A simple parametric EQ should not be this hard.

[stonerbox]

Hi gents,

The 1M pot does not change the frequency and I can't seem to find any reasons to why that is. The buffer, boost/cut and resonance does work.
When pin 3 (non inverted) is measured the voltage starts at 5.36 and then slowly begins to drop (about 0.05V per second). This also makes pin 1 (Output 1) and 2 (inverted) to lose voltage.


OPA2134 A
U1B
1. 5.36  -->  drops when pin 3 is measured
2. 5.36  -->  drops when pin 3 is measured
3. 5.36  -->  drops when measured
4. 0

OPA2134 B
U2A - Simulated inductor
5. 0.8
6. 8.56
7. 0.8
8. 18.63

[stonerbox]

On second thought, the Resonance and Boost only seems to affect the background noise and not the actual signal, strange. I've been at this one all day and starting to lose faith in ever solving this. Is there any other circuits out there that does similar things? What I need is a fixed 10-15dB boost, a fairly wide Q and with adjustable frequency.

[robthequiet]

Um, seems weird that the meter is sucking down the voltage. What are the measurements on U1A?

Also, what are you getting for Vb?

[stonerbox]

I solved the mystical drainage of pin 3, probably some components must have connected on the breadboard. The frequency issue is unfortunately still present, nothing happens when I adjust the 1M pot.

Here are the buffer voltages, the other half of this opamp is configured as a sallen-key LPF (currently unused).
U1A
1. 6.5
2. 6.5
3. 6.5
4. 0
5. 2.3
6. 18.1
7. 18.2
8. 18.5

OPA2134 A
U1B
1. 5.50
2. 5.38 
3. 6.07 
4. 0

OPA2134 B
U2A - Simulated inductor
5. 1.11
6. 8.44
7. 0.11
8. 18.58

[robthequiet]

Well, my guess would be to try to re-bias the circuit so that your op amp pins other than the power and ground are more like 1/2 +V, so around 9V instead of 5-6V. Why the frequency pot has no effect may reveal itself once the biasing is adjusted. How do the voltages on U2A vary if you turn the 1M pot?

[stonerbox]

When adjusting the 1M pot pin 6 varies from 8.46 - 9.24, the other pins voltages remains unaffected.

[blackieNYC]

Take a look at the sabrotone three (?) band eq. Based on this, just a different schematic. And a couple different values.

[robthequiet]

Pins 6 and 7 on your simulated inductor should be at the same voltage. I would re-check the wiring, I think.

[stonerbox]

Thanks for the tip Blackie! It looks very very similar to this one, except for the dual supply.

Rob, I rechecked the wiring and noticed some obvious errors, had the non inverted input connected to output. What was I thinking!?  Still no luck with the eq though. Damn, I've built STMs Three Band Eq without any issues before so why can't I get this one right? Questions, questions..

I swapped the 720k on the buffer non inverted input to a 100k and now I get about half the supply. How do I go about to re-bias U1B? The voltages looks dodgy.

U1A (Buffer)
1. 8.71
2. 8.71
3. 8.71
4. 0
5. 2.3
6. 18.1
7. 18.2
8. 18.5

OPA2134 A
U1B
1. 0.99
2. 1.08
3. 0.99
4. 0

OPA2134 B
U2A - Simulated inductor
5. 9.25
6. 9.31
7. 9.31
8. 18.62

[robthequiet]

Yep, looks like you went two steps forward and one sideways  but two out of three chips look much better.

On the Sabrotone the mixer opamp (U1B + input) is biased from the output pin of the buffer opamp through the 10K resistor.

In your schematic, the U1B non-inverting is coming from the junction of the 3K and 10K pot, maybe check that wire and signal path. For the 3K to drop the output of U1A down to around 1V might be a bit much. Maybe check both sides of the 3K resistor and the 10K pot to make sure they are as they should be.

If you wanted to post a photo of your breadboard maybe one of the eagle eyes on the forum might spot something.

[stonerbox]

Excellent Rob, I'll check that tomorrow it's getting really late here. Thank you very much for all the help I do appreciate it a lot.

[stonerbox]

High score! Had a broken cable in between the 3k to the non inverted input of U1B. Everything works now. Thank you Rob for pointing out the low voltage around the 3k resistor. 

I got another question, the wider the Q is the less dBs in boost/cut or are my ears playing tricks on me? I can't seem to get a really wide and fat punch in the mids with C1=0.15 C2=0.022 (124Hz to 566Hz. Q varies from about 3 to up to 16).

[robthequiet]

It's always something like a wayward wire or something,  Congrats!

I'm no engineer but from my experience there are a few key frequencies that resonate more than others in someone's ears, plus room and other signal chain factors.

Maybe one approach would be to put a mild shelving low pass filter, with a corner freq of 800Hz or so, in front of the parametric, then sweep the frequency to get a good peak where it does the most good at a moderate Q. That or maybe another parametric band. You may have to play around with the cap values to get the best mix.

[ElectricDruid]

I got another question, the wider the Q is the less dBs in boost/cut or are my ears playing tricks on me?

Not always, but yes in this design.

It's possible to design EQ circuits that don't have this property, but this isn't one of them.

There's a good article on the Rane site coming at it from the point of view of graphic EQ design:

http://www.rane.com/note101.html

HTH,
Tom

[Eddododo]

I had a rane parametric eq schematic somewhere, but in it's absence http://sound.whsites.net/articles/state-variable.htm
I love parametrics based on state variable filters, and they're pretty versatile

[blackieNYC]

The very narrow q setting is worthwhile - great before dirt, like a fixed wah. And narrow is also the best kind of notch IMO.
Minor problem I have:  I have one of these, I like it, but if I happen to power up the pedal while boosting(max) a narrow (Max)band, it can go into oscillation. Even if it's in bypass - it will bleed a little if oscillating- into other nearby pedals. I just dial back the gain and q before (or after) power-up, and then turn them back up if that's the setting I want. Swapping in different op amps for the gyrator is said to fix this. I got a little improvement then did t worry about it. Tell me if you get this issue.

[stonerbox]

Interesting reads, all of them. On Toms site I found the analysis of the dreaded MT-2 and its mid range boost/cut and in turn that lead me to the Wien-bridge over at ESP.
http://sound.whsites.net/project150.htm

It can sweep through frequencies without changes in the bandwidth. I'm planning to do a fixed mid range boost (with a Q of 1 or less) that can sweep through 100hz-700hz, Wien-bridge looks like the way to go.

[stonerbox]

Wait a minute.. the Wien-bridge is not game with a single supply, right?

[ElectricDruid]

Didn't you say you'd just seen an example from the MT-2? That'd be single supply...



T.

[stonerbox]

I built this one instead due to my lack of 50k dual pots at home.