Control 3 Parameters at once. Maybe Digital Pots???

[Bill Mountain]

I have a simple graphic EQ that I like.

It's your basic 6 band graphic EQ but I tied two of the mid bands (500Hz and 1kHz I think?) to a single pot and it works great as a broad mid control.  I could also use dual gang pots but the one pot solution seems to work fine at the moment.  I run it anywhere from 9V to 24V depending on my set up and what I'm using before it.

I have a third Hi-mid band (2kHz) that I like to boost at the same rate that I either boost or cut my dual mid band control.  It adds some nice spark when I cut and helps make the mids more aggressive when I boost them.  So the basic settings I use are:

1. Mid and Hi-Mid flat or close to it.
2. Mid cut and Hi-Mid boosted.
3. Mid Boosted and Hi-Mid boosted.

I know that the easiest solution would be to leave well enough alone and just use separate controls but where's the fun in that?

I've spent the last couple of days reading up on digital pots.  I'm still not sure if a 5V digital pot and be safely used in a (possibly) 24V EQ but what perplexes me the most is whether a digital pot can be programmed or controlled in the manner I described above.

It would need to Boost or cut two bands of EQ and boost only one band of EQ (in either direction).

As you can tell I'm out of my depth.

Any ideas?

[Phoenix]

So, if I'm understanding you correctly, what you're wanting is a mid control that affects a broader frequency range, right? If so, you want a mid control with a lower q, rather than overlapping mid controls to fill out the frequency bands. No need to mess around with multi-ganged pots or digital pots. You can even make the q adjustable and the frequency sweepable. Check out Geofex.

[stonerbox]

Hi Bill. I hope I did not misunderstand what you are trying to do here. However if you feel like ditching digital potentiometers you could perhaps do this with vactrol? Check out Anotherjim's alternative in this thread of mine.
http://www.diystompboxes.com/smfforum/index.php?topic=116502.msg1079207

[Phoenix]

Actually, on a second read-through, it sounds more like you're after 3 basic "pre-set" eq curves:
Mid flat, high mid flat
Mid cut, high mid boosted
Mid boosted, high mid boosted

Are you hoping to have some sweep between these different settings on the one pot or something? I've no idea how you could implement that in any way that would be at all intuitive - starting at zero, what, both mid and high mid are flat, then as you rotate towards 12 o'clock, progressively mid is cut and high mid boosted? Then from 12 o'clock to max, mid is boosted but the upper mids stay locked at full boost? Seems awfully convoluted to me...

Why not use a three position switch and some simple logic to change the parameters of whatever eq circuit you're using? That would be very straightforward.

[Bill Mountain]

So, if I'm understanding you correctly, what you're wanting is a mid control that affects a broader frequency range, right? If so, you want a mid control with a lower q, rather than overlapping mid controls to fill out the frequency bands. No need to mess around wiht multi-ganged pots or digital pots. You can even make the q adjustable and the frequency sweepable. Check out Geofex.

That's part of it and definitely a consideration.  My only concern is that I'd like to avoid the huge peak at the center frequency with a wide Q.  That's why I used two bands.  1st, I like those bands and second, I liked the flatter peak at the center frequency.  Visually speaking, instead of one big mountain (lower Q at 750Hz), I have two little mountains (higher Q's as 500 and 1k) that cover the same amount of ground but not as tall.  Or the inverse at the same width it's a shallow lake vs. a trench.  Does that make sense?  This isn't right or wrong.  It was just my thinking during the process.

I'm famous for being confusing.  All the details were simply background information for my main question about how to get a circuit to boost a frequency regardless of which way I turn the knob.  My first guess was digital but I'm hoping there's other options.

[Bill Mountain]

Actually, on a second read-through, it sounds more like you're after 3 basic "pre-set" eq curves:
Mid flat, high mid flat
Mid cut, high mid boosted
Mid boosted, high mid boosted

Are you hoping to have some sweep between these different settings on the one pot or something? I've no idea how you could implement that in any way that would be at all intuitive - starting at zero, what, both mid and high mid are flat, then as you rotate towards 12 o'clock, progressively mid is cut and high mid boosted? Then from 12 o'clock to max, mid is boosted but the upper mids stay locked at full boost? Seems awfully convoluted to me...

Why not use a three position switch and some simple logic to change the parameters of whatever eq circuit you're using? That would be very straightforward.

That's an idea.  It's not meant to be intuitive like a regular EQ.  It's more for situations like:

1. Man I'm too clanky/flubby better cuts some mids (but thankfully the hi mids are boosted a hair to keep my sound from getting too dark), or
2. Man I'm just not cutting.  Better boost some mids (and thankfully the high mids are boosted a hair to keep my sound from being too nasally),or
3. I have no issues with my mids today but I want to use the low boost so I'll keep the mid control at noon so it's not in my way.

That's all.  As with most of my posts the circuits aren't meant to be practical.  I just want to learn something new by solving a made up problem.

[ElectricDruid]

If I were trying to put it on one pot, I'd want it like this:

8 o'clock: Mid cut, high mid boosted (A sloped EQ)
12 o'clock: Mid flat, high mid flat (Flat central)
4 o'clock: Mid boosted, high mid boosted (Mids Boosted)

That seems more logical to me. For the first half of the rotation, the slope should decrease and approach flat. For the second half, the flat 0dB level moves up and becomes a general boost of +x dB.

As to how to do it...well, that *is* tricky. As you said, a uP and digital pots is one way, but it'd throw up some problems for sure. I might instead be tempted to do it with a crossfader circuit. The first half would have to crossfade between a sloped signal and a flat one (giving a variable slope), and the second half is a crossfade between a flat signal and a boosted one. Bonus points if you can include switching so that you use the same crossfader for both jobs.
The crossfader could be an LM13700. There are plenty of good crossfader circuits using that. You can either use an inverting op-amp to flip the control signal so as one goes up the other goes down, or you can use a differential pair to do the same job, which gives you a different control response option. A full-featured example is here:

http://electro-music.com/forum/topic-54633.html

That's a full synth module, so you can pull some bits off it for your purposes.

HTH,
Tom

[Bill Mountain]

You taught me something new today.  I didn't know they made IC's for cross fading.  I've been wracking my brain trying to come up with discrete ways of doing that.  Thanks!

[Phoenix]

My only concern is that I'd like to avoid the huge peak at the center frequency with a wide Q.  That's why I used two bands.  1st, I like those bands and second, I liked the flatter peak at the center frequency.  Visually speaking, instead of one big mountain (lower Q at 750Hz), I have two little mountains (higher Q's as 500 and 1k) that cover the same amount of ground but not as tall.  Or the inverse at the same width it's a shallow lake vs. a trench.  Does that make sense?

That's not how q works - as the q gets lower, the peak gets broader and shorter, as the q gets higher, the peak gets narrower and taller.
It seems to me like two bands, lower mids with a low q, and upper mids with a higher q is exactly what you're wanting. No need to even make the q adjustable, or the frequency, you could make this just a two-knob affair, level control for low mids, level control for upper mids. If you really wanted you could include internal trimpots for tweaking the q of each band to taste, and then it's set and forget, never to be touched again.

[Bill Mountain]

Thanks!  I'll have to sim it again.

[slacker]

If it's your typical graphic EQ design like shown here http://www.geofex.com/Article_Folders/EQs/paramet.htm#graphic_eq then you might be able to do it with a dual pot. Wire one half like you have it now for your mids and wire the Hi section like shown below. Make the resistor half the value of the pot, then with the pot at 12 o'clock Hi will be flat and which ever way you turn the pot its resistance decreases and Hi will be boosted. This should work but it depends if you like the way the amount of Hi boost tracks the changes in boost/cut to mids, what ever solution you come up with that could be an issue though.

[Bill Mountain]

If it's your typical graphic EQ design like shown here http://www.geofex.com/Article_Folders/EQs/paramet.htm#graphic_eq then you might be able to do it with a dual pot. Wire one half like you have it now for your mids and wire the Hi section like shown below. Make the resistor half the value of the pot, then with the pot at 12 o'clock Hi will be flat and which ever way you turn the pot its resistance decreases and Hi will be boosted. This should work but it depends if you like the way the amount of Hi boost tracks the changes in boost/cut to mids, what ever solution you come up with that could be an issue though.

This is very interesting!  Thanks.  I look forward to playing around with this idea.

[slacker]

Be interested to hear if it works, by the way I just noticed a slight error, the resistor needs to be a quarter of the pot value not a half.

[Bill Mountain]

My only concern is that I'd like to avoid the huge peak at the center frequency with a wide Q.  That's why I used two bands.  1st, I like those bands and second, I liked the flatter peak at the center frequency.  Visually speaking, instead of one big mountain (lower Q at 750Hz), I have two little mountains (higher Q's as 500 and 1k) that cover the same amount of ground but not as tall.  Or the inverse at the same width it's a shallow lake vs. a trench.  Does that make sense?

That's not how q works - as the q gets lower, the peak gets broader and shorter, as the q gets higher, the peak gets narrower and taller.
It seems to me like two bands, lower mids with a low q, and upper mids with a higher q is exactly what you're wanting. No need to even make the q adjustable, or the frequency, you could make this just a two-knob affair, level control for low mids, level control for upper mids. If you really wanted you could include internal trimpots for tweaking the q of each band to taste, and then it's set and forget, never to be touched again.

Thanks for the help with Q!  I tried to adjust Q other ways before didn't have much luck with getting the same frequency response.  But by revisiting and using RG's resonance control I could tweak it to the same ball park.  Thanks!

[Bill Mountain]

Be interested to hear if it works, by the way I just noticed a slight error, the resistor needs to be a quarter of the pot value not a half.

Thanks again. I'm going to take a stab at simulating it later tonight.

The Q adjustments to the other mid band actually effected the hi-mid band so I'm thinking I might put that in a separate EQ stage so those values can stay unmolested.  I could also design that stage to be boost only (pot in the feedback loop instead of between the inverting and non-inverting inputs).

[Bill Mountain]

Be interested to hear if it works, by the way I just noticed a slight error, the resistor needs to be a quarter of the pot value not a half.

Thanks for the idea.  It looks like it would work out great!  I just need to experiment with taper.  I sim'd it but will need to order some parts to experiment with it on the breadboard.

Might end up sticking with separate controls but this idea is very intriguing to me.  Thanks!