[GibsonGM]

I modeled this in Electronics Workbench, ran a Bode plot...your -3 dB cutoffs are at about 147Hz and  4.3kHz, and with the variable R set to 0, you have something similar to a Big Muff mid scoop.  The signal is most affected in the first 5% or so of rotation (0 to 10K).   The scoop flattens right up quickly, leading me to think something like a 100k audio taper pot might work better...you'll need the 100K to bring the mid back up to flatline so 'full mid' is actually flat, anyway.  Looks good, breadboard and test this for sure!

[vanessa]

So you think a 100k over a 50k? Why Audio vs. Linear?

[GibsonGM]

Try 'em both, keep the one you like ;o)    Most of the 'action' is happening in the first 10K or so.  After that it really slows down (the proportion of rotation to how much the curve changes gets larger and larger).  So, you'll have much happening from say 9 o'clock to 11 o'clock on the pot.   Then less & less.   A REVERSE audio taper (now that I re-think this) has a logarithmic sweep; it will decrease resistance much faster than linear, and make more control happen over more of the range of the pot instead of just that 9 to 11 range.  I can't simulate log pots, unfortunately.  I'm sure someone more experienced than I will add to this, but this is a place to start.  There's more info on GeoFX "the secret life of pots" (see links).   

At 50K, you still have a little bit of mid scoop, so you'd never get the mids back to "flatline" again. If you're switching it out, tho, that may not bother you.  With a  100K pot, they are pretty flat, as if the control was almost no longer there, allowing you the full range of mid cut.  That's where I get that value from.  Again, I'd try both 50k and 100k, linear and reverse log taper, and see what you like (the beauty of bread boarding, and you will get to hear what I mean).  Just making suggestions as to what I'd try, I'm not a guru or anything, but I've come across things like this before!   Well thought-out design given that you haven't even tweaked it yet.  I might try this one out, too 

[vanessa]

I wish... Leo Fender designed this one. The mid cut is from Shin-ei (via the Superfuzz), and the values for that applied to bass for a mid cut are from Mark Hammer.
I'll only be able to use linear trimpots for this one. It's design is to allow the player to set the mid cut to a desired taste and then it is put into the guitar's cavity. Then you just let the switch do the talking.

[Mark Hammer]

There are two ways to attack this.  One is certainly via selecting a suitable pot taper and value (or tapering to taste).  Another is to figure out what you want and go with a 3-position toggle.  So, imagine the variable resistor (R12) is replaced with a 1M fixed resistor.  The centre/common lug of a SPDT 3-position (on-off-on) toggle goes to the junction of C10 and R12.  One outside lug of the switch goes to ground or some selected small-value (<1k) resistor.  The other also goes to ground through a somewhat larger value resistor (e.g., 10k).  In the middle position, the R to ground is 1M, so essentially "flat" according to GibsonGM (if you need a bigger R to accomplish flat, go with that).  In one of the side positions, you get full scoop or as much as you can stomach.  In the other position, you get mild scoop.  If your friend is okay with 3 possibilities, that could work, be easily repeatable and predictable, and would occupy less panel space.

Now, one of the more interesting things the functionally "infinite" resistance that our new R12 makes possible is skipping R9 and the bypass switch, because by not placing anything in parallel with R12, you effectively HAVE bypassed the scoop.  Where this gets interesting is the possibility of having a volume-compensated scoop. 

Let's say you have a fixed resistor (which we'll call R14) between C11 and R13.  If, for example, R14 was 10k, then R13 would behave like a 35k pot that could never be turned up full.  As R14 gets smaller, then R13 at maximum starts to behave more and more like a pot set to max.

Now, since the scoop is produced by bleeding off mids, you have this "yoking" of level and scoop/tone, which would be nice to get rid of.  You can probably see where I'm going with this.  If you had a 3-position DPDT toggle, the second set of contacts could be used to simultaneously adjust the value of R14 as it adjusts the value of R12.  When R12 is at minimum value and produces the most scoop, then R14 should be at essentially zero ohms to preserve as much signal as possible.  As the value of R12 gets larger and preserves more signal, then R14 should also get larger to ease our "virtual pot" made of R13/R14 below maximum and provide what sounds close to equal volume level.

You will still be able to GET variable volume, but your friend will be able to flick a switch and change the tone without having to dicker around with the volume pot.  I think that would probably appeal to them.

Happily, our needs conform perfectly to the way the switch works.  So, make R13 something like 8k2 to 18k (you'll need to experiment here).  When the toggle is in the middle no-scoop position, this value of R13 will set maximum volume.  When the toggle is in the max-scoop position, the corresponding side lug from the other set of contacts will now shunt R13 with a straight wire.  The value of R13 needs to be selected with R13 at max, so that full scoop and no scoop come out as close in volume as possible.  In the other side position, you'll get medium scoop, so here you'll need to select a parallel resistor (we'll call that R14a) that, in combination with R14, produces some attenuation but not tons.  I'm guessing that a suitable value might be whatever you've used for R14, or perhaps the common value just below that (e.g., if R14=18k, R14a=15k).

Make sense?

[vanessa]

It would seem to me that you would want to go about it by switching in/out the R14 resistor. And to make it simple for the player you would want to make that R14 a variable resistor so that the player could find the desired mid-cut they want and then adjust R14 for unity gain for when the mid-cut is not engaged. You can't really work with fixed resistors here as the player might add something to, or change up their rig at a later date. At that time they might want to re-adjust the mid-cut.
What kind of switch would you use for that? Would you not want to bypass the R14 resistor when the mid-cut is engaged? I would think you would want a 2 position, ?PDT to do it. 4PDT?

[Mark Hammer]

Well, you're certainly welcome to try a rotary switch or whatever meets your needs.  I often opt for 3-position toggles because they get a lot done for very little cost and effort.  Of course, it's understandable if it doesn't get enough done....

The fixed value of R14 simply assumes that some value can be identified that "unhooks" level from tone.  If you're lucky, it's a common value.  If not, then a trimpot may be called for.  Of course, if adjusting the level to compensate for passive signal loss through the notch is NOT a hardship, then simply skip it.  The reason I suggested it in the first place is because of my experience with my modded Shin-Ei.  Increasing the value of R12 can produce a big change in level.  If adjusting it is something you do when no one is listening, it's no big deal.  But if it's a gig situation, or the player wants to change tone mid-song, the amount of volume adjust required is disruptive.  Hence the search for something that delivers equivalent volume at different tone settings.

Having said that, a 4 or 6-position two-pole rotary will do just fine if the 3 toggle positions is insufficient.  The same principles that guide the toggle will work with a rotary.  Heck, for that matter, you could wire up a variety of different  scoop circuits (different location and/or depth) and select with the rotary.

[vanessa]

I did not even think of a rotary switch. Now I have to go back to the drawing board! Mark you are the best, always getting me thinking deeper!