0
\$\begingroup\$

The image below features the EQ section of a DJ mixer with -26db gain reduction on each frequency band (hi/mid/low). My goal is to modify this circuit so that the op amp filters are "full kill" (aka isolators) - meaning that when you turn the VR all the way down, it kills the frequency band completely. What can I do to accomplish that (or at least increase the gain reduction beyond -40db)?

EQ section

\$\endgroup\$
6
  • \$\begingroup\$ The true null only occurs at 1 frequency then the width at -3 and -26 dB depends on Q which cannot notch the entire width @ -26 dB as this would require more tuneable notches like a half octave EQ rather than a 3 to 4 octave EQ \$\endgroup\$ May 29 '19 at 16:03
  • \$\begingroup\$ From which Yamaha gear this diagram is coming from? And the maximum/minimum gain will be equal around \$\large \frac{R_{120}}{R_{116}} = +/- 28 V/V = +/-29dB\$ or R112 and R108. \$\endgroup\$
    – G36
    May 29 '19 at 16:23
  • \$\begingroup\$ Thanks G36. This is from a Denon X400. I realize based on some of the answers that the db reduction rating is not for the entire band, and in this circuit theres a reduction curve across the frequencies of a given band. For my purposed I'm happy to roughly increase gain reduction per band without much concern for frequency precision (after all DJ EQing is crude). So it sounds like I can just increase R120 from 5.1K to 10K to change all three EQs from roughly -29db to roughly -50db? This should create a more audible "cut" within each given band in my understanding of this answer? \$\endgroup\$
    – wirl
    May 29 '19 at 19:14
  • \$\begingroup\$ But for 10K the gain is only 10k/130 = 77V/V which is only 37.7dB not a big improvement. And the R120 sets the "boost" gain. To change the "attenuation" gain only you need to change R106. But I do not recommend this way \$\endgroup\$
    – G36
    May 29 '19 at 21:22
  • \$\begingroup\$ Gotcha, 37.7dB for the mid range. I'm assuming the R106 your referring to is the resistor preceding the components in the screenshot i've uploaded (from the full schematic)? Just want to make sure that's not a typo. Also yes that's correct, I really just want to increase the attenuation gain, not the boost. Lastly - why wouldn't you recommend adjusting the attenuation gain with R106? \$\endgroup\$
    – wirl
    May 30 '19 at 4:43
1
\$\begingroup\$

My goal is to modify this circuit so that the op amp filters are "full kill" (aka isolators) - meaning that when you turn the VR all the way down, it kills the frequency band completely.

Redesign using tunable notch filters as opposed to gyrator inductors and series capacitors. Gyrators are great for emulating inductance but it comes at a price of series resistance (R108, R112 and R116).

However, killing a signal completely is largely unfeasible - you might be able to get down to maybe 100 db reduction with careful circuit layout and decent components but you have to ask yourself if this is really what you want.

\$\endgroup\$
2
  • \$\begingroup\$ Anything under 40db gain reduction would meet my needs. Ultimately these bands should be close to inaudible when turned down, which doesn't necessarily mean they need to be non-existant. To clarify - i'm modifying a prebuilt circuit in a DJ mixer. This doesn't allow much space for new components. Is there a way to simply change some of the R/C values in this circuit to increase gain reduction? \$\endgroup\$
    – wirl
    May 29 '19 at 15:53
  • 1
    \$\begingroup\$ No, there is no way to achieve what you want with minor or lightweight modifications. \$\endgroup\$
    – Andy aka
    May 29 '19 at 16:46
0
\$\begingroup\$

Your immediate question is answered, but I wanted to add some thoughts:

  1. You cannot achieve "full kill" across a band of frequencies, even in theory -- at least not if you restrict yourself to filters that don't have an infinite amount of delay. There is always a delay vs. completeness of attenuation in play. This isn't just defeatism -- this is a direct consequence of the underlying mathematics.
  2. Even for something theoretically possible, like a notch filter (which gives "full kill" at exactly one frequency) or a collection of them ("full kill" at a finite number of frequencies) you cannot practically achieve total cutoff -- in the real world, signals always leak through.
  3. The closer you push a real-world system to the ideal, the more nasty artifacts you'll create. In the case of audio, this will take the form of delay, or phase distortion, ringing, or pre-ringing. You may not mind how these artifacts sound -- or they may drive you (or some fussy recording engineer) up the wall.
\$\endgroup\$
1
  • \$\begingroup\$ Thanks for this reply. I think what might be throwing some of these answers off is that the term "full kill" is a term commonly used in DJ mixers. It's not a scientific term. It's also not an unusual feat I am trying to accomplish. It's very common. It just means that when you turn the low mid and high eq knobs down, you don't hear anything anymore. There are many DJ mixers that do this, including a similar model the Denon X500. I'm simply wondering if some simple component swapping can turn a traditional -26db eq section into one that reduces frequency bands more dramatically. \$\endgroup\$
    – wirl
    May 30 '19 at 4:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.