The ARP Module 1047 Multimode Filter/Resonator

by Chris Meyer of LearningModular.com

The ARP Module 1047 Multimode Filter/Resonator introduced several significant features to the modular synthesizer landscape in 1970, including:

• It was one of the very first multimode filters (along with the EML Electrocomp 100), offering four different filter output responses simultaneously, with a particularly interesting Notch response.
• Its bandpass output had a gentle 6 decibels/octave slope on either side of its voltage-controllable “corner” frequency, making it ideal for replicating acoustic resonators. At the other extreme, it could be set to a very high “Q” or feedback, allowing only a very narrow band of frequencies to get through.
• It could act as a percussion module, with a dedicated input for a gate or trigger signal to “ring” its filter, with adjustable decay.

Let’s dive in a bit deeper into what made this filter unique.

credit: EMEAPP & David Baron

The 1047 had simultaneous low pass, high pass, band pass, and notch outputs. Particularly interesting is the notch filter, as it provided a substantial 40 decibels of rejection, plus the ability to offset that notch from the filter’s normal corner frequency (Fc). When Resonance (feedback or Q) was turned up, this meant the 1047 emphasized harmonics at the corner frequency, and cut them at the notch’s offset frequency. Some modern modular filters recreate this by using two or more independent filter elements.

scanned by Jammie Logan

The primary method at the time to create a band pass filter at the time was to “couple” together a low pass and a high pass filter. This is seen in the Moog 904A/B/C module combination (low pass, high pass, and coupler), and the Buchla Model 191 Sharp Cutoff Filter. These filters had steep 24 decibel per octave slopes, meaning harmonics above or below the corner frequency were quickly attenuated. That steepness was not ideal for simulating acoustic instruments and spaces.

The 1047 featured a more gentle and realistic 6 decibel per octave slope, which made it better suited for creating sounds that reference acoustic resonators such as a pipe or the body of a violin or guitar. Indeed, a common patch on the 2500 was to send an oscillator’s signal through both the Module 1066 Low Pass Filter and the 1047 in series, to combine a typical enveloped-filter synthesizer sound with a natural resonator effect.

The 1047 also had the option to dial in much higher resonance than other filters of the time, narrowing its band pass response to be as narrow as 1/32 of a semitone. This meant it could be used to pick out individual harmonics in a sound. Combined with its external audio input, it could even be used as a piece of precision audio test and analysis equipment! Indeed, its creator Dennis Colin noted that its circuit was evolved from a design used in analog computers at the time.

Another interesting feature is the 1047’s “Keyboard Percussion” input. A trigger or gate signal that signified the start of a note could be patched directly into the 1047, causing the filter to oscillate briefly in its own. The bandpass output was preferred for this application. The duration or decay time of this oscillation was set by its Resonance control, and it could be tuned with the Frequency control.

scanned by Jammie Logan

Most analog drum machines of both yesterday and today use this ringing filter trick to create their “shell”-based sounds such as kick drums and tom-toms. You will also see some modern modular filters with dedicated “ping” inputs to replicate this effect.

The Module 1047 Multimode Filter/Resonator was so significant at the time that Dennis Colin presented a paper on it at the 1971 Audio Engineering Society (AES) conference, which was later reprinted in the Journal of the Audio Engineering Society.

credit: Journal of the Audio Engineering Society

Advanced synthesists still revere the 1047 today as an example of what a versatile, good-sounding filter should be like.

The ARP Module 1027 Clocked Sequential Control

by Chris Meyer of LearningModular.com

One of the main attractions to using a synthesizer is its ability to create sounds ranging from the familiar to those that cannot be created by traditional instruments. Similarly, some synthesizer modules allow the composer or performer to create strings of notes that would be impossible to play by hand.

The most common tool for this is a module known as a sequencer. For those unfamiliar with them, a sequencer has a number of steps, each of which can be adjusted to output the parameters or voltages required to define an individual note. These parameters usually include the pitch but could also include voltages to control the loudness, articulation, or timbre of the note. The composer or performer sets the values of these parameters ahead of time. They would then use either an external trigger signal or a built-in clock to step through the resulting sequence of notes.

credit: Alex Ball and jondent

The ARP Module 1027 Clocked Sequential Control featured 10 “Positions” or steps, with controls to set three different voltage output parameters per Position. It too could be stepped through its Positions using an external trigger from another module, or by using its internal voltage-controlled clock with a “Pulse Repetition Frequency” that could go from 20 steps per minute (20 BPM) to 400 per second – well into the audio range!

In addition to outputting the three voltages associated with each Position, the 1027 could also send a “Clock Out” signal that was used as a gate signal to signify each new note event. The “Pulse Width” of this gate is set either by a front panel control or a separate voltage control input. When this gate signal is patched to an envelope generator, the Pulse Width determines how long the envelope is held at its sustain phase and is set as a percentage of the note’s overall duration – from 0% to 100% from the front panel, and according to its brochure from 5% to 95% under voltage control (in other words, from staccato to tenuto). This feature was unique to the 1027 at the time.

Since the voltage parameter rows of the Module 1027 could be patched back to these voltage-controlled clock rate and pulse width parameters, a composer had considerable control over the duration and articulation of each individual note in the sequence.

The 1027 could also output a separate “Position Gate” for each individual step, one of which was patched back into the 1027’s Reset input to set the length of the sequence. Other ARP 2500 modules – such as the Module 1036 Sample & Hold Random Voltage – could be patched to the 1027’s Reset input (as well the 1050’s Step input), in this case to create random sequences.

The 1027 could also be used in conjunction with the ARP Module 1050 Mix Sequencer to select one parameter row at a time to send to other modules, resulting in sequences up to 30 steps long.

credit: EMEAPP

It’s amazing to think the Module 1027 could do all of this with essentially transistor-based technology. Sequencers in today’s modular synthesizers can range from basic units that are simpler in function than the 1027, to microprocessor-controlled composition stations that include and go beyond the capabilities outlined above. However, when most electronic music listeners think of “sequenced” music, they’re usually thinking of the repetitive strings of notes that the Module 1027 in its simplest application made possible with the ARP 2500.

Patching with Switch Matrices

by Chris Meyer of LearningModular.com

In many cases, a fully patched modular synth looks at best like an old-fashioned telephone switchboard … or at worst like a rat’s nest of cables. As a result, it’s nearly impossible to understand a patch with just a quick glance; each cable needs to be traced from its starting point to its destination, with cables often weaving in front of and behind each other. And along the way, those cables often obscure the controls on the individual modules.

ARP co-founder David Friend once shared with me: “We looked at that and said, you know, there’s better ways to do this…” The better way ARP came up with for the 2500 was a set of switch matrices positioned above and below each module.

credit: Alex Ball

Each module is labeled to show which of its inputs and outputs are mapped to each vertical column of the switch matrices. The composer or performer then moved each color-coded switch vertically to decide which horizontal bus that input or output was connected to. Switches set to the same horizontal rows shared the signal on that bus between their associated modules.

credit: David Baron

As an ARP 2500 brochure of the time noted: “Matrix switch interconnection of the modules allows the panel areas to remain uncluttered and accessible no matter how complex the interconnection pattern…. Set up time is drastically reduced and quick visual checks of all signal and control voltage paths are possible.” This also made it much easier to notate and recreate a patch on the 2500.

Another advantage of the matrix switch system is that the player had access to essentially unlimited “multiples” with the ability to set as many modules inputs as they wanted to the same bus. The earliest models had 10 positions switches above the modules and 20 position switches below; later models features 20 positions both above and below. Although this was sufficient in most cases, a very complex patch could result in running out of horizontal bus lines available to make unique connections.

The biggest problem with this system was that the mechanical switches suffered from electrical crosstalk between adjacent busses, sometimes causing audio leakage and the occasional unexpected result. Alan Pearlman himself later conceded that their system was “a bit noisier” than the typical patch cable approach, and reverted to patch cables for the ARP 2600 – but with the addition of normalized connections pre-patched behind the front panel, creating a common default patch to save a musician time while setting up.

credits: Rob Currier/Retroaktiv

Most modular synthesizers today are still burdened by the use of traditional patch cables. However, there has been a minor movement to reintroduce matrix switches and even matrix mixers. The difference is that many of today’s modules feature a different hardware approach compared to the 2500. Whereas the switches in the 2500 carried the actual signals, modern matrix switch modules instruct other circuitry which connections to make, greatly reducing the chance for electrical crosstalk. Advanced matrix modules use computerized memory for instant recall of their routings, and in some cases allow voltage control over those connections. The front panel of the non-modular Arturia MatrixBrute is a good example. However, they all still point back to the switch matrices in the ARP 2500 as one of their ancestors.

Analog Computers and Modular Synthesizers

by Chris Meyer of LearningModular.com

Many fans of modular synthesizers look at the pre-digital analog computers of the 1950s and 60s, and assume they must have been inspiration for the first analog modular synthesizers created by Bob Moog and Don Buchla. But that’s not entirely true.

For those unfamiliar with them, analog computers solve problems in real time by simulating physical objects with electronic modules. These general-purpose modules are patched together and tweaked to get an idea of how a physical system would respond. Many of their components – such as attenuators, oscillators, amplifiers, integrators (filters), adders (mixers), and the such – have analogies in the modular synthesizer world.

Marcin Wichary/WikiMedia Commons

Take the problem of simulating an automobile suspension with an analog computer: The weight of the car is set using a bias or offset voltage. A bump that it hits in the road can be simulated by a transient or envelope generator. The bounce and rebound of springs are similar to a resonant filter (hopefully not too resonant!); the damping of a viscous shock absorber is similar to a slew limiter. It is easy to test variations on a suspension’s design by tweaking the parameters of these electronic circuits.

However, neither Bob Moog nor Don Buchla are known to cite analog computers as an inspiration for their work. Herb Deutsch has mentioned in interviews that Moog was inspired by an automated automobile assembly plant nearby, which used voltages to control the movement of hoists and the such. And when asked how he arrived at the voltage control approach, Buchla – who had backgrounds in music, engineering, and physics – replied “I don’t know. I just did it.”

Instead, it was Alan R. Pearlman who had a very direct connection with the analog computing world. A co-founder of Nexus Research Laboratory in Canton, Massachusetts, he designed multipliers and amplifiers for military- and space-grade analog computers. Pearlman applied this technology to create the more stable circuits that ARP modular synthesizers were known for, especially when compared to the early Moog and Buchla models. In a nod to his days at Nexus, exponential converter circuit components were even listed as items for sale in the early ARP 2500 catalogs.

scanned by Jammie Logan

Another example of the connection to analog computers can be found in the Audio Engineering Society paper engineer Dennis Colin wrote on the 1047 Multimode Filter Resonator. In it, he wrote that “the filter is basically an analog computing circuit consisting of summers and integrators, set up to solve a second-order differential equation. The circuit is well known and documented in analog computer and servomechanism fields.” An example includes the EG&G Parc Model 189 Selective Amplifiers shown below. The difference was that Colin added voltage control to several parameters, making the design more usable in a music creation context.

credit: Hainbach

Today, some electronic musicians are going back to the early days of electronic music, integrating electronic test equipment and – yes – analog computers into their music-making process. But it was Alan R. Pearlman who most directly bridged the gap from analog computers to modular synthesizers.