Audio clips: 3 Band EQ and Acoustic Preamp box

[Transmogrifox]

This little acoustic sketch (actually about a 2.5 minute recording--big download) was recorded using a Tacoma acoustic guitar, Dean Markley Humbucker Acoustic pickup--and anyone who has used a Dean Markley knows they're kinda dull sounding.

I built this little preamp because 1, the pickup has very low output and needs a relatively low noise front-end amplifier, and 2, the sound is rather bland, so it needs some EQ.

From the recording you will see the preamp/EQ unit I made satisfies the need quite well.  I apologize for the rough sections and the fact that it's not a shreddin' distortion track--but I know some of you do acoustic stuff--so a simple EQ/preamp may be of interest to you.

Here's the link:

www.geocities.com/transmogrifox/acousticsketchmp3

If you want a schematic to build one, reply to this post and request it and I'll probably scan in a hand-drawn schematic.  The EQ section can be found on National Semiconductor's TL082 datasheet.  The preamp I designed keeping noise reduction in mind.  It works as a distortion pedal for electric guitar--has a gain of 40 dB from the preamp, and another potential gain of 20dB if all eq knobs are maxed for a total of 60 dB = 1000 V/V

I am pleased with this little active EQ unit that is provided from the datasheet.  It is well suited to guitar.

[lovekraft0]

Very nice - great sound quality! Is that direct from the preamp to line in? I'd like a look at that schem, if it's not a lot of trouble - mostly the preamp, I've got the 3-band from the National datasheet, but I'm really curious about how you approached the lo-noise 40 dB boost.

[PeterJ]

Excellent recording and playing. I'd also be interested in the schem, as I have a low output pickup on my banjo that could use a lot of boost. What kind of impedances are you seeing at the in and out?
Thanks,
Peter

[Incubus]

I'd be interested in a schem for sure....sounds really nice.

By the way, how would it sound when used for a guitar with no strings ?

[erio fraga]

I'm also interested in the schem.
Congratulations for your guitar performance

Erio

[Satch12879]

Sounds fun, I'm down :mrgreen:

[Transmogrifox]

Ok, I think I got enough interest to make it work posting...and yes, it was recorded direct from the preamp output to line in to the sound card.  Thanks for the compliments.

Here's the schematic link:

http://www.geocities.com/transmogrifox/acceq.jpg

It's hand-drawn and scanned in, but I think it's clear and legible.

Input impedance is essentially the 511k resistor you see.  Note that you have to add a coupling capacitor if using an active pickup system--since active guitar systems generally have sufficient output, this unit would not be practical for that any way--but if you're using as say, and EQ for a distortion pedal, you would want to capacitively couple the input and it would be good to reference the input ring to ground in this case.

Output impedance is relatively low.  The 47 ohm resistor is a short circuit output condition protection as a bit of extra help to the op amp.

I used a MC33079 op amp in mine--this is the quad chip.

Again, as you could tell from the recording, there was not much (if any) detectable background noise, and the preamp was turned up almost all the way, probably a gain of 75 = 37.5 dB.

The only time I was able to detect an additional noise at all was when I plugged it into the lead channel of my amp, and the noise level did not overpower anything.  Definitely more quiet than an average OD box, though the gain is not much less than an OD.

[lovekraft0]

Nice clean design! Just a quick question, though - in the interest of optimized ultra-low noise (at least in the case of use with a passive/piezo pickup), wouldn't you theoretically get better noise results using a JFET input opamp (OPA604, maybe?) for the first gain stage, coupled directly to the pickup, and eliminating the 511K input shunt resistor completely? Don't misunderstand me, I didn't hear any noise at all in the clip, I'm just wondering how you decided on this particular design. Thanks for taking the time to draw it up and post it!!

[Transmogrifox]

I determined the 511k resistor was irrelevant to the noise issue.   Here's why:  The effect of the input noise current and input noise voltage are most reduced when Rin is 0.   Given the nature of physical circuits, this cannot be so.  There has been derived an optimal Rin for achieving the best noise figure for an op amp circuit, and this is expressed as noisevoltage/noisecurrent.  For the MC33079 this ends up being about 9k.  Well, a guitar pickup output impedance may be in that range...at least within 5k of that, either way, within a range where it doesn't significantly increase noise by varying.  

The 511k resistor appears in parallel with the guitar pickup, so its noise addidtive effect is likely not even measurable.  In fact, this circuit IS as you suggested--the guitar pickup is coupled directly to the op amp input.  For humbuckers, a somewhat lowered input impedance is favorable to keep them happy as well.  With excessively high input impedances you can have a low damping condition, allowing the parasitic resonance effects to become more of an isssue as the Q is increased with increased input Z.  

The main reason the 511k resistor is there is to keep the op amp biased at the V/2 reference when the guitar cable is disconnected at the guitar end--just to keep the circuit happy.

I think a JFET input op amp would be a good idea if using a piezo system, however, notice that at a gain of 10, the input impedance of a BJT input op amp is still several Megs, so simply removing the 511k resistor would yield sufficient input impedance for piezos.  However, the other thing is that piezos may have output impedance characteristics that are better suited to JFET input op amps for superior noise characteristics, since the optimal Rin for a JFET input amp goes up significantly.

[lovekraft0]

Thank you, that was exactly what I needed to know! It's funny how obvious a lot of that is once I've had it properly explained.  :lol: Just one more dumb question and I'll stop bothering you - what value did you use for the unmarked cap between the high band pot and the inverting input?

[Transmogrifox]

Oh geeeezzzzzz.  Sorry, I totally spaced that:

It's a 5 nF cap.  4.7 nF would probably be just fine.  

Again, this EQ section is a direct cut-and-paste job from National Semiconductor TL082 app notes, so you can reference that if you get lost.  The combinations of parallel and series resistors were done the way they were to get the values I need based on what I have in stock.  The build would be more simple and neat if you used single resistors equivalent to the values that they combine to equate.

[lovekraft0]

Thank you, sir - you're a scholar and a gentleman !!