IOM Motomouth 2017 rev 5
OPERATING MANUAL
1 | P a g e
IOM Motomouth 2017 rev 5
SAFETY FIRST!
Motomouth has been designed to be a very safe (and reliable) product. However a few things
need to be considered for the safe and continued use of the product:
DO NOT power Motomouth with more than +/-15 VDC. Use a +/- 12 VDC (preferred) power
supply that has a current output of at least 100 mA and one that is able to power capacitive
loads (e.g. 10uF). Linear or Semi-linear power supplies are preferred due to their low noise
injection and fast/smooth reaction to changing loads.
Motomouth is an ELV (Extra-Low-Voltage) rated product. However, it is also good practice, for
all ELV modules, to use only approved isolated (SELV) mains power supplies in the Eurorack
enclosure: Look for any of these types of recognisable safety symbols on the power supply
and/or power supply transformer:
DO NOT power with 12 or 24 VAC power – This is ‘alternating current’ and it could stress the
internal capacitors and/or create hum.
Operating
DO NOT operate in the rain or in very damp conditions - keep Motomouth dry. (You won’t get
electrocuted, but galvanic corrosion may take place if powered and the PCB is wet). Operate at
between -10 oC to +40 oC, <95% Relative humidity (non-condensing i.e. no dew). This will also
apply to other modules.
DO NOT over-tighten the screws, nuts and stand-offs provided.
DO NOT over-turn the control knobs (they are analogue potentiometers, which have ends stops
at ~300 Degrees rotation) or over-push the toggle switches.
Cleaning and the environment
DO NOT clean with solvent based or abrasive cleaners. Use a lint free cloth or a damp cloth
with mild detergent. Solvents may affect the silk screen paint.
DO NOT dispose Motomouth in ‘land fill’ – the majority of the components are recyclable and/or
can be reclaimed and reused.
The PCB May contain sharp edges and sharp points. Keep away from young children.
2 | P a g e
IOM Motomouth 2017 rev 5
THANK YOU FOR CHOOSING MOTOMOUTH
FOR YOUR EURORACK SYSTEM!
Thank you for choosing to ‘back’ Motomouth through the Kickstarter campaign, and for bringing
the project to life. Also, thank you to those that missed the campaign, but still wanted to back
the project.
Motomouth has taken a since October 2016 to realise, and I do hope that you really enjoy using
it as much as I have enjoyed designing and making it.
IMPORTANT NOTE!
There will be the overwhelming need to immediately unpack it, connect a monotone squarewave into it, then scan back and forth through all the vowels for some time, to a point where it
becomes annoying. I know this, because that is exactly what I would do.
BUT, simply consider how you would use you own ‘voice box’, and ‘mouthing’ as many vowels
as you can over a few minutes (using a monotone voice) would also have the same outcome for
any listener. If you realise this, then you will quickly become an expert at using Formant
Filtering.
Thank you.
Ren.
Designer and creator of Motomouth.
A big thank you to Patrick Schisler and John Jacobs for ideas, proof-reading, and assistance
in order to complete this manual.
Not to forget, a big thank you to all you terrific backers for help with designing the panel
graphics (and to George, my main guidance throughout the process – even down to the Jack
positioning). And to Ben Hex for some great PR coverage.
You, the terrific MW community and many others have all helped greatly with this project.
Finally, a thank you to ‘slic’ (Sean) and ‘rhythmdial’ from the MW forum for making the project
launch a great deal easier – as well as ‘slic’ testing out some parts and prototypes for me.
3 | P a g e
IOM Motomouth 2017 rev 5
FORMANT FILTER BASICS
F1 F2 F3
Motomouth is a voltage controllable analogue formant filter, which uses just one control voltage
range to form the different vowels. There are 16 IPA type vowels to choose from.
Formant filters are different from lowpass (LP), highpass (HP) and bandpass (BP) filters in that
they comprise at least 2 bandpass filters (F1 and F2) with very narrow bandwidths (generally
fixed at around 60 to 100 Hz). F1 and F2 are moved to different resonant frequencies to mimic
the resonance changes in the ‘mouth cavity’ when forming the different vowels. A third
bandpass filter (F3) is preferably added to introduce the ‘nasal chamber’, but it’s generally kept
static.
The 3-bands F1, F2 and F3
Sedra & Espinoza BP Filter
The Sedra & Espinoza filter design, often known as a Dual Amplifier Bandpass (DABP) filter, is
perfect for formant filters, because the gain, at any quality (Q) or frequency (f) value, is always
2, and the bandwidth stays relatively static at each different frequency. These filters can have
very high Q levels (basically, the sharpness of the point), even up to around 120. In
Motomouth’s case, the Q value varies from 5 to 25, depending on the frequency.
4 | P a g e
IOM Motomouth 2017 rev 5
BLOCK DIAGRAM
5 | P a g e
IOM Motomouth 2017 rev 5
POWERING UP
250mm
10-pin
(Module)
16-pin (Rack
power-bus)
Red line denotes Minus
Twelve (-12) Volt end.
10 pin header
-12V
+12V
0V
Red line at the bottom
Connected.
Correct orientation to connect
to the rack’s power bus
If you have this connected the
wrong way round – no damage will
occur, the module simply won’t
operate. When correctly powered,
the RED LED will be on.
The power cable
Included is one Thonk™ 16 pin to 10 pin, 250mm connector cable for +/- 12V power.
Connection
When looking at the back of the module, the 10-pin power connector is on the bottom right (as
seen in illustration).
6 | P a g e
IOM Motomouth 2017 rev 5
TRIM AND ADJUSTMENT
INPUT GAIN
The audio input is calibrated to receive a 10V peak to peak audio signal. This is the ‘factory set’
value. If your input to Motomouth is higher and/or there is notable distortion (clipping), and you
can’t attenuate the incoming signal in any way, then the trimmer shown below, can be turned
anti-clockwise (counter-clockwise) gradually until the distortion disappears.
Should you need to use audio ‘Line’ levels (~4.5 V p-p or less, typically +/- 1.7V), and you
require higher output levels, then the trimmer can be turned fully clockwise (but remember the
initial trimmer position so that you can return to it later if you need).
7 | P a g e
IOM Motomouth 2017 rev 5
PATCHING
Sequencer
CV
CV
Gates
VCO
CV
Gate
Out
MIDI-CV
CV
CV
Gates
MIDI
CV Gate
MIXER
CV
Gate
In Out
VCFF
Note CV
Vowel CV
Vowel snap control option
Unaffected Sound source option
Affected Sound source
(Vowel)
Alternative Control Options
The advantage for modular systems is that you have a vast number of ways you can connect
and use each module, as well as the wide selection of module types to choose from. With
Motomouth, it’s best to consider it as a voltage and gate controlled formant filter (VCFF) that is
preferably used in parallel with the VCO train, which may comprise its own VCA and VCF, yet
sharing the same source VCO. The VCO-VCFF link may also have an insert for an additional
distortion module or perhaps a mixed noise source. Using a separate VCA train allows you to
activate the sound when required, and not have it running continuously.
However, you will soon find your own way to connect and use Motomouth, and the various
forums may be a good platform to share/discuss set-ups and patching – that’s part of the
attraction.
However, to get things started, here is a basic patch, with some options for snap control and
mixing the source sound plus the affected formant output from Motomouth:
Moto’
8 | P a g e
IOM Motomouth 2017 rev 5
THE CONTROLS (QUICK START GUIDE)
Increase or decreases the
‘slew’ between each
vowel, or the rate of the
filter sweep.
Turns on (or off) the
‘slew’ for F1 and/or F2.
Constant glide (rightselect) between vowels, or
it may glide and snap (leftselect) to that given vowel
whenever a gate is
activated. Analogy: Gated
Mono Keyboard.
Selects either vowel or
band-pass filter (VCF)
operating modes.
Filter F1, F2 and F3 Mixed
audio output (mono). See
block diagram.
Source audio input. Up to 10 Vp-p.
Blue LED Vowel code for
vowels 1-16, is coded in Binary
(LSB at the top). In VCF mode,
it simply sweeps, in dot mode,
up and down through the
frequencies (low at the
bottom, high at the top).
Manual adjustment of the
Vowel selection or the filter
sweep. This is inactivated
with a CV control Jack
inserted.
Red = powered.
Yellow flash = gate trigger.
0-1.25V CV input
(preferred) or Biased and
attenuated LFO input. This
deactivates the manual
control when a Jack is
inserted. See also page 11.
0-5/12V Gate or Trigger
input.
9 | P a g e
IOM Motomouth 2017 rev 5
MOTOMOUTH CONTROL CONCEPT
Time
(sec.)
1 2 3 4 5 6 7 8 9
1V
0.5V
0V
Slide
ON
OFF
ON
Morph
ON, mid-scale
Gate
ON
OFF
ON
F1
frequency
F2
frequency
Snap
Snap
The first Synthesizers for commercial use were monophonic and designed to be played from a
monophonic voltage generating keyboard. The keyboards basically used a ‘resistor ladder’ for
the keys to generate an oscillator control voltage for each note. This method, of course, was
adequate to play a monophonic oscillator. However, it was important to have a note-on gate, so
that envelope generators could be triggered for the various amplifiers and filters. Each time the
key is pressed, a gate voltage goes high (or low), where this is also used for setting things like
the sustain level. When the key is released, the gate voltage goes low, which is useful for
instigating the ‘Release’ part of the envelope. The gate trigger is simply a switch (or a change in
voltage detector) on each key that is made (or broken) when the key is pressed, and this is
configured as OR logic. As long as the note is held, the gate remains ON. When one key is
pressed, the gate is ON, when the key is released, the gate is OFF. However, when one note is
pressed, the gate is ON, and if another note is pressed without releasing the first note, the gate
remains ON. So you can see that the envelopes will not be retriggered, and it remains at the
Sustain point in the envelope, for this type of transition. This is a particularly useful method for
adding expression into the music – by default. If you want to trigger the envelopes on each key
press, then simply play each note with a gap (OFF) in between each note.
Motomouth uses this principle too, in a rather useful way, because it uses a gate trigger (leading
edge) to bypass the morphing function, so that is ‘Snaps’ to the current input voltage (CV).
Having a slewed ‘morph’ to each vowel makes the transition sound more natural and realistic.
However, from one vowel that may be off, due to the VCA being off, as soon as you go to
another vowel, you often don’t want it to slew, you want it to instantly start at that vowel, and this
is achieved by the ‘snap’ function.
Timing example
10 | P a g e
IOM Motomouth 2017 rev 5
LFO and CV Input voltage range and bias
(Vowel and Filter mode)
LFO NOTE: IMPORTANT
Although the BP filter (VCF) is mentioned as being controlled by a biased and attenuated LFO,
it is not necessary to do so.
You can play the notes to glide though the filter range, or you can play two notes, where you
want the filter to span, and add portamento or add some Morph time in, using he Morph control
knob (+). This way, you can ‘play’ the filter in a more musical way. Using the F1 or F2 morph
switch to turn the morph off for F1 or F2, can also get some interesting results.
11 | P a g e
IOM Motomouth 2017 rev 5
SNAP EXPLAINED
GATE (or Trigger)
Internal Pulse (LED Flickers yellow)
Rising edge only
Falling edge.
Nothing happens until the
next rising edge.
When a Gate signal is applied to the Gate Input, and the Mode switch is selected in Snap Mode,
the rising edge is used to trigger an internal pulse, which discharges the ramp (Morph)
capacitors whose ramp rate is governed by the Morph Potentiometer value. The discharge
means you then snap to the exact input voltage applied at that time, which is the intended vowel
or filter F target. It’s complicated to describe the function, but the effect is useable in both
Vowel and Filter (VCF) Modes. This can be applied for some very usable effects.
12 | P a g e
IOM Motomouth 2017 rev 5
IPA VOWEL TABLE
On (Blue)
Off
FILTER MODE LEDS
Low f High f
VOWEL CODE LEDS
(See note 1)
NOTE:
1. The IPA vowel locations are for sequence guidance, rather than exact positional location due
to analogue potentiometer non-linearity. Exact location is read from the LEDs.
2. The CV voltage setting is not as critical as it is for note tuning, and there is some dead-band
between the vowels so that aliasing (jitter between vowels) is eliminated. It is quite easy to set
up vowels using a span of 0-5V, and even possible with a span of 0-12V. A span of 0-2V is
preferred, and a span of 0-1.25V is perfect.
13 | P a g e
IOM Motomouth 2017 rev 5
Tolerance and dead-band table for Voltage Input
Vowel
Ascending
Ideal
Descending
mV
mV
mV 1 0 0 39 2 63
83
125
3
148
166
212
4
231
250
295
5
315
333
379
6
396
416
458
7
481
500
540
8
564
583
627
9
652
666
714
10
738
750
795
11
820
833
879
12
901
916
964
13
987
1000
1046
14
1072
1083
1133
15
1154
1166
1205
16
1237
1250
1250
ALL Notes above D# remain the same.
C to D#
0 – 1.25V
(16 semitones)
(Test Data)
Keyboard range
14 | P a g e
IOM Motomouth 2017 rev 5
Note for MIDI users
0
500
1000
1500
2000
2500
3000
0 5 10 15 20
Hz
Note Number
Some octaves are ‘labelled’ differently, i.e. C0, C1, C-3 and so on. This also applies to control
voltages as there is no specific ‘octave’ standard, other than ‘Volts per octave’. Therefore, you
may have to use the octave selection button to find where the active CV range begins and ends,
but this will always be from 0 V to 1.25 V in real input voltage terms for Motomouth, and it still
uses the 1V/octave standard i.e. ~83.33 mV/semitone.
F1 (bottom) and F2 (top) frequency plot against note number (or voltage).
Note 1 = 0V, note 16 = 1.25V:
15 | P a g e
IOM Motomouth 2017 rev 5
TEN TIPS FOR GETTING THE BEST OUT
OF FORMANT FILTERS
1 Don’t overuse it
The desire to use it continually overpowers the need to use it sparingly and effectively*.
Some reference examples on how to use Motomouth, in a very impressive and attention
grabbing way, can be found below in these rather amazing examples (the intro’s are
long, but they ‘build’ up really nicely):
https://soundcloud.com/street_ritual/detox-unit-systems-offline-1
https://soundcloud.com/sixis-3/wandering-deeper
https://soundcloud.com/swamp-music-1
https://soundcloud.com/marc-adamo/marc-adamo-sublow-spazm
https://www.youtube.com/watch?v=phpFvOsatPs
https://soundcloud.com/marc-adamo/product01-ready2rage
* Of course, continuous use to change the character of a given sound is quite effective.
But this should be limited to a one or two vowel changes. Continuous use through a
‘speech’ phrase is necessary of course, but try not to over-repeat.
2 Change the pitch, and bend notes
Changing the pitch and bending (portamento/glissando) some occasional notes is how
we speak (and sing). So for interesting sounds, not just for speech, it’s great to add
some note changes and/or bends.
3 Chose the best input, and keep it low
A saw-tooth wave closest resembles a human voice-box, but you can enhance this by
adding in some more High Frequency harmonics i.e. sounds similar to a Clavinet type
sound. ‘Zipper’ type sounds are terrific. FM ‘reso’-bass sounds are often great too.
Using Square-waves are not ideal, but they can be used to sound more robotic – try to
shorten the duty cycle of a pulse wave (10 to 20%), and sometimes, using PWM can
sound quite good. Avoid Triangle or Sine waves – they don’t really work, because they
have very low harmonic content (i.e. they can’t ‘kick’ the resonators into action). Lower
frequencies work best, because these will be below the resonant frequencies (~200 Hz).
4 Alternate between snap and glide (morph)
A mixture of Gliding and Snapping thought a piece can add unique interest, because this
is quite a new concept that’s not been used before.
16 | P a g e
IOM Motomouth 2017 rev 5
5 Add in plosives and sibilance
Adding or mixing in ‘puffs’ and ‘essess’ at each relevant vowel, at strategic points, can
make Motomouth ‘talk’ really quite convincingly – and even ‘sing’. However, this will
take a lot of your time to perfect – but the result could be quite staggering. These are
best mixed in ‘post Motomouth’, and could be sourced from a sampler for example.
6 Mix with the original
Motomouth does not have a mix control for the unaffected input sound. This is because
you may not actually want it to operate continuously, or you may want to process the
input sound quite differently and post mix it (or have it mixed in un-distorted, and so on).
However, mixing it eventually with the original sound can add to the subtle use of the
effect, where you have the option of automatic control of both with differing filtering and
amplifier shaping for each.
7 Try different transition vowels
There are a lot of vowels to choose from, where some vowels may sound very similar.
However, the ‘similar’ sounding vowels are often useful as a substitution transition vowel
(one type of vowel inserted between the first and last vowel), which can make a
difference to the ‘accent’ or ‘pronunciation’ of the phrase.
8 Try not to fast transition vowels through a signal
step point - i.e. avoid the ‘Blip’
Changing resonant oscillation frequency mid-flight is, and is heard as, distortion, and it’s
at that same frequency as the change – in other words, it can be heard as a blip. This is
a natural problem for discrete modular systems that have no means for total
synchronisation. This doesn’t always happen, and in the mix, it’s not really noticeable.
It’s less pronounced with closer vowels too. When using VCA’s and VCF’s, then the
issue disappears. It also disappears with gate operated wave triggers. A cure for the
occasional time you come across this, is to apply a small amount of morphing to a step
sequence, so that the vowels transitions fast enough to be effective, but slewed enough
to eliminate the ‘blip’.
9 Double-track, shape and affect
Double tracking and panning left-right will add real depth and surprise to the sound.
Alternating between mono and stereo with formant filtering can enhance this effect.
Also, try to shape the sound with VCA’s and VCF’s, and add in reverb or delay
(preferably stereo delay).
17 | P a g e
IOM Motomouth 2017 rev 5
10 Explore other features
Of course, ‘speech’ is not all that Motomouth can do – you can make it sound like a
modem or a telecom component, add some initial attack resonance to a 24 dB/Oct LP
filtered bass line, or even keep it static to obtain that perfect timbre that works best for
that mix.
18 | P a g e
IOM Motomouth 2017 rev 5
PRODUCT SPECIFICATION
Depth: 42 mm from base of front plate to tip of underside power connection
pins.
Width: 60.6 mm (12 HP). PCB = 60.0 mm (central).
Height: 128.5 mm. PCB = 100 mm (central).
Panel: 2 mm thick, Silver Anodized Aluminum finish.
Silk screen: Black.
Power connection: 10-pin universal (not keyed) header, ±12 Volt and 0 Volt
connections. 2.54mm pitch.
Reverse Polarity Protected.
Consumption: +12V, 80 mA max (0.96 W). -12V, 4 mA. (1W total).
CV Input (Vowel Mode): 0 Volts to 1.25 Volts at 1V/octave scaling for Vowel 1 to Vowel 16.
±12 Volt tolerant.
CV Input (Filter Mode): 0 Volts to 1.25 Volts ~200 Hz (1.25V) to ~2 kHz (0V) respectively.
±12 Volt tolerant.
Gate voltage: 0V = OFF (<2.5V).
5V = ON (2.5V to 12V).
±12 Volt tolerant.
Input sensitivity: 0.1 Vp-p to 10 Vp-p.
±12 Volt tolerant.
Output: 10V p-p max (depending on resonance).
Short circuit tolerant.
Filters: 3 x classic Sedra & Espinoza Dual Amplifier Bandpass Filters with Q's
set to between 5 and 25 and F set to between 200 Hz to 3,000 Hz.
Signal to noise: 60 to 66 dB (10V p-p input).
Resonance resolution: Analogue.
Input/output Jacks: Monophonic, 3.5mm Jack*.
CV and Gate Jacks: 3.5mm*.
System Voltage: 5 VDC, Linear, regulated. +/- 12VDC for the output amplifier.
No switch-mode power supplies are used.
*Tested during R&D for 500 operations, with no degradation.
19 | P a g e
IOM Motomouth 2017 rev 5
EU CONFORMANCE
Self-declaration
Health and the environment
All component and metal casings/fixings are Lead-Free and Cadmium-Free. All electronic components
are conformant and ISO 9000/1 regulated.
The Low Voltage directive
The device is designed for, and operated within the Extra-Low-Voltage limits, so the Low-Voltage
Directive will not apply. The power supply will come under the Low Voltage Directive, but this item will not
be supplied with this product. However, advice on selecting a suitable and conformant power supply is
given in this manual.
EMC
There are no LC switching components or switch-mode circuits used in this product, and the
microcontrollers are each decoupled with a 10uF electrolytic capacitor and a 100nF ceramic capacitor,
where this is further blocked by the incoming linear regulator, which is tantalum capacitor decoupled
(10uF). All PCB trace designs and decoupling for each component is followed in accordance with the
manufacturer’s recommendations. The PCB comprises segregated digital and analogue ground-plains
connected in a ‘star’ topology.
Purpose
Motomouth is classified as a ‘Musical Instrument’, and normally sub-grouped with ‘Synthesizers’. It is
intended to be used for recreation and entertainment, where its functionality is demonstrated through the
various examples on the web site.
It is designed to be a high quality and reliable product when used in accordance to this
instruction/operating manual.
Not suitable for young children (as with all electronic PCBs): It may have ‘sharp’ edges or points (i.e.
graze/cut/pierce hazard). When installed in a rack, these hazards are generally inaccessible.
Most importantly – have fun!
END
20 | P a g e
Loading...