Solid-State and Multiplex Pipe-Organ Control Systems

Scanning replaces one wire per contact with one question per contact, asked over and over. A theatre-organ society glossary describes a multiplexed electronic relay as scanning every key input of every manual and pedal at a given frequency, and gives the considerable reduction in wiring complexity as the main advantage of doing so. The scanner does not care what a contact means: a key contact and a stop sense are alike to it, each simply a position in the sequence whose state, on or off, is read on every pass. Whether piston contacts are scanned with them is a system-by-system matter, taken up below.

The reduction is visible in one supplier’s manuals for two earlier products. Syndyne’s keying system multiplexes manual keys, stop senses and piston contacts into a console processor, which then sends everything into the organ over a four-wire data cable, and its single-board combination action multiplexes the general and divisional pistons through eight piston inputs. Those wire counts and capacities belong to those two products; they show what multiplexing does, not what every system provides.

A multiplexed transmission is still a switching system, the same layer that a relay cabinet or a diode panel occupies between console and chamber, and not a different kind of instrument. Peterson’s own name for the OrgaPlex, a time-division multiplex switching and coupler system, keeps both jobs, switching and coupling, in the description.

Once the console has been scanned, the states travel in time order rather than in parallel. Peterson states that the OrgaPlex converts keyboard and stop data to serial form for transmission on a main cable, and its installation manual documents that main cable as carrying 30 data conductors arranged in groups A through D, with modular telephone-style data cables joining the components inside the chamber. Peterson’s same catalog document, as of September 1, 2026, states that OrgaPlex development began in 1985 and that nearly 2,000 systems are in service worldwide; that is the manufacturer’s count of its own product, not an industry figure.

Solid State Organ Systems describes MultiSystem II differently at the console end but identically in principle: the system reads the keys, stops and expressions from the console and routes everything through a network to a processor. The serial data network is the cable and protocol that carry the scanned console state to wherever the outputs are.

At the chamber the stream is unpacked back into individual circuits. Peterson states that the OrgaPlex demultiplexes the serial data back to parallel outputs in the organ chamber, and its installation manual lists demultiplexer sizes from 13 to 109 notes, so a demultiplexer is sized to the portion of the organ it serves rather than to the whole instrument. Each parallel output is a driver output: the stage that actually switches current to a chest magnet so that the windchest action can open the pipe valve.

In Solid State Organ Systems’ description of MultiSystem II, the network delivers the console data to a processor which activates the correct chest drivers and expression. The chain therefore ends at the driver, not at the data. A key press that reaches the chamber as a bit in a stream does nothing until a driver output turns it back into current.

The sources do not agree on how few conductors a multiplexed rank needs, and the disagreement is worth keeping. A 1976 patent for an electronic pipe-organ control system claims serial multiplexing of keyboard data such that as few as three wires per rank, carrying data, clock and synchronization, take the place of what the patent calls a substantial multiplicity of wires between console and pipe chamber. The theatre-organ society glossary, by contrast, describes the scanned result as carried on a single wire per rank. One account is a claimed design at its filing date and the other is a glossary description of the class; neither settles the count for any particular installation, and a patent establishes a claimed design, not what was built, sold or installed.

A 1980 patent assigned to an electronic-organ manufacturer claims a further step: multiplexing several manuals and the pedalboard simultaneously into synchronized serial data streams. Its stated motivation is that separate scanning circuits and wiring for each manual made the circuitry complex and unwieldy, increasing size, cost and maintenance problems. Again this is a claimed design at a date; it does not show adoption or field performance.

In a converted organ the multi-conductor harness between console and chambers can be replaced by one data cable. One identified builder, describing its own conversion work, states that a multiplex conversion changes the console-to-chamber cabling to a single standard computer-grade LAN cable, and that this is why a converted Möller console can be moved, or made mobile, in a way the original cable harness did not allow. That is one practitioner’s observation about its own installations rather than a universal rule, and the mobility consequence follows from the cabling change rather than being a property of multiplexing itself.

On the OrgaPlex, Peterson states that the entire organ is refreshed approximately 120 times per second, which is one scan every 8.3 milliseconds. Peterson also states that the system carries extensive self-diagnostic capability with built-in LEDs for monitoring the data streams, which is the manufacturer’s description of its own product rather than a fault-finding method. That refresh rate is Peterson’s figure for its own system and does not describe other manufacturers’ systems.

Two accounts give different answers, and they describe different designs, so this page presents both without choosing. A practitioner author identifies variable scanning latency as the most controversial issue in electronic transmission. His reasoning is mechanical: a key pressed just after the scanner has passed its position waits a whole scan cycle before it is read, so a system scanning at 25 Hz yields a delay that varies randomly between zero and 40 milliseconds. He recommends a minimum of about 1,000 scans per second together with sample-and-hold.

In a professional-association roundtable on solid-state switching design, a hard-logic multiplexing design is described as adding only a fixed 10 millisecond delay, while computer multiplexing is said to be able to introduce delay when many keys, pistons or MIDI functions are called for at once, with some systems reported to give up under such load. The fixed 10 millisecond figure and the variable zero-to-40 millisecond figure are not measurements of the same system; one is a designer’s statement about hard-wired logic and the other is a worked example at a stated scan rate. What both agree on is that latency in a scanned system comes from the scan cycle, and that the rate matters.

Some systems route every scanned input through one processor and some have no central processor at all, and the manufacturers say which they built. Solid State Organ Systems describes MultiSystem I as the first software-based pipe-organ relay, developed from 1990 with first delivery in 1992, and its manuals index dates the product from 1992 to early 2000. Its MultiSystem II description keeps a single processor in the chain: console data routes through a network to a processor that activates the chest drivers.

Syndyne, as of September 1, 2026, documents the MS8400 the other way. Its Organ Builder Manual states, ‘The MS8400 Master System does not have a central processing unit. Instead, computing is done by multiple cards throughout the system’, and the product page adds that ‘If one part of the system fails, the remaining parts continue to operate.’ That is Syndyne’s statement about its own product. A distributed controller of this kind is one architecture of the switching layer, not a different function; the cards together do what one processor does elsewhere.

Distributed processing puts computing near the chests as well as in the console. Peterson, as of September 1, 2026, describes the ICS-4000 as a distributed system in which a main CPU cage in the console is paired with Satellite card cages in the organ chamber, each Satellite cage combining a microcontroller board with Input, Output and Stop Action Controller boards, and names a Satellite Streak Processor that it credits with increasing data throughput more than tenfold over earlier versions. The throughput comparison is Peterson’s, against its own earlier product.

Solid State Organ Systems’ MultiSystem II Installation Guide, in its revision of 2 May 2013, states that processors read information from planes, that each processor may operate up to two planes, and that planes connect through 25-way D-type connectors carrying the information and power the processor requires. Syndyne documents an MS8400 console architecture in which each keyboard and pedalboard is wired to a separate keying input card, the system accommodates up to 224 pistons and up to 1,280 stop controls including couplers per console, and the console and chamber chassis are joined by a CAN bus over ethernet cabling. Those are design capacities of one product generation; no particular console carries those counts, and they say nothing about other manufacturers.

The cable between locations is, on the documented current systems, ordinary network cable, and each manufacturer sets its own limits on it. Syndyne documents the MS8400 as connecting its cards on a CAN bus carried over Cat5e or Cat6 cable with RJ45 connectors, with an MS8400DTR termination resistor required on the last board of each CAN run and on every otherwise unused CAN port, an MS8400DJMP jumper for an unused touchscreen port, RJ11 connectors for expression shoes, crescendo shoes and rotary selectors, and an MS8420 fiber optic converter offered to extend distance. Those are the manufacturer’s stated requirements for its own network, and they mark where one distributed system’s boundary is drawn.

As of September 1, 2026, the MultiSystem II product page states that the system uses CAT5 100BaseT Ethernet cabling between locations and Krone punch-block connectors for field wiring, and describes its inputs as high resistance with built-in surge protection and debouncing, accepting 10 to 36 V. As of the same date the MSM Setup & Operation Guide from Solid State Organ Systems specifies Cat5e or Cat6 data cable with a maximum of 80 meters between devices, and states that data cables must not be chained or ‘Tee’ jointed but must run point to point. A network that looks like office cabling still has organ-specific rules attached to it, and those rules are the manufacturer’s.

The output stage is a set of modules, each with a fixed number of circuits, and each circuit switches the current for one chest magnet. As of September 1, 2026, Solid State Organ Systems states that MSM offers single-column modules of 32, 64, 96 and 128 circuits and a four-column plane of up to 512 connections. The MSM Setup & Operation Guide, as of the same date, states DC organ power of 12 to 36 volts and negative output modules switching up to 1.5 A per circuit. Those figures are one manufacturer’s, for one product, and the guide’s own revision history marks it as prerelease.

What the figures show is the boundary between data and current. Everything upstream of the output module is a bit in a stream or a state in a card; at the output module that state becomes a switched circuit at organ voltage, and the chest magnet on the far end of it is the first thing in the chain that moves.

In a multiplexed system a coupler does not need a relay of its own. A practitioner account of electronic transmission describes coupling as a data operation: when a coupler is drawn, circuitry in the console multiplexer inserts a duplicate set of ‘on’ pulses into the scanned data stream, so the coupled division’s notes arrive at the chamber as if they had been played on that division’s own keyboard. The same account places keying, coupling and stop scanning together inside the transmission scheme. Peterson’s description of the OrgaPlex as a switching and coupler system says the same thing from the manufacturer’s side: coupling is a function of the transmission, not an add-on to it.

On the documented systems, board addresses, polarity and a number of behavioral settings are configuration rather than wiring, but not all configuration is reachable from the bench. Syndyne’s CS5608 combination action, a 2006 product generation, places settings such as the number of general pistons, multiplexing, pulse timing and feed polarity behind onboard DIP switches and jumpers that require physical access to the board, while the organist retains memory level selection and Set and Cancel operation. On an older solid-state system of that kind the technician boundary is physical rather than credential-based.

Syndyne’s MS8400 Organ Builder Manual documents Backup All Cards, Backup Single Card, Restore All Cards, Restore Single Card and Firmware Update as system functions. Its Quick Start Guide places backing up the system configuration as the final step of a twelve-step builder bring-up sequence. That sequence begins with unboxing and inspection, DIP switch addressing and polarity jumpers, and proceeds through mounting, wiring, powering and LED verification, testing, chamber relay connection, console and chamber configuration, organist setup and MIDI configuration. How much of a distributed system’s identity lives in configuration shows in that list; the steps themselves are the builder’s work.

Syndyne also includes a Test for Address Conflicts function to identify duplicate board addresses. Its existence documents that, in a distributed system, a configuration conflict is a recognized cause of misbehavior alongside component failure. It establishes that the condition is recognized on that product family, not how often it occurs in service, and it applies to MS8400 systems only.

The combination action usually does. The same practitioner account states that combination pistons are not usually part of a multiplex transmission scheme unless part of the mechanism is in the organ chamber. Solid State Organ Systems draws the line the same way for its own product: MultiSystem II’s combination action and sequencing are supplied by a separate module from the switching that plays the pipes. The practical consequence is that a console’s combination behavior can be changed without changing the system that scans, transmits and drives, and the reverse is equally true. Combination action, MIDI and console software are a separate subject in their own right, as are the questions of identifying an original or retrofit control system and of retrofit scope.