Balancing impedance of piezo electric pickup with magnetic in box

[simit]

Hi.

First post so I will introduce myself!

I'm an electrician and understand about electricity and making good connections etc... I'm also a sound engineer and understand about signal processing. I would like to combine these skills into something I can use for my other interest, playing guitar! I have a first project in mind but, coming into this as a beginner I don't know where to start.

I'm looking to combine the output of a magnetic pickup with a piezo into one signal which I can then process further or bung into one DI box to go through the desk. With large impedance mismatches (magnetic humbucker is about 12K and the piezo around 1M) I guess I need some active circuitry to do this.

So my specification is this:

The circuitry is going in a stomp box (or just box without the 'stomp!) rather than on the guitar. This is a vintage guitar and modifications to the instrument should be reversible. My plan is to run the magnetic output down the tip of a 3P 1/4" jack, piezo output down the ring and a combined ground on the sleeve. This will then go direct into my 'pickup combining box'.

The piezo is being mounted internally to pickup the soundboard sound. The magnetic pickup is in the bridge position and should fill in the trebles that won't be heard through the piezo.

Other than that, my plan is open! And I have NO IDEA where to start in terms of a circuit...

I would like to retain clarity as much as possible but with the ability to attenuate each signal for balance by a pot per pickup. Ultimately I will want to EQ this signal. I wonder if being able to EQ each channel independently is beneficial? I am thinking that an instrument level output and a DI level output (~600Ohm) would be ideal but I'm open to suggestions why one would be better than another?

Does something already exist? Am I crazy trying to build this?

Do let me know what you think, even if you do thing I'm mad attempting this! I won't be offended!

Thanks everyone!

S.

[ashcat_lt]

Does THIS help?  It is presented as an on-board solution, but there's no good reason you couldn't put it in a box on the floor.  I guess cable capacitance will change the frequency response compared to what's shown, but the mag pickup will end up sounding more like it "should" that way.

[simit]

Thank you for that link. It's going to take me a little while to 'digest' that page so I'll report back once I have read and understood.

In the meantime I found this page: http://www.till.com/articles/GuitarPreamp/

In theory, would two of these preamp circuits in the same box then wired together at the output give me what I'm looking for?

[PRR]

What if it was two guitars (instead of one guitar two pickups)?

As a sound guy, you would run them to two inputs on a mixer.

Or since mixer input impedance is low relative to guitar pickup impedance (wound or piezo), you want two DI or two hi-Z buffers. Maybe different piezo or wound, and maybe not.

IMO, you are likely to want different EQ on each pickup. This leads to a $59 Beeringer mixer with maybe two buffers out front.

You can simplify or complicate to taste.

[Quackzed]

fwiw i've done a few buffers for piezo's and magnetic pickups as well as the tillman which is the 'goto' solution for many. imho the tillman isn't ideal in that it can distort due to the relatively large signals a piezo pickup can put out. magnetic pickups on the hot /loud side can get up to @1v where a hot piezo can do @5v or more even. pretty hot for a 9v circuit... thats why i like rail to rail opamp buffers. like the ne5532. rail to rail meaning the opamp chip will swing to the full voltage rail before clipping, where most ordinary opamp chips will distort before they hit the voltage rails, and jfets like the tillman definately will. so for headroom / no dist. i'd say go opamp. you can even run off of say 18v rather than 9v (same thinking to avoid distortion)
   the other thing i'd mention is that cable loading and low imp will tend to dull or kill the highs on a mag pickup, while on a piezo low imp will actually kill lows. its the opposite of what you expect, so be aware of that when tweaking your design..
at the end of this thread theres a setup similar to the one you're proposing using opamps, the first 2 opamps are the buffers (up to r3 r4) and each has a volume control (vr1 and 2) for mixing signals as well as taming overly hot piezo signals which might be usefull as well, the grey schem blue writing at the end... and sub a 5532 dual opamp for the tl072 dual opamp... you could probably build it up to the r3 r4 resistors and use it as is.
http://www.diystompboxes.com/smfforum/index.php?topic=113077.msg1045717;topicseen#msg1045717

[PRR]

> ne5532. rail to rail

FWIW: '5532 is not rail-to-rail. Not real different from a '741 or a TL072.

But if you're happy, that's all which matters.

[ashcat_lt]

Yeah, honestly if it was me, I'd probably just use a couplefew opamp stages and call it good. 

[Quackzed]

FWIW: '5532 is not rail-to-rail. Not real different from a '741 or a TL072.

I don't know why I thought they were rail-to-rail ?!? i ordered a bunch of em for different stuff JUST because i thought they were rail to rail.
all i can think is that i had my eyes cross over some low noise and full swing stuff.
well i fooled me, lol!
in that case rail-to-rail might not be TOTALLY necessary but can't hurt? 

[ashcat_lt]

Spose it depends what you're driving.  A signal big enough to clip a TL07x at 9V supply is going to crush any guitar amp, and a little opamp crunch will barely be noticeable. 

[PRR]

> thought they were rail-to-rail ?!?

Datasheet does not lie. (Never tells the whole truth, but no lies.)

"Peak-to-Peak Output Voltage Swing 26 V Typ
With VCC± = ±15 V and RL = 600 Ω"

+/-15V is 30V. Perfect R2R would give 30V p-p output; typical R2R with light load maybe 29.7V p-p. Instead it falls 4V short of the total rail voltage.

4V shy (2V each side) is no big deal at +/-15V, but looms larger at 9V.

That's "typical". "Min" is 24V p-p under specified conditions.

The internal circuit suggests ~~0.7V loss each side plus 0.6V loss when working high current. You could round that to "1V each side for typical small audio uses". So 9V supply may give 7V p-p.

The difference 7V p-p to 8.7V p-p of a "true R2R amp" is less than 2dB. If 7V is not enuff, 8V or 9V won't make you deeply satisfied, you need lower goals or higher battery.

[Perfboard Patcher]

Hi simit,

I have built a similar thingy. I don't know exactly what kind of piezo you're using, but in my case it was a cheap piezo disc. These cheapos don't have an impedance that high, mine measured 23nF (138k @ 50Hz). In my case the piezo was installed in a solid body guitar mounted on the guitar body in the bridge pickup cavity. So you have to decide for yourself how this applies to your project.

I had to boost the piezo signal about 30 times to match the signal of my high output humbucker.
I got a very decent sound but I was getting a lot of noise when rubbing the guitar body. I haven't pursued my project but if were I would take only the high end of the piezo and mix it in with the humbucker sound.

[Gus]

Quackzed
You can pad the input of a jfet with a cap

An overbuilt buffer sim

[PRR]

> piezo was installed in a solid body guitar mounted on the guitar body in the bridge pickup cavity.
> I had to boost the piezo signal about 30 times
> getting a lot of noise when rubbing the guitar body.

You have miked the solid body, far-far from where the strings hit it. So of course you have more body-bump than string-sing.

Vibration pickups should favor the bridges. Usually the bridge is loosened and the piezo slipped under one end.