advice on RG's parametric EQ

[tempus]

Hey all;

I'm looking at building a 5 band parametric EQ based on RG's design here: http://www.geofex.com/article_folders/eqs/paramet.htm. I've got a couple of questions:

1. Obviously, U1b should be a low noise opamp, but does U2a and U2b need to be also, or would a TL074 do?
2. I plan on using this in my fx loop of my amp. Is U1a still necessary, or can I forgo the buffer?
3. Any special considerations for the caps, or will ceramic be fine?
4. If I were to use a split supply (+9v/-9v) would +Vb be connected to ground? If I don't use a split supply, will +Vb of 4.5v give me enough headroom if I'm feeding the input from the FX loop?

Thanks

[blackieNYC]

Keep the buffer.The TL072 will be fine. If you are concerned about the fx loop levels being a little higher, you could get a rail to rail op amp for a little headroom.  The sabrotone 3band parametric uses a max1044 to get + and - rails.  My EQ is just 9v and doesn't clip, but I haven't tried it in the loop.  Look up that sabrotone schematic, but - don't use the 2 10k resistors, one in series, one in the FB loop of the output.  Use the 2.7k in the geofex schem.  Use film caps from small bear electronics.  These are large values for a ceramic, and ceramics aren't all that accurate sometimes.
What is that rail to rail op amp - TL2272?  I think.
5 band is a lot.   Almost like a graphic Eq, you could leave out 5 pots if you just go with a fixed,somewhat narrow Q.   Likewise, do you really need each of these five to be variable frequency as well?  That's another five pots. Breadboard it with pots, try it in your loop, and maybe you'll find some fixed values you like.    Maybe just one fully functional parametric for the all important middle frequencies?

[tempus]

Thanks for your reply. Good advice on the breadboarding, but I think I'll go with the fully parametric EQ. I plan on using this with different speaker combinations for both live and recording, so the ability to change things up is kind of important. I currently use a graphic, and have 5 bands that I usually alter. The parametric will allow me to hit exactly the frequencies I need to fine tune my preamped sound. Thanks too for the tip on the film caps. I usually think of electros as having wide tolerances, but I guess ceramics suffer from that as well. Also, I'll try it with a 9 or 12v supply at the breadboarding stage and see if there are any issues.

Thanks again

[tempus]

Also, RG states:

The resonant frequency may be calculated as Fr=1/(2*pi*SQRT(L*C)), then later:

The center frequency of an LC filter is F0=1/(2*pi*L*C).

Which one do you use to calculate cap/inductor values? It seems to me that the resonant and centre frequencies should be the same thing. Is the second equation a typo?

Thanks

[PRR]

> "The center frequency of an LC filter is F0=1/(2*pi*L*C),..."

This must be a typo. After you multiply L*C, you have to take the square-root.

[tempus]

Thanks PRR - figured it must be something like that.

[tempus]

Just about have the design finished here, but I was wondering a couple of things:

1. Should the pots be linear or audio taper?
2. I've been working my design around parts that I have here. Will changing the values of the 51K or 470 ohm resistors affect the performance of the circuit (other than changing the range of frequencies)?
3. Will raising the supply voltage (to 12v or to a +/- supply) require changing the values of any of the resistors?


Thanks

[PRR]

The pots are complicated.

Frequency almost-always favors REVERSE-Audio taper. (Alternately, normal audio but put the numbers on backward.)

The boost/cut really wants a *custom* taper. The best standard option is LINear. The fault is that most of the action is scrunched-up at the top and bottom.

The useful range of Q (resonance) is so small you use Linear.

Try to keep all resistor values within 10%-20%, or re-compute the whole thing with your new values and see if it is tolerable. (Right-size resistors are surely "cheaper" than brain-work.)

[tempus]

Hey all;

I've been working on this project and I'd like to add a gain control to make the gain consistent across different boosts and cuts. I've tried just putting a pot (with a limiting resistor) in the feedback loop of the output opamp, but this gives more of a mix kind of effect. It doesn't actually decrease the gain, it lessens the effect of the boosts or cuts. Is there some way to add the gain control within the given schematic, or will I need another opamp after the output to do it properly?

Thanks

[PRR]

> doesn't actually decrease the gain, it lessens the effect of the boosts or cuts

Good ears.

Re-strap the INput amp for a gain of 2 or 3, and put a volume pot in front.

[Eddododo]

Hey all;

I've been working on this project and I'd like to add a gain control to make the gain consistent across different boosts and cuts. I've tried just putting a pot (with a limiting resistor) in the feedback loop of the output opamp, but this gives more of a mix kind of effect. It doesn't actually decrease the gain, it lessens the effect of the boosts or cuts. Is there some way to add the gain control within the given schematic, or will I need another opamp after the output to do it properly?

Thanks

Yea that isn't going to do it, I'm pretty sure. First two simplest options are 1) a passive  output vol pot  (no boost, obviously), and/or 2) make the first stage variable gain

edit prr beat me to it, more or less

[Eddododo]

The 'blend' effect- I assume it is because you are increasing the clean gain as well as altering the response 'centering' of the eq bands

[tempus]

Thanks for the quick replies. I'll give these a try. Can you suggest values for the pots? I'm planning to use this in my amp's FX loop, if that makes a difference...  PRR - am I correct in deducing that in your suggestion, this would give a gain of about 6 dB, but would be attenuatable ( down to silence?

And for both posters, are we adding the gain at the first opamp simply out of convenience, or is there a reason to add the make up gain before the EQ? Also, I'd have to set a fairly high input Z if I want to maintain the buffering of that first stage.

Thanks again

[PRR]

> set a fairly high input Z if I want to maintain the buffering of that first stage.

Why??

The EQ needs to see a low impedance (<<2k7). What you do on the front of that stage is up to you. If it is inside an effects loop, it isn't critical.

[Eddododo]

are we adding the gain at the first opamp simply out of convenience, or is there a reason to add the make up gain before the EQ? Also, I'd have to set a fairly high input Z if I want to maintain the buffering of that first stage.

Thanks again

0
Convenience, mostly. Any number of ways to add makeup gain / volume attenuation...... gain up front > eq > passive volume works, buffer-as-is > eq > variable gain works fine, but to me, variable gain input stage is pretty great because A) if the gain is AFTER the eq, then all noise from the previous stage is amplified, whereas gain BEFORE it is better for noise floor B) you keep the opamp count as-is C) with +/-9v bipolar supply, you have some error room  before your subsequent boosted mids, for example, cause nasty clipping.

Look at R.G's 'adjusticator' (its in his... whats it called.. circuit sweepings? section). Look at the high and low-impedance version...  What is your amp ? The nature of the FX loop comes into play with required input impedance of this eq, but it should PROBABLY be ok... the only issue with the high-impedance version is the instinct of 'higher resistance higher noise' but you'd have to ask someone else how much that matters (probably not alot in this application)

> set a fairly high input Z if I want to maintain the buffering of that first stage.

Why??

The EQ needs to see a low impedance (<<2k7). What you do on the front of that stage is up to you. If it is inside an effects loop, it isn't critical.

I think he meant of the second op-amp in the shown circuit (post-filter / out) rather than 'into' the input buffer.

[tempus]

Hey guys

Finally completed this build after an unusually lengthy debugging process. Anyway, it's done and sounds/works great. Thanks to all for your input and advice.