Re: Can I make my own active pick-up?

kindyroot wrote:
Sorry Cliff i sent my last post before to read yours, can you please
explaine a bit why the fet is not suitable here? i recognise that i'm
seeking a way to avoid the common-base option because i never got to
make a common-base work in my whole life, i get always lost in the
calculus ... maybe if i find some good schematics, if i don't i think
i will replace the 741 with something better, let's say an opamp with
fets in the input for example, is that ok? i found these in a
catalogue: LM387 and LF351, what do you think about them?
and finally if you have an idea about the tone shaping that i want to
do and if you know how it's done please let me know.

First I'll deal with the pickup itself, so you understand what you're working
with and trying to mimic.

0.4mm wire is about 24AWG, which has a DC resistance of about 0.16 ohms/meter.
You've used 250 turns of about .2m each, so you have a DC resistance of around
*8 ohms* - one thousandth of what a typical p/u will have.

In addition, the inductance of a coil rises with the square of the number of
turns - because every turn couples with all previous turns to increase the
inductance. The same thing happens with capacitance, each turn introduces some
self-capacitance with every other turn. The use of scramble-winding vs tight
winding affects this also. Anyhow, when you get 5000-10,000 turns as in a
normal p/u, the inductance and self-capacitance produce a resonant peak that's
proportional to 1/sqrt(L*C), which means the frequency drops with the square
of the number of turns. For the same copper mass, the number of turns drops
as the square of the wire diameter, so you have a resonant freq that changes as
the 4th power of the wire diameter. With a low resistance, the peak will be quite
sharp, but at 8KOhms, the peak is very broad (low Q factor). This explains the
behaviour of normal pickups.

Check out <http://en.wikipedia.org/wiki/RLC_circuit> for more info.

Now, for the pre-amp...

A common audio preamp might have an input impedance of 50Kohm, but a piezo is an
extremely high impedance output - 1Mohm input impedance is more suitable. More
actually, but them you run into other noise problems (Johnson noise), because
you have so few electrons flowing you start to "hear" them individually.

For minimum noise, you need an amplifier that has an impedance somewhere near
your pickup's output impedance. That's why I mentioned a common-base circuit.
If you employ 20-1 emitter degeneration (use a 150 ohm resister in series with
the p/u), and feed negative feedback from the collector to the base as normal
to set the desired gain, you should get a very linear low-noise circuit.
Best to avoid capacitive coupling on the pickup at these impedances, they'd
need to be big electrolytics. I'm not a purist that avoids electros in all
audio - they can be quite acceptable - but since you have a preamp circuit
that's so closely coupled to the P/U there's no reason to avoid it. If you want
DC isolation with a response down to say 20 Hz, into a CB setup with an input
impedance of 75 ohms, you need a C around 100uF (hence electro - but don't use
a tantalum).

You probably only need a gain of 20-50 before you feed the signal into tone/volume
and further amplification. You can massage your signal any way you like from there.

It probably doesn't make sense to try to reproduce the resonance at the pickup
by adding parallel capacitance with a series resister... but that's something you
might try anyhow, if you can rig up some way to measure the response curve.

Clifford Heath.
.

Relevant Pages

  • Re: 250mA voltage clamp
    ... >>And in answer to Jim's question: it takes about 300mA draw for the internal ... >>impedance of the supply to drop the voltage sufficiently. ... >>a shunt regulator if I can, so as to avoid unnecessary heating. ... >load would solve all your problems at 80% efficiency. ...
    (sci.electronics.design)
  • Re: 250mA voltage clamp
    ... >And in answer to Jim's question: it takes about 300mA draw for the internal ... >impedance of the supply to drop the voltage sufficiently. ... >a shunt regulator if I can, so as to avoid unnecessary heating. ... load would solve all your problems at 80% efficiency. ...
    (sci.electronics.design)
Loading