12

Basics 

Introduction

To understand the synthesizer techniques, either the analogue or digital synthesizers, you need to go back to the basics.
With basics I even do not mean the musical meanings of the synthesizer philosophy. You have to go back to the basics on a level of the principals of VOLTAGES.
You might think that Voltages are only to be there for an engineer to understand but nothing is less true than that.

The whole idea of the realization of the philosophy of the analogue synthesizer was to VOLTAGE CONTROL each parameter of each unit. Of course, first there was the idea to create various sounds by using the principal of subtractive and additive synthesis. This means that the base of a sound is there and by manipulating (removing or adding 'parts') of this constant base, the final sound is created.
In case subtractive synthesis, a so called VCO (Voltage Controlled Oscillator) produces a tone with a constant amplitude (volume) over time and a fixed amount of upper tones. By using a VCF (Voltage Controlled Filter) the upper tones amount can be dynamically decreased to initiate the timbre of the sound. With a VCA (Voltage Controlled Amplifier) the volume curve of the tone can be arranged and thus by using a keyboard to control this at command it may sound as a musical instrument. In the case of additive synthesis several wave forms are add dynamically to end up in a final wave form. Either way is achieved by tying up certain modules which are all voltage controlled.

An analogue voltage - Digital (only counts a LOW and HIGH level

Voltages

All of this is based on VOLTAGES. Either analogue levels of voltages or logic voltage jumps. Analogue voltages may be an automatically repeating wave form or a constant voltage level which can change to another constant level by a manual action (turning a control pot) or on automated commands. Digital voltages are used to initiate and control the events in sense of a musical scenario. When a key at the keyboard is depressed a logic (digital) level is produced to command a certain process to start - or to stop, if you wish because this all depends on how the commands are used.

With a modular synthesizer (either analogue or digital) you have the freedom to decide how certain parameters of your setup are controlled by certain events. When you do not know what sticks behind this on a Voltage level in conjunction to the different building blocks of a synthesizer then, you easily will fixate a certain possibility of using signals as being the reason of a certain output sound or effect. But, the result of the sound only might be so because a certain combination of controls is used in a patch with certain units and this does not mean that all of the controls are also only meant to function as such.
It does not matter whether we speak about analogue or digital synthesizers because, at first instance, the digital synthesizer was nothing more than an imitation of the analogue one. The designers have tried to imitate all the functions with digital electronic principals. This digital development has lead to functions which before did not exist in analogue synthesizer techniques but nevertheless you might understand how they work on a basic level and thus can give you ideas in how to use the possibilities of your devices beyond manufacturer standards.

Glissando (Example easily to understand)

Each synthesizer has a so called PORTAMENTO effect to have the notes you play on a keyboard slide to each other. What actually happen on a VOLTAGE level thing is that the jump from one voltage to another (one tone to another tone because the tone-height is determined by a Voltage Controlled Oscillator - tone generator) is not made abrupt but sliding from one to the other level according a certain time setting.

To achieve this effect, the abrupt jumping voltage need to be changed in a charging and discharging timeline from one voltage to another. This we can call the SLOPE time or SLEW time. The time set that is necessary to flow from one voltage to another. The result: a Portamento effect when you play the keyboard.

This effect you know as it is and you do not think about how it is achieved in context to treating a Voltage change which represents the change of the notes and their way to get from one note to the other.
I can tell you that the electronically treatment of the voltage which determines the tone is the same as done to achieve a BASS control function at an amplifier or to set the speed of a random voltage or to set the ATTACK and RELEASE parameters of an envelope generator. It only goes about the context of the input signal and the parameters input you like to control. Once you get the point of how voltages can be treated and function to control certain parameters you then understand a modular synthesizer and can use this knowledge to extend your creativity of using all the modules of a modular synthesizer or a complexes of devices together.

Time Effects (Another example)

Every musician know about the effects like: Echo, Doubling, Reverb, Phasing, Flanging and a special Phasing filter. What you do not know is that they are all based on a variation of a single parameter. You can buy for all of these effect separated devices or a digital device which has all in one. What you do not know that when you bought all these devices separately you actually have been 'cheated' and when you buy the digital all-in-one box you think that you have been presented a gift for little money.
The truth is that it is very easy to achieve all of the mentioned effects because they are all based at a single parameter and their dynamic change over a certain time period and this parameter is called DELAY TIME.
When you take a delay line which can deal both with short delay times and with longer delay times you can, by simply changing the delay time and/or the dynamics of the delay time control, achieve all of the mentioned effects.

Using a delay time of 50 Milli Second or higher doubles the sound. At 50 Ms this is called DOUBLING and used to make the news reader become better to hear at low volume without accentuate an effect (you actually do not hear explicit). When the delay time becomes longer it started to sound more like a single ECHO. The time between the original sound and the repetition of is becomes longer. When also the output is in feedback to the input, then the sound will repeat more than one time. Depending on the FEEDBACK amount this will generated certain ECHO effects.
Do exactly the same but now with a very short delay time (5...10 Ms) and you get a filter which softly changes the colour of the sound - depending on the feedback control which all of a sudden now is called "Q" or "RESONANCE". Vary the Delay time control constantly by using a LFO (Low Frequency Oscillator) all of a sudden turns the Filter into a Phaser and when in this setup the delay time is decreased to 50 to 100 Ms it becomes a Flanger. When you have a Flanger, you most likely have the possibility to switch off the LFO (the dynamically changing of the sound) and turn it into a filter. You know, when the manufacturer would add a switch to change the delay time range they also would have provided you with a PHASER. The same electronics - just a switch extra.
When we continue this story about delay based effects with this single electronic principal we can achieve many functions (effects as you like) by just varying a single parameter and its range and/or dynamics control over the time line. With the principals of a modular synthesizer everything should be and most of it is voltage controlled. Thus, the applied voltage to an input determines the effectiveness of it. By varying the voltage in certain ways over certain time lines, certain effects can be achieved.

Whether you understand these abstracts or not does not matter at all but do plant into your head that it all goes about the appearance of voltages and that you can use any voltage to control any voltage controlled input of a modular system. A way of working that invokes the so called Trial an Error method is very valuable and will lead to more success once you have tried to understand the basics. Even when you do not understand it as you should it still will be of value. Better a wrong fantasy than no fantasy at all.
Fantasy! This is what it all is about. You have it so you might as well use it also in context to some boring abstracts about voltages. You find out that voltages may become a fairy tail which you can write yourself. Join Alice.

Some theory about Voltages

About AC and DC you probably have heard. You find these terms written on the adapters you are using for musical devices. AC means "Asymmetric Cut" and DC means "Direct Cut". AC is a constantly varying voltage (as with the 110 V / 60 Hz supply from the mains contacts at your walls). DC is a constant voltage as from a 9V battery.
In analogue synthesizer technology we also know AC and DC voltages but we usually call them differently (frequencies (tones), control Voltages which are dynamic (AC) and static control voltages (DC).
An AC voltage is principally a voltage level which is changing over a certain period of time and a DC voltage one that is stable at a certain level.
When you set the Course frequency control of a VCO at a certain point, you actually set a certain DC voltage level. The frequency (tone) of the VCO is determined by a Voltage (it is voltage controlled). Now what happens when you vary this Coarse frequency control constantly by turning the control knob up and down? Of course the frequency will go up and down but, also you create manually an AC voltage because the Coarse voltage control is now varying on a periodically based rhythm. When you are changing the Coarse control very slowly you might question whether this could be called an AC voltage or a slowly changing DC voltage. It does not matter so much how you call it in this case. It is Voltage but, when you read a specification about a certain synthesizer unit which says that it is AC coupled it becomes important because an AC coupled voltage control input cannot deal with a constant voltage at the input. Instead only the variations of the input voltage will go through.
When designing the "TD" as much as possible all control inputs can deal with DC input signals. This means that beside responding to the change of a control voltage input (AC) also a static or very slow changing static voltage do have effect.

Frequency

When a voltage is varying periodically and thus can be seen as an AC voltage the repeating time of it is called frequency. Whether you deal with something you can hear as a tone or not does no matter. A voltage which is changing levels 100 times a second can be heard as a tone with a frequency of 100 Hz but when it is changing only 10 times a second you cannot hear it but still it is an AC voltage with a frequency. Thus, Frequency and Tone are related but we only call it a tone when we can hear it - when it is within the audio range. A frequency of 10 Hz or less usually is used to modulate and thus it often is called modulation-frequency. But, realize both are just nothing more than changing voltages.

Wave Forms

When a voltage is changing periodically, it does this with a certain curve. This we call the wave form. Your outlet (mains) for example is a sinusoid wave form because the changing curve is a sinusoidal one (Sine).
Also in context to a synthesizer we have wave forms. Not only a Sine is a base wave form. Also other base wave forms are available. Triangle, Saw Tooth, Ramp and Square are other base wave forms most synthesizers count.
When the frequency of these wave forms lay within the audio range and you listen to them you hear a difference in their appearance by the colour (timbre = the Harmonic content). A sinusoid sound the most basic (round). It has no upper tones (higher harmonics). A triangle sounds pretty round but has some odd harmonics. A Square has many (odd) harmonics prominently to be heard (think about an electronic organ). A Saw Tooth or Ramp wave form have the most harmonics - both odd and even and thus do sound very rich (think about the strings preset on your synth).

Sine
Harmonics: 0
Amplitude relationship: 0

Triangle
Harmonics: 1,3,5,7
Amplitude relationship: 1/9, 1/23 1/49

Saw Tooth
Harmonics: 1,2,3
Amplitude relationship: Exponential

Ramp
Harmonics: 1,2,3
Amplitude relationship: Exponential

Square
Harmonics: 1,3,5,7
Amplitude relationship: 1/3, 1/5, 1/7

When a repeating voltage - a wave form / frequency - is used to modulate a certain parameter (Voltage Control), then the wave form shape can be seen in another context - depending on which parameter of the sound is modulated.

Let’s take a simple example: modulating the frequency of a VCO. A VCO of a modular synth always has a so called FM input. FM means Frequency Modulation. When you apply a voltage to this FM input, the frequency of the VCO is modulated in function with the voltage change of the FM control signal. When your FM modulation signal (voltage) is a sinusoidal wave form with a frequency of approx. 7 Hz (Hertz) and the modulation depth is small, you can hear a vibrato effect. The tone is slightly going up and down - 7 times a second.
When you lower down the modulation frequency to 1 Hz and turn open the FM depth control, you get a sound like a siren.
Now, when instead of a sinusoidal wave form a Saw Tooth shaped form is used, you will hear another kind of serin. The tone will jump high abruptly and then goes down during one second (when modulation frequency = 1 Hz) and repeats this curve again.
When you reverse the Saw Tooth wave it becomes a Ramp wave form: the tone will rise during one second and then abruptly jumps down.
When you apply a 'normal' square wave the modulation will let the frequency of the VCO jump between two tones because a square wave only has two levels (low and high).

This all may look completely normal to you when you do have some experience with using a synthesizer - in particular a modular analogue one. But, ever thought about that the tone you hear from the VCO when attached to an amplifier basically is nothing more then, a repeating voltage - repeating itself following a certain curve of going up and down and that the modulation signal is basically the same but only with a different speed and possibly a different curve. The reason that the frequency of the VCO that you hear and the modulation signal only modulates is because you have chosen to let this happen as such. The generator which generates the tone you hear is called a VCO and the generator which produces the modulation signal is a generator which also can be a VCO (when this generator is also Voltage controlled). Although it often is called a LFO (Low Frequency Generator) or VC-LFO because it is voltage controlled, you simply deal here with two changing voltages. Only the use of them - their purpose - is put differently.
It is important to think about it for a while in the abstract manner of having only to deal with voltages, their level change over time (frequency and wave form). A synthesizer is nothing more than a bunch of voltages tight up together in a certain way to achieve several functions and because it is this simple you may have infinite fantasies about how to tie these voltage up together. That is the fun which finally should result in a sound you like to play with.

Whether we speak about a VCO, VCF, VCA, Sequencer, Sample and Hold, Noise, Amplifier or whatsoever in context to a synthesizer, all these units only generate and or modulate voltages and you tie them up with each other as you like to. Whether this results in a workable sound depends on how you tie them up and how much you understand of what you are doing.
When working with a modular synth is new for you, you should explore the basic experiments as just were described because it helps you to also understand the use of voltages in other ways.
Using, for example the modulation signal (wave form) for AM (amplitude Modulation) instead of FM gives you other effects while using the same voltages. The reason is that a VCA (Voltage Controlled Amplifier) does not let you control the frequency of a tone but instead the amplitude (volume when the AM modulated VCO output is routed to an amplifier).
If you understand enough about this you can conclude that when you AM modulate a wave form which is used to FM modulate a tone generator (VCO used as such), you then get another effect: a frequency modulation which varies in depth because the amplitude of the modulation signal is made smaller and larger by the AM modulation signal.

FM modulate the time period of a wave form - the frequency and therefore named Frequency Modulation (FM). The higher the FM voltage input, the shorter the time period and thus the higher the frequency.

AM modulates the amplitude. This is just another parameter of the wave from. Instead of the Time period (horizontally represented) it changes the vertically represented parameter: the TOP-TOP height of the voltage).

Logic Signals

Logic signals and levels are voltages which are used to control the processing of certain events of a module or a patch.

Think about the Clock signal of a Sequencer module which commands the step change or the GATE signal of a keyboard which commands the Envelope Generator to start producing an envelope wave.

A logic signal also is nothing more than a voltage but the essence of it is that it only knows two voltage levels: Low and High. Mostly the Low level is a zero voltage and a High level a voltage of 5 Volts but this is not principally and may be different (Low -5 Volt / High +5 Volt). To which values the Low and high state is determined is, in theory - not important. Important is to realize that a logic signal only has two states. Depending on the characteristics of the parameters to control by the logic signal, the response to a control signal might be activated by the Low level, The High level, both or at the change from Low to High or vice versa.

Trigger

When a logic signal is used to initiate the start of a process, then it is called a Trigger signal. Only the moment the logic level goes from Low to High state is seen as the moment to start an action.

For example: the Trigger output of a keyboard which starts an Envelope generator (ADSR generator).

Gate

When a logic level is used for to activate a certain state or process at the moment that the level goes High and keeping that state or process going for the whole time being of the High level, then we speak of a GATED logic signal.
Again, an Envelope Generator is an example. The GATE signal of the keyboard starts the Envelope Generator to produce an envelope wave form and when the generator has reached the SUSTAIN level (after the ATTACK and DECAY periods are finished), it will remain in its SUSTAIN level as long as the GATE signal is at High level. Then when the GATE signal goes to Low level again (because you release the keyboard key), the Envelope Generator turns in its RELEASE phase.

The difference between a Trigger and Gate signal also can be demonstrated with an Envelope Generator. Many analogue synthesizers do have a Gate signal from the keyboard to control the ADSR (Envelope Generator). Mostly you can choose for the ADSR with a switch how to have the ADSR respond to the logic Gate signal from the keyboard: GATE or TRIGGER (sometimes "trigger" is called "Single Shot").
When you use 'Gated' also the active High level duration of the logic signal is involved in the response. When you use 'Triggered' only the moment that the logic signal goes to High level is involved.
The practical difference with the example of the Envelope Generator is that when you Gate the ADSR, you will hear the voice as long as you keep the key in (when a SUSTAIN level is set) and when you Trigger the ADSR the Envelope wave will be produced without taking notion whether you have the key depressed for a longer time. Some ADSR modules then produce an ADR wave - thus without a SUSTAIN level/period - and other modules do have also a control to set the SUSTAIN time duration and thus also produce a SUSTAIN period automatically.
The way ADSR modules (Envelope Generators) deal with Gate or Trigger principals and how they respond to a series of Gate changes or Trigger commands during the process of generating an Envelope wave varies.
Also, when you have a keyboard which delivers both a Gate and a Trigger signal the difference is that the Gate signal remains high as long as one or more keys are depressed and thus when you play legato the Gate signal remains all the time high. The Trigger output however usually produces a trigger signal each time a new key is depressed - also when you play legato. Some synthesizers have a New Key Voltage detection output which also functions like this (to generate Multi triggering when playing legato).

Clock (CK)

Sequencer and S&H modules are controlled by a steady repeating logic signal that we call a Clock (CK) signal.
A Clock signal actually is a repeating trigger signal.
Mostly a CK input responds at the moment that the logic signal goes from Low to High level. Then is called "positive edge triggering".
The output of a Clock generator may be a pulse train of short pulse or simple a square wave with a duty cycle of 50/50 (equally high and low level periods). Here the response of the receiving module is more important in how you name the logic signal (Trigger, Gate, Clock).

Other ways

I have tried to distinguish Gate, Trigger and Clock logic signals only to point out the basic differences of the response of the module that is controlled. This does not mean that when a module has a control input named CK that you may not use other signals to control the CK input.

SEQ Keying

When you connect a keyboard Gate or Trigger output to control the Sequencer Clock, then the Sequencer will step forward each time you depress a key - depending on what Gate and/or Trigger signals your keyboard delivers this can imply that when you play staccato and legato combined, the Sequencer step changes by following only the staccato parts or also the legato parts.

S&H Keying

The same you can do to control the S&H CK input to generate random voltages to control a filter synchronized with your keyboard playing.

LFO as Clock generator

A Square wave output of a generator (fGen or mVCO or LFO) also can be used to CLOCK the Sequencer or S&H unit. In this case the generator is turned into a CK generator. In context to the TD the use might be that the mVCO for example is a voltage controlled generator (VCO) and thus implies a CK speed control by voltage versus the static (manual control only) control of the specific CK generators of the TD.

Clock Generator as Modulation generator (LFO)

The CK outputs of the TD's specific CK generators (for SEQ and S&H) also can be used as modulation generators - using the CK output to feed an AM or FM input. Or you feed the ENV generator of the TD with it in synchronization with the SEQ and the S&H module and then use the ENV output to AM or FM other modules.

RESET & HOLD

The Sequencer module also has a RESET and a HOLD input which are logic control inputs. This means that they respond to a logic level (Low and High). The RESET input is Trigger based and the HOLD input is GATE based. Both can be controlled by any signal which counts a zero voltage level (Low) and a +3.5V or higher level (High). Thus these control inputs are typically logic controls too because they start an action based on a logic level.
As a specialty of the TD, these logic inputs also can respond to analogue voltages. This means that the input stages detect whether the analogue input voltage must be seen as a logic Low or a logic High level depending on the voltage level. Roughly this means that when an analogue voltage exceeds the level of 3.5V the input stage sees this as a High level and when the analogue voltage goes down again below 2.5 volt it is seen as a Low level.

By the way: also the CK input of the Sequencer can deal with analogue input voltages.
Can this be useful? Using a triangle wave form from a generator to control the Sequencer would be the same as using a Square wave (logic wave from). In this case it only would be useful because you do not have a square wave available or not available to patch it to the Sequencer.
More interesting is using analogue voltages wherein the voltage level represents a certain stage of musical impact. A keyboard voltage level for example represents the position of the note that is played (monophonic keyboard to make it easy). A 5 Octave keyboard most likely has a keyboard voltage output which ranges a 0 to 5 Volts output. Zero volts for the most left key and 5 Volts for the most right key. Having in mind that the logic inputs of the Sequencer and the S&H unit can distinguish analogue levels to interpreted them as logic levels this could give the following result: Playing half way the 4th Octave at the keyboard will trigger the Sequencer CK input to make a step forward or causes the S&H module to generate a new random voltage. Once the keyboard voltage is seen as a logic high level due to the fact that it has exceeded 3.5 volts, you first would need to play a note at the keyboard in the middle of the 3rd Octave or lower to cause a low logic level again. Then playing a note half way the 3rd Octave or higher will generate a High level again and thus caused the Sequencer again to step forward. Now, playing with both hands a melody using the low and high half of the keyboard can generate interesting synchronized Sequencer and/or S&H patterns.
You have to experiment with this and do not take the voltages (keyboard positions) which I used as an example here too literal. First of all because I never tried this myself and secondly because the exact High and Low level detection of the logic input are not yet verified by me.
What I want to demonstrate here is (again) that understanding the basics of the signals that flow from one point to another point in a synthesizer patch are nothing more than voltages and voltage levels and you freely can fantasize in how to use them and not take so much the name the voltage or controls carry too literal. It all goes about the context of using the voltages within a certain patch.

VCO / VCF / VCA

The most common modules in analogue synthesizer technology are the VCO, VCF and the VCA.

• VCO : Voltage Controlled Oscillator
• VCF : Voltage Controlled Filter
• VCA : Voltage Controlled Amplifier

In modern digital technology they are often called DCO, DCF and DCA. The "V" of Voltage simply is replaced by the "D" of Digital which is related to the way the parameters of these modules is controlled (analogue Voltage versus Digital code).

VCO

A VCO is basically meant to generate the ground wave form(s) to either produce a tone or to produce a modulation signal to modulate other modules (vibrato (FM), tremolo (AM) etc..). When a VCO is configures/designed to produce modulation voltages in particular then is often is called VC-LFO (Voltage Controlled Low Frequency Generator). LOW because modulation signals mostly are usually of a low frequency which is below the hearing range of human beings. When such a LFO is not Voltage controlled, then it simply is called LFO.
The TD, as already was mentioned has a mVCO module (Modulation VCO). This simply is a deviation of the terminology VC-LFO and VCO to state the multipurpose aim of this module). The fGenerators of the TD are actually VCO's but the differences with classical designs did force me not to call them VCO's.

A VCO, VC-LFO, mVCO or fGen. Whatever they are called, what they do have in common is the fact that they produce a wave form which repeats itself automatically and the speed of this repetition depends on the applied Voltage at the control input and the manual control (Coarse and Fine controls).
Besides using these modules to generate wave forms to create tones to be heard in context to a musical piece, the signals also can function for all kind of other (modulation ea..) Purposes.

What the module VCO provides us is:

• Wave form (the graphical curve in which the output voltage raises and goes down again).
• Frequency (the repetitive rate of the wave form).

On top of these basics there are many ways to put it into a contextual behaviour with other modules. Either modules with another basic function or to have more VCO's combined with each other to produce complex wave forms (think about the FM synthesis with the once popular Yamaha digital FM technology wherein DCO's were combined to create a certain base sound - DX synths).

Wave Forms and Harmonics

The reason that the classical designed VCO's do have several output wave forms (Sinusoidal, Triangle, Saw Tooth and Square/Pulse) lays in the fact that each of these wave forms do have - from a musical point of view - a different ration from harmonic related frequency above ground tone (frequency) in the frequency spectrum range.

• Sinusoidal Ground tone only - no harmonics - sounding very 'round' (pure).
• Triangle Odd harmonics on to of the ground tone according a level ration of 1/6,1/25,1/49.... Resulting in a sound which is round with some 'upper tones'.
• Saw Tooth Odd and Even harmonics on top of the ground tone with an exponential level ratio resulting in a very rich sound (base for strings). Also a good base to imitate acoustic piano.
• Square Even harmonics on top of the ground tone in a prominent level ration (1/3, 1/5, 1/7). Sounds much like an electronic organ - quickly boring when a tone is held for a longer time - in contrast to a Saw Tooth as a base for a string-sound which sound more natural.

By mixing these wave form and/or modifying and filtering them first, many resulting wave forms can be created which all represent a certain base form to create a certain sound with.
Through time many technologies were developed to create complex wave forms which needed to serve - together with other treatment - for a certain sound.

It should be clear to you that the base wave from is an important starting point to achieve a certain sound. Using a sinusoid as a base is meant to create a 'round' and soft sound and using a Saw Tooth is meant to be chosen for a rich sound. Of course - using wave form shapers and or distortioners can turn a sinusoid into a very rich sounding wave from.

Trough time, designers of synthesizer technology have tried to extend the possibilities to produce other wave forms (more complex) than could be achieved by using and mixing the base wave forms.

The first technology to produce very complex wave forms was to Frequency Modulate (FM) one VCO with another VCO and having them both dynamically synchronize (correlate) in there frequency shifting - based on (for example) the keyboard position of the played note. This technology became popular in a digital form with the release of the Yamaha DX series. Nowadays everybody thinks that Yamaha invented the 'FM sound'. This is not true. Both the principal of synchronizing by logic means and due to the effect of a fast FM response of the VCO's involved were done before Yamaha came with their DX series.
First of all it was me, who long before Yamaha did it build my own modules, to be suitable for this idea. Then I discovered that Sequential Circuits had done something similar in a nice way with their Prophet T8 synth. Sequential Circuits achieved a similar sound by using the Sync method which they based on the possibilities of the Curtis chips (CEM chips) which did have the sync control invoked. I, first of all reached the 'FM effect' by designed my VCO's in such a way that they could respond very fast to FM signals (on audio frequency level) without degradation of the all-over performance. Then I did invoke the full scale of the sync possibilities of the CEM chips to explore 4 different ways of synchronization (hard, soft, positive and negative). Using 3 VCO equipped with these possibilities and route them to control each other gave a sound which made the Yamaha DX series as a laugh. Of course, the DX was polyphonic and mine was only monophonic so no bad things about the DX but from a wave form production point of view, the DX was not a revolution at all, unless you take the large audience for it into account due to possible mass production.

VCF

A Voltage Controlled Filter (or DCF when digitally controlled) controls another aspect of the sound. This is called the Timbre.
A VCF modifies a base wave form - coming from the VCO. The base wave from already has a timbre range but can be more precisely determined by using a VCF which is tracked by the keyboard. This is typical for a modular analogue synthesizer because this is mostly a subtractive synthesis. This means that things already in its basic appearance are there are selectively removed.
A VCF is basically meant to cut out upper tones (harmonics) from the basic complexes of wave forms and to do this dynamically in ratio with the position on the keyboard (the played key).

Because a VCF is voltage controlled, it becomes possible to - besides the keyboard tracking to balance the spectrum of tones to be played - to also modulate the timbre by other controlling voltages. Injecting a certain amount of Envelope Generator signal to the filter will also change the timbre curve dynamically when a note is played. Injecting a low frequency signal from a LFO or VC-LFO (mVCO on the TD) will modulate the timbre constantly in all kind of wave form shapes.
Whether you make the modulation effect subtle or dramatically depends on your aim but one could say that a VCF is one of the most effective modules within a synthesizer setup.

Many variations exist. In analogue technology the BP (Band Pass) type is most wide spread. Furthermore we can see LP (Low Pass), HP (High Pass) and Notch filters as modules involved. Some synths do have a filter that can deal with all of these characteristics by setting a switch or do have separated outputs for each of these filter functions.
Combining these filter functions is very much worth laying with. The designers of the KORG MS series did understand this very well. Instead of designing a single Band pass filter they designed two filters: A LP (Low Pass) followed by a HP (High Pass). Depending on the settings and dynamic Voltage control you could create many filter functions with the combinations of the LP and HP filter.
Although using filters is extremely fascinating to come up with a sound that is special, the TD does not have any filter modules. The reason is that this is such a wide territory that it could not be included within the front panel space available with a 4 height unit 19 inch rack. Thus, the TD concentrates on generating wave forms rather than to treat the timbre of the sound. This must be done by additional filter modules or by using the filters form other modular synths.

VCA

A VCA (Voltage Controlled Amplifier) is mostly used to give a dynamic volume change to the sound. In combination with an Envelope Generator - triggered by the keyboard- it does able you to let a played key appear with a certain volume curve.
More abstract: A VCA is an amplifier which has its amplification depending on an applied control voltage.
A VCA also can be used to control the amplitude of any input signal by the control voltage. Thus, a control voltage also can be used as an input and then controlled by another control voltage to determine the amplitude ratio of it (If the VCA can handle both AC and DC signals).

A LFO signal which is used to modulate a voltage controlled filter (VCF) can be routed through a VCA to have the modulation depth vary by the control voltage applied to the VCA. This control voltage on its turn also can be a LFO wave form. When the LFO signal to modulate the filter is 7 Hz and the LFO frequency which modulates the amplitude of the VCA output is set to 0.5 Hz then, the filter is modulated with an up and down going depth every 2 seconds.
Also, with this idea you could let the modulation depth of the filter depend on the position of the played key at the keyboard. The more to the right a key is played the more modulation depth or the other way around when the control voltage is reversed. In all modern digital synths you can find this option to scale an effect with the keyboard position.

VCA as 2-quadrant multiplier (half ring modulator

A VCA also can function (when designed as such) as a so called two quadrant multiplier. When using the VCA as such with audio signals you get the known "ring modulator" effects.

To achieve this effect both the signal input and the control voltage are within the audio range.
When using a external signal source - like a voice - and the VCA control voltage input is fed with an audio frequency from a VCO then the typical robot like and weird voices can be created by using your own voice.
Using for both the signal input and the control input VCO's which are related in their tonality scaling (both controlled by a keyboard for example) leads to a sort of SYNC effect. What it then basically does is modifying the timbre of a sound.

Not all VCA's on a synth are capable to generate these effects. Sometimes they are designed only to function for the purpose to give a volume curve (envelope) to a played key (relatively slow changing control signals).

Envelope generator (ADSR)

An ADSR is an Envelope generator. ADSR stands for Attack, Decay, Sustain and Release. Also so called AR envelope generators are known in classical synth design (Attack and Release only).
When digital synth came the ADSR idea was extended to more periods and levels to generate a more complex envelope signal (curve).

An Envelope generator produces a curve, triggered by a logic signal (usually form the keyboard) and this envelope curve can be used to control any other control input of any other module to generate certain effects. Thus, an envelope generator not only is meant to control the volume curve of a played key. It also can control a timbre change curve when fed to the control input of a VCF (as most of you know). Besides these quite know use of it you can may apply an envelope curve to any control input an have the effect of a module depend on the curve. This is why it is in synthesizer technology so important to have as much parameters depend on a control voltage (or digitally simulated s such).

An Envelope generator starts to generate a curve whit receives a logic command (Gate and/or Trigger pulse) but there is another Envelope idea. This is the Envelope follower. Most analogue modular synth's do have such an envelope follower which is designed to follow the average amplitude of an audio signal in such a way that it can function as a control voltage for other synthesizer modules.

Playing a guitar and feeding the guitar signal to a VCF synth module and also to the envelope follower (ENV) and using this ENV signal to control the frequency parameter of the VCF will lead to the effect of a function between the loudness you play on the guitar the filter frequency becomes higher (or vice versa when you reverse the function by using an inverter).

You also could use your voice envelope to control the filter in junction with playing the keyboard. Then, each sound you will produce with your voice will turn the filter more or less open. Depending on the Attack and Release setting of the envelope follower this gives certain controlling effects.

An Envelope Follower combined with a Trigger generator (producing a Gate or Trigger signal based on a certain voltage level reaching and dropping) also delivers interesting signals based on an external input.

13

Additional 

Introduction

In basic analogue synthesizer technology, each unit is meant to be used into a specific order to generate the final sound. We already saw that for this a minimum of 4 units is required: VCO, VCF, VCA and ENVELOPE FOLLOWER. To make this base setup more useful for a normal musician a Keyboard that delivers a key depended voltage and Gate/trigger signal is necessary.
But, beyond these basic modules, traditional analogue synthesizer came up with a S&H and SEQUENCER module and on the manual control side a ribbon control and pedal.
Furthermore Wave form shapers, ring modulators and additional filters, delay lines and mixers are typically additional modules.

A module that cannot be controlled by a Control Voltage does not really belong to the Class of analogue synthesizer modules but nevertheless can be very useful in an analogue synthesizer setup.
Most delay line based effect pedals for example do not have an external control voltage input but are never the less very useful in a synthesizer setup. But, more interesting would be when the delay time would be controllable by an external Voltage because then the module becomes a dynamic part of the whole setup.

One can say that anything is allowed to be used in a synthesizer setup. Not only those modules that treat audio frequencies but also that can treat digital levels (trigger, gate, Clock pulses).

Do not forget that the synth's you buy in a shop are designed for the masses and thus do not cover everything that is possible.

Digital Delay Line

All delay based effect pedals - build until the nineties - are actually internally based on a control voltage to set the delay time. They easily can be modified by an electronic engineer or by a capable hobbyist to add an external control voltage input. These modules are: Phaser, Flanger, Reverb and Echo. So, this is worth thinking about.
You could let the delay time for example decrease when the frequency of a VCO becomes higher or vice versa. Or, vary the delay time in function of a VCF or VCA setup.

An extra VCA for example could vary the delay line output volume in function to other dynamic parameters of a synth setup and then mixed together with the dry sound. This principal you can do with many non-Voltage Controlled Modules: adding a VCA to at least be able to vary the volume-in-mix amount in relation to the dynamics of the other parameters.

SVF (State Variable Filter)

Computer

A computer for example can be excellently turned into a control source for certain parameter controls. I do not speak about MIDI control. No, much more simpler.
A simple BASIC code program could toggle a printer port pin up and down (high/low) and this could be used to trigger or gate a module.
Using the computer as a programmable divider for a Clock signal only provides that you feed a Clock signal to the computer (printer port pin) and let a BASIC code program count the pulses and respond on it by sending on another printer port pin a pulse out. A printer port has 8 bits and thus you could use 4 of them as an input for certain pulses form the synthesizer setup and use the other 4 to send the pulse back with having them put into a certain relationship together by a small programmed computer algorithm.
Once you learn how to use your computer hardware and a suitable programming tool (one that first to your level of knowledge) then you can come up with many ideas to play around with analogue synth modules.

Analogue Reverb

In many old organs and guitar amplifiers you can find a analogue reverb unit designed around a mechanical reverb string element. This element consist of a metal case in which one or more springs are hanged which are brought into vibration by an audio signal trough a moving coil system. One end of the string(s) hangs into a coil and vibrates based on the audio input. The other end hangs into another coil which functions as pickup element connected to a preamplifier to amplify the delayed audio signal that did vibrate the string(s) and thus was delayed due to the a mechanically delay response of the string.

This principal to delay an audio sound to convert it first into a mechanical movement and then - after going through a metal string - converting back again to an electronic audio signal, has many disadvantages from an audio technical point of view. But despite all of its imperfection creates a certain unique sound that nowadays many of us forgot about.

Analogue mechanical reverb systems are designed in many ways - most of them in a bad way due to commercial aim.
The first matter in the design is the kind of spring element used. There are short ones (little delay) and long ones (a longer delay time) there are single, double and even triple spring elements.
Without diving too deep in all the mechanical-electrical specification of various produced strings, we can simply conclude that the most common quality string element is the so called "Hammed" string element. This element was built in many Hammond organs. It is a double anti-parallel spiral string element. This means that certain resonance disadvantages of a single string were compensated by combining more strings together in opposite winding direction (two pairs).
Unfortunately the Hammond designers and many of their successors did not invoke this high quality string into a high quality and maximized electronic circuit. I guess that commercial targets did prevent them from doing so. Secondly the spring element is in many devices situated in a wrong way. It is often rotated sideways which effects the sound quality negatively.

I have studied about all the designs on using a mechanical string element and combined this knowledge with some own ideas to see whether I could improve and the sound quality of this kind of analogue reverb to give a more wider usage and to determine the maximized sound more to the needs of a certain piece of music. I will now try to explain how this is achieved and stimulate you to use such a reverb together with some additional units to improve this kind of analogue mechanical delay and have more control on the sound character.

As already pointed out the mechanical spring element must be of a "Hammed" type to achieve a reasonable result.
In my Unit, which I designed in 1985 as a final version 4 of these spring elements are involved. Actually the unit counts two identical Reverb units - each equipped with two string elements.

Why a two elements units?

Playing with my experimental setup, I came to the idea to dampen the strings (which produce quit a long delay decay) to make the decay shorter (a cellar effect). This I simply did by dampen the springs mechanically by putting a dot of soft wool (make-up or medical type supplied). Originally I had the intention to be able to control the amount of damping by mechanically put more pressure onto the wool which push onto the springs. This in principal also would be possible to be electronically be regulated by a Control Voltage converted into a mechanical movement to put more pressure onto the damping material.
With a balance control I can balance between long and short reverb. Also the output of each spring are brought out to the front panel and thus I could decide to Voltage control their balance by using two VCA units and thus dynamically control it based on musical requirements.

Frequency Compensation

A second features - which you only find in the better spring reverb units is to compensate the frequency response. The coil that drives the mechanical string has the nasty habit to be more strong for low frequencies than for the higher frequencies and thus the result - without extra compensation - will be that the low frequencies do have more reverb than the high ones. A typical problem with many spring reverb units build into guitar amplifiers.
At the time that the spring reverb units were designed into organs and guitar amplifiers compensation of this frequency response was relatively expensive and thus only could be found in the expensive models. With the availability of integrated circuits (chips) in the early eighties this changed but at the same time new delay (reverb) techniques came up and so optimizing the mechanical reverb with a string element was hardly updated.
In my design I compensated the frequency response of the spring fully by choosing the right so called Op-Amp chip and putting the spring(s) in a compensating feedback-loop. Practically this means that the amount of amplification grows with the frequency becoming higher. Low frequencies, which easily can drive the mechanical spring are amplified little and higher frequencies, which need more power to drive the spring, are amplified more to equally move the spring in an equal vibration.
At the other end of the spring a high quality MIC amplifier is put to amplify the weak response signal from the spring to become a high quality signal again.

The result is a reverb signal that sounds much more bright than those of many implemented reverb systems in organized and guitar amplifiers. Simple and very effective.

What can you do to improve you reverb system?
This depends very much on how separately you can patch the reverb unit. When the reverb is circuited without being separately being used from the original sound, then it becomes hard. But, when you can feed the reverb with an separated sound source, then you can first feed the original source to an equalizer or HP (High Pass) filter and make sure that the frequency curve amplifies more on the higher frequencies. Thus: up going amplification curve.
Putting also an equalizer or filter to correct the reverb output is advisable.

Resonance Compensation and effect

The third problem is the own resonance of the mechanical spring element. No matter how good you compensate the frequency curve of the element, the spring also has its own resonance. This means that at a specific frequency band the spring gets 'neurotic' and amplifies this frequency band extremely. This resonance extreme give to many reverb units a very specific sound (error).
When you can tune out this frequency at its source (so before feeding the sound to the element) then, the more pure response of the spring element comes at front.
This can be achieved by first having the input signal go through a so called NOTCH filter. This is a filter type that lets you dismiss a particular frequency band. Of course, this should be exactly the band to which the spring element overreacts.
In my reverb unit such a NOTCH filter is included. This filter can be controlled for its center frequency by an external control voltage (thus is a VCF in NOTCH mode). A VC-LFO is included to vary the NOTCH center frequency. The result is amazing. Besides being able to compensate the resonance frequency band of the mechanical string elements, it also features a dynamic control in this frequency response that goes beyond the resonance compensation --> it is very effect full too in a musically sense.
If you do have a module which feature the NOTCH filter function, you should experiment with it and find the frequency band that need to be dampen to get a more natural sounding response from the reverb unit.

Additionally

In my reverb unit I also designed two VCA module and Envelope follower/generator from an AR type (Attack - Release).
This was done to be able to control the mix of original and reverb signal by a control voltage which is related to the musical setup and intentions.

Because I have two of these units into one case - each with two string elements and all the described additional units and features - a very complex analogue mechanical delay effect can be setup into a dynamic context.

Practically

An very good example of the features of my system was to imitate the church speech effect. As you probably know all digital based modern delay effect do have several so called 'church' like effects. They are fun to play with but do not reach the feeling of the reality one experiences when being in a church to listen to a priest and as a musician being cached by the tremendous reverb effect of the voice and therefore being impressed (no matter what the 'idiot' has top say).
Well, I with my setup as described could get close to this sound with old fashion techniques that I improved. My compliments to the designers of digital delay line but, it does not cover the real thing. It is all a suggestive simulation without going deeply enough into the complexes of emotions connected to very specific responses of waves reflecting to walls.

Besides 'Religious' sounds
The example above is a practical translation that is based on a concept to fulfill a certain sound to be needed to be imitated because at that time this was a personal item for me as the designer of the unit. I so much wanted to imitate the sound of a priest in a church and after achieving this made several recording with this 'priest effect'. I managed quit close to get the priest sound that I knew from the visits to church. An important add was to use so called 'anti VCA's' to have the reverb signal be suppressed during the spoken words and have it come up during the intervals (the pauses between the words and sentences) in a certain envelope curve that did suits to the memory. It is great to have a device build that offers you to set all the parameters in a flexible and Voltage controlled way. The way you can start playing and researching the aspects of a certain sound effect and slowly - step by step - start to understand it and achieve a result that satisfies within the needs is worth the energy that is necessary to get to it. It makes the combination of parameter settings so personal and therefore exclusive. This is hardly to be achieved with digital based instruments as being on the market today.
This is only one aspect in concrete of what is possible with a reverb unit as described above. The suggestions on how to improve your own reverb unit are fundamental and you really should take advantage of it by trying out certain suggestions within your range of technical possibilities.

Mixing

As a last point I state that reverb often is mixed together with the dry sound in a static way or semi-static by being manually edited by the sound engineer. I pleat a more dynamic control based on the total setup of a synthesizer complexes wherein a dynamic voltage control scheme also controls the reverb character (depth and timbre). Of course, when the reverb signal is recorded on a separated track, the sound editor always can balance (edit) it afterwards for the total sound.

More

Under construction

Previous part    Next part