Schematic and description of new octave/heterodyne fuzz, the "Subfuzz"
morris....@gmail.com
preamp which incorporated an analog multiplier to generate the square
of the input signal and mix it in, followed by filtering to enhance
bass - but that was and is a clunky gadget. It required a split 24
Volt supply, used precision op amps, was basically a clean but
nonportable device that plugs into the wall. Decided that the effect
was cool, and needed to be redesigned for economy, simplicity, and use
of a 9V supply. Now it's working and I figured I'd write it up.
While maybe I should keep my designs 'secret' and try to profit, I
don't honestly see the point, as I am starting a new job soon and
won't have much time for hobbies and side projects. So schematics and
description follow.
This effect basically IS an octave fuzz. But it's got a very smooth
transfer function, not a typical fullwave rectifier circuit, but an
approximation of a true voltage-squarer. This prevents many of the
'trashy' harmonics that typical Octavia-type units produce when chords
are played, and allows smooth isolation of the low-frequency
components. So you can play, say, an C# and an E and have a nice
smooth low A come thrumming out the speakers to accompany them, or
play an Eb and an A together to get a low B . . . it can be used for a
synth-bass sound, basically, entirely different from a usual octave-
down pedal. Single notes have a strong octave-up component as well,
and you can get that Hendrixy vibrato squeal if you want. It's like
an Octavia only smoother and warmer, I guess that's the description.
Schematic is here:
http://img172.imageshack.us/img172/3917/subfuzzge4.gif
The prototype - well, I don't want to show the circuit board cuz it's
ugly. Didn't etch a PCB for this one, just used a RadioShack
'universal' pcb board. Good lord, those things are a pain. Mainly
cuz you only get 3 holes for each pad. But here it is, from the
exterior, living in a Radio Shack project box. Very plain. Unlabeled
Mix, Tone, Drive, Level knobs but hey I know which is which:
http://img187.imageshack.us/img187/390/subfuzzboxqz4.jpg
Sound files - I will record some soon. But basically to get the idea,
you'd have to play it. It does generate bass from diads and chords,
it's output easily goes down as far as a bass guitar, but the feel is
just something you have to get used to. I have showed it to people in
person, and it does take a few minutes to get used to it. Fun,
though!
I think the circuit is pretty elegant, at least for a hobbyist
project. In my previous preamp project got a nice voltage-squaring
effect from an AD633 multiplier IC, but those things do need a lot of
volts to be happy, to the point where 2 9-Volt batteries are still
only marginally enough. They are expensive items too. This one uses
- just 2 germanium 1N34 diodes. Much cheaper way to get an
approximate square of the input voltage.
Only 2 cheap ICs are used for this project, an LM324 quad op amp and a
TL082 dual op amp, both available at Radio Shack. In fact, every
single part in this is available from Radio Shack, except for - the 2
1N34 germanium diodes. Thankfully I had a couple left over from a
previous thing, mail-ordered. There's something sickeningly ironic
about that, Radio Shack no longer carrying them - those are THE main
component of a crystal radio, THE classic Radio Shack project.
They're not expensive, not especially rare, I have no idea why Radio
Shack discontinued them. I don't understand Radio Shack.
The LM324 is a neat op amp because it doesn't have a problem with
single-supply operation - doesn't mind having it's inputs slammed up
against the rails - stable as a brick most of the time. In the upper-
left hand corner of the schematic you can see the first of the 4 LM324
stages, wired up to supply 1/2 the supply voltage as a 'ground'. By
the way, this makes SPICE throw tantrums, if you try to simulate this
circuit and set 'ground' to be an op-amp output, so you'd have to set
'ground' to be the '0V' line and just use a different label. I like
calling it 'ground' because the output of that op-amp is the voltage
reference for everything else.
You can also see that the negative terminal of the battery is
connected to the board (and also case ground, not shown) by the
connection of a mono plug to the input, typical effect stuff so that
the battery won't run down unless the effect is actually plugged in.
The 9V and 0V supplies must be connected to the LM324 and TL082 ICs of
course, but I didn't bother drawing that.
The second LM324 stage is just a unity-gain buffer for the input
guitar signal. I don't want gain here, because the voltage-squarer
circuit would just become a full-wave rectifier if overloaded. I want
the smooth square function, no need for overdriving too much here.
Guitar level works just fine. This buffered signal is then sent
through a 47K resistor (to limit clipping/overloading that might
happen if the signal goes over 0.3 V) and fed to the two 1N34
germanium diodes. These diodes, when biased below their 'conduction
threshold' of 0.3 V positive bias, do allow small currents through
that are roughly proportional to the square of the input voltage.
Each diode can only handle one half-cycle of the input voltage,
positive or negative.
The TL082 dual JFET op amp is used for current-voltage conversion due
to it's very high-impedance inputs, which allow accurate amplification
of the small (microamp or nanoamp) currents through the 1N34 germanium
diodes. It's not a great op amp for general purpose single-supply low-
voltage operation, it can sound very bad if both inputs are driven to
the supply rails, but it's great for this stage of the circuit.
I tried ordinary 1N914/1N4148 silicon switching diodes first, but they
didn't work well. Very, very noisy, too much noise compared to the
actual signal current. Possibly you could use some fancy Schottky
diodes here, but germanium 1N34 does sound good.
The third LM324 stage is a differential circuit, with moderate gain,
and combines the positive and negative half cycles from the 2 TL082
stages. The output of this stage is the square of the input signal.
It is blended with the uneffected Dry signal by the Mix control.
Following this, there is a high-pass filter composed of the 47n
capacitor and 100K resistor which serves to remove DC components and
subsonics generated by the squaring operation. Next is the Tone
control, which has a variable high-frequency cutoff. The 4.7 K
resistor prevents instability of the third LM324 stage if the Tone is
turned all the way up, op amps don't like to drive a capacitive load
without SOME series resistance, it would oscillate ultrasonically and
make some bad noises.
The final LM324 stage boosts and clips the output signal, with the
Drive control serving to set gain. The 68p capacitor just smooths off
the clipping the tiniest bit, from the 2 1N914 silicon diodes -
honestly the level could be a bit higher. I should have used LEDs
instead or maybe 4 1N914s in 2 series pairs, and maybe the gain could
be pushed a LITTLE higher. It's nice to have a gain stage here so
that you can crank it up and smooth out the output if you like, not
lose sustain like you naturally would from squaring the signal, and
emphasize the bass further if you've got Tone rolled down.
What's left is just a capacitor to remove the 4.5 V voltage offset,
and a level control. The 100 ohm series resistor, again, is to keep
the last LM324 stage from becoming unstable if you try to drive a long
cable at full level. Op amps don't like to see capacitance at their
outputs, they just get mean, at least I've heard them do that. Oh
yeah, and a SPDT stomp switch. I know that it's not a 'true bypass'
because the input stage still loads the guitar signal but it's not
that big a load. If it bothers you, maybe make that first 470K
resistor a Meg or 2, and if you'd like cut that initial 47n capacitor
down to 22 or 15.
Guess that's my report. If anyone has suggestions or comments, I am
always happy to hear them. Hope everyone is working on cool things
and new sounds,
morris
underwoodblog
Looks a little like an ampeg scrambler with OPs and an and extra
distortion unit.
greets
Jürgen
morris....@gmail.com
On Jul 25, 4:48 am, underwoodblog <underwood.de...@googlemail.com>
wrote:
Yeah, kind of sounds like this Ampeg Scrambler sample below, but less
noisy, more whooooomy type bassy.
http://www.tonefrenzy.com/sound_files/ampeg-scrambler.mp3
I guess I gotta record it soon.