Core Shaper Parametric Eq [documentation]
Posted: 25 Mar 2021, 20:12
After showing you a couple of not-clones, here's a more original circuit I'm proud of. It's a parametric mid frequency eq with also bass and treble controls.
I think equalizers are very very powerful effects, altough they shouldn't be overestimated when post production is an option. Usually diy-ing a graphic eq is hindered by the amount of controls needed for a decent amount of bands, being them pots or worse, sliders. So I wanted to look for an alternative, and one which didn't feel like a downgrade, and a parametric or at least sweepable eq seemed the best solution.
Parametric eqs aren't new in the guitar world, the boss sp-1 being an example, and many others through the years. My approach to the bandpass is the Wien bridge eq, in particular a switched Q version through the means of positive feedback very much like in the JHS haunting mids, but which I want to point out I've taken from the book "Small Signal Audio Design" 3rd edition by Douglas Self directly. Unlike JHS, I haven't just copied the circuit as is, but I've tweaked the values to get the gain, Q and frequency ranges I liked, while using only E6 values because that's how I like it. To do this the Q switch shorts a series resistor instead of adding a parallel one for example.
The gain is +-16dB, the frequency sweep is about 300 to 3.3K, the Q is 2.4 in HI and 1 in LO. The beauty of this circuit is that the Q stays constant across the whole sweep and so does the gain.
Before this, there's a low gain input stage which gives a strong input signal and buffers the input while avoiding clipping at guitar levels throughout the circuit. The charge pump helps with the headroom, providing a +-9V supply so the boost levels can be reached without problems.
Following the bandpass, there's a very standard Baxandall 2 band EQ to take care of low and high frequency and add quite a lot of flexibility to the circuit when combined with the bandpass. The cutoffs seem to work well in my experience but they can easily be switched by switching caps if more controls want to be added. I chose the one-capacitor version for both bass and treble networks because that's the least amount of components and I'm cheap
but also because I like the cutoff being more constant and the loading on U1B seems to be good. C8 limits the gain at ultrasonic frequencies and is not always found altough it's very good practice. 100p is probably generous but that's not an issue for guitar especially.
Knowing that EQs tend to struggle with noise, either because the amount of gain they have especially at high frequencies, or because of the circuit itself, I prioritized noise over headroom, and this is reflected in the choice of a fixed gain input stage and a volume control at the end, which in the end isn't that different from what most eq pedals do. Headroom is still very good, I've measured input headroom throughout in "worst case" scenarios with max boost, and you have to try hard to make this clip, with an hot input signal, maximum amounts of boost, and maybe overlapping the mid band with the baxandall. Personally I didn't have any issues, and it needs to be said that some eq pedals get away with a single 9v supply and similar amounts of boost somehow. For line level signals an input pad might work well, and maybe 12V input. The impedances are low throughout to minimize noise, while still avoiding overloading the op amps. For the input stage TL072 is recommended because it's quieter with guitar inputs, while the band EQ uses 5532 which is quieter overall. I think both noise and headroom have been handled well and I don't have issues.
Here's some real measurements with ARTA as demonstration.
Flatness vs bypassed signal:
Low and high Q gain sweep:
High vs Low Q:
Bax Bass:
Bax Treble:
I have no frequency sweep pictures because I was testing this with a 50K pot which meant a narrower range, and didn't save pictures of the 100K sweep, but I can assure it's about 300 to 3.3K as expected.
No demo of this because it's such a versatile pedal it's hard to do it justice. It's not the end all tool, but works surprisingly well in shaping the sound of clean and distortion alike, and that was the goal. For finer tasks post production can take over for me, even a graphic eq won't be enough.
Finally, if you're wondering about the version 2.5, here's the short story: V1 was a gyrator based three band eq inspired by an ESP project. It works well in theory and has the advantage of being able to add more bands easily. Sadly, I wanted to have one or maybe two sweepable bands, and while gyrators can be sweeped across a decent range, the Q variation is too much, I calculated it to be equal to the frequency ratio.
V2 used another kind of Wien bridge bandpass, a simple one that required a single op amp, and the input stage was a darlington buffer to keep the op amp number even. For V2.5 I also considered even another version of Wien bridge, one that required two op amps but managed to keep Q roughly constant with a single gang pot. After finding available 100KC dual pots I went with this one with a very constant Q and the switchable Q, altough the single gang is charming.
Here's a not pretty but very compact vero for it:
And some pics of my build just to show off 
Hope you enjoy and let me know if you try it
I think equalizers are very very powerful effects, altough they shouldn't be overestimated when post production is an option. Usually diy-ing a graphic eq is hindered by the amount of controls needed for a decent amount of bands, being them pots or worse, sliders. So I wanted to look for an alternative, and one which didn't feel like a downgrade, and a parametric or at least sweepable eq seemed the best solution.
Before this, there's a low gain input stage which gives a strong input signal and buffers the input while avoiding clipping at guitar levels throughout the circuit. The charge pump helps with the headroom, providing a +-9V supply so the boost levels can be reached without problems.
Following the bandpass, there's a very standard Baxandall 2 band EQ to take care of low and high frequency and add quite a lot of flexibility to the circuit when combined with the bandpass. The cutoffs seem to work well in my experience but they can easily be switched by switching caps if more controls want to be added. I chose the one-capacitor version for both bass and treble networks because that's the least amount of components and I'm cheap
but also because I like the cutoff being more constant and the loading on U1B seems to be good. C8 limits the gain at ultrasonic frequencies and is not always found altough it's very good practice. 100p is probably generous but that's not an issue for guitar especially.
Here's some real measurements with ARTA as demonstration.
Flatness vs bypassed signal:
No demo of this because it's such a versatile pedal it's hard to do it justice. It's not the end all tool, but works surprisingly well in shaping the sound of clean and distortion alike, and that was the goal. For finer tasks post production can take over for me, even a graphic eq won't be enough.
Finally, if you're wondering about the version 2.5, here's the short story: V1 was a gyrator based three band eq inspired by an ESP project. It works well in theory and has the advantage of being able to add more bands easily. Sadly, I wanted to have one or maybe two sweepable bands, and while gyrators can be sweeped across a decent range, the Q variation is too much, I calculated it to be equal to the frequency ratio.
V2 used another kind of Wien bridge bandpass, a simple one that required a single op amp, and the input stage was a darlington buffer to keep the op amp number even. For V2.5 I also considered even another version of Wien bridge, one that required two op amps but managed to keep Q roughly constant with a single gang pot. After finding available 100KC dual pots I went with this one with a very constant Q and the switchable Q, altough the single gang is charming.
Here's a not pretty but very compact vero for it:
