Electromechanical Pipe-Organ Relays

The relay sits between the organ key and the pipe-controlling magnets, and it turns one closed contact into several closed circuits. The description comes from a 1916 electromagnetic relay patent written expressly around a pipe organ, US Patent 1,204,469, which places the relay in circuit with the pipe-controlling magnet, a stop switch and an organ key. Depressing the key completes a circuit through the relay’s electromagnet. The energized magnet moves an armature, and the armature closes a plurality of auxiliary circuits, so that a single key contact ends up energizing several pipe-controlling magnets at once.

A patent establishes a claimed design at its date. This specification describes one relay construction, built around a rotating cylindrical armature; other relay constructions exist, and the document says nothing about how many relays of this pattern were made or how they behaved in service. What it does fix is the vocabulary and the routing: key contact to relay coil, relay armature to magnet circuits.

The stated reason for interposing a relay is current: the patentee’s position is that a key contact by itself cannot supply enough current to operate several pipe-controlling magnets at once. Routing the key current through a relay first means the key contact carries only the relay coil’s current, while the armature completes the heavier circuits to the magnets. That is the patentee’s rationale for one design, not a general engineering limit that applies to every key contact and magnet combination, but it explains why this generation has a relay at all rather than a direct wire from each key to its chest magnets.

Key contacts perform keying. Each contact is the make-and-break point operated by a key; closing it is the command, and everything downstream exists to carry that command onward. In a relay installation the key current goes to the relay’s electromagnet rather than directly to the chest magnets, which is the arrangement described above. The key contact is a distinct component from the relay: it originates the circuit, and the relay multiplies it.

Coupling, which makes one division’s stops sound from another keyboard, is performed by coupler relays: ganged switches operated by solenoids or pneumatic motors. Drawing a coupler operates the gang, so that the key circuits of one keyboard also complete the circuits belonging to the coupled division. Because the coupler is a separate mechanism with its own switch bank and its own driving solenoid or pneumatic motor, coupling in this generation adds hardware, not just wiring.

Stop selection is performed by stop switches, also called ladder switches, which select which relay contacts are live. Drawing a stop does not send a note anywhere; it enables a set of relay contacts so that the key circuits passing through them reach the chest magnets of that stop. This is why keying and stop selection are separate hardware sets in electromechanical practice: the key contact decides when, and the stop switch decides which.

The stop-by-key arrangement was itself the subject of patent claims. US Patent 1,788,963, granted to Harold L. Steinfeld in 1931, claims a switching base plate whose contact members are arranged in longitudinal rows, each row corresponding to a coupled or unified stop, and in transverse rows, each corresponding to a key, with conductors arranged in echelon to connect contacts in octaves. Claims 7, 9 and 10 of a patent filed in 1929 establish a claimed design only; they do not establish that the apparatus was manufactured, adopted by any builder, or performed as claimed in service. What they show is the shape of the problem this generation solved in metal: a physical grid with stops along one axis and keys along the other.

The governing rule for electromechanical keying is that the number of contacts operated by each key must equal the number of circuits that key controls. Every circuit a key can call for needs its own contact under that key, because a single contact carries a single circuit. The 2014 wiring-diagram article at colinpykett.org.uk that states the rule gives about ten contacts per key as a reasonable working limit, beyond which key relays become necessary; that figure is one practitioner’s rule of thumb, not a standard.

Key relays are the relay mechanism applied to the keying problem: instead of stacking more contacts under a key, the key operates a relay whose armature carries the additional circuits. The current rationale in the 1916 patent and the contact-count rule in the practitioner article are two views of the same constraint.

Because each contact carries one circuit, the wire count grows with every function, and the physical result is the thing a reader notices first. A theatre-organ society explainer works the arithmetic for one rank at one pitch on a three-manual instrument: a switch for each of 183 manual keys plus 32 pedal notes, before any further stop or pitch is added, and a cable leaving the console that could be eight inches or more in diameter.

A 1977 patent for a computerized unit organ relay, filed 29 July 1977 and granted in 1979, characterizes the prior art it set out to displace in the same terms: a keyboard relay occupying a space many feet square by a foot deep, wiring demands reaching about 6,100 individual wires for a hundred-stop organ, and electronic gate-logic replacements that still required some 6,100 AND gates plus many OR gates for the same instrument. A patent characterizes prior art to motivate its own claims, so the figures are the patentee’s framing rather than a survey, but they give the order of magnitude.

On unified theatre organs the relay grew into a room. A 1986 article in Theatre Organ describes the original unified relay as a complex switch matrix that allowed any rank in an instrument to be played on any division at the pitches the specification required, housed in relay rooms containing hundreds of multi-contact pneumatically or magnetically operated relays with thousands of soldered connections. Ray De Vault, writing in the November/December 1986 issue, calls the caseworks “cumbersome, exceptionally heavy” and “filled with chattering magnetic relays (some had pneumatic action)”, and describes “a myriad of hundreds of multi-contact pneumatically or magnetically operated relays with thousands of soldered connections and miles of wire”. He is describing unified theatre-organ relays generally and names no builder in that passage.

Key contacts in this generation are metal-to-metal surfaces, and in the documented examples they are renewable as components. Létourneau reported that Aeolian-Skinner Opus 1257 of 1955 had original silver wire and bronze plate key contacts, and that in restoration the contacts were renewed with new silver wire; the instrument’s original pneumatic stopknob actions were replaced with electro-repulsion solenoids in the same project. That is one documented restoration, reported in The Diapason, not a general specification.

Alan Jackson, writing on servicing Casavant organs in the Journal of American Organbuilding, states that Casavant replaced under-key contacts with a roller key contact system for manuals and pedals using self-aligning silver key contacts. Casavant’s own history separately claims patents for a tracker touch contact system in 1903 and for a roller relay contact system in 1918, and states, as of September 1, 2026, that the roller relay is still in use today; the company gives no patent number, jurisdiction or filing date for either claim. Jackson’s roller key contact and Casavant’s roller relay are consistent with each other, but neither source says they are the same thing, and Jackson’s article is undated.

The magnets and relays of this generation were supplier products with their own patent record. W. H. Reisner Manufacturing Company’s organ-relay work dates from 1915; its US Patent 1,204,469, a 1916 specification, is written expressly around a pipe organ and describes the relay connected in circuit with the pipe-controlling magnet, a stop switch and an organ key. The patent establishes a claimed design at its date. It does not establish that Reisner manufactured or sold organ relays commercially, or that any named builder bought them.

A Reisner trade brochure issued under the Organ Supply Industries name states, as of September 1, 2026, that for over three quarters of a century the Reisner chest magnet has been the standard of the industry, and lists pedal pouch units PPA-1 to PPA-5, standard and C-17 chest magnets, the 601 Direct Pallet Magnet and Direct Valve Action units. That is the company’s claim about its own product, in an undated brochure; it is not independent evidence of how many instruments carry Reisner magnets.

The oldest of these patents shows the magnet itself being engineered around the contact problem. US Patent 595,660 of 14 December 1897 claims an electric organ action, in the sense of the keyboard-to-chest connection, using electromagnets with multiple independent windings on shared cores so that only one contact per key is required, a metal veneer of uniform thickness to standardize armature travel without hand adjustment, and resistance circuits to cut power consumption after the armature engages. It is a claimed design at its filing date, not evidence of production or prevalence.

How a console gives access to its contacts is itself a design feature, and the documented examples show that design changing to make access easier. Jackson states that pre-1958 Casavant consoles required removal of heavy keyboards for access, while the later design allows access by raising the lid and unclipping the piston contact splice blocks. Console weight was a design target as early as the 1897 patent, which claims a swing-door console reduced to about 330 pounds for a sixty-stop organ. These are descriptions of what the hardware was built to allow, not service steps.

Electric action, meaning the electrical connection from key to chest, is older than the relay that organized it. A practitioner history of electric actions at colinpykett.org.uk dates the concept to the 1820s and observes that electric organ actions and the electric telegraph developed in parallel, because both consisted of a contact operating a remote electromagnet; early organ actions were unreliable largely because of battery technology. The same history states that after 1950 mains-driven transformer-rectifier sets gradually displaced the low-voltage dynamos and accumulators that had powered electric actions, and notes that some 1970s power supplies sold in the UK were hazardous through defective grounding and exposed live parts. On that account the power supply is one of the things that changed within the generation: dynamos and accumulators first, transformer-rectifier sets after 1950.

As of September 1, 2026, Organ Works / Classic Organ Works publishes a control-system history that dates relay and electromagnet practice to the mid-1920s, after pneumatic experiments in the 1840s, tubular pneumatic in the 1860s and improved electro-pneumatic systems around 1900. That is a manufacturer’s chronology; the 1916 patent described above claims a relay design, and the 1897 patent an electromagnet design, earlier than the mid-1920s date it gives.

When contact switching gave way to transistor switching is not settled, and the two published accounts give different dates. The practitioner history at colinpykett.org.uk dates solid-state switching in organs, meaning transistor switching, to about 1960, and states that transistors reduced the current a key contact had to carry from roughly 100 milliamps to roughly 1 milliamp, which eliminated sparking at the contacts. Organ Works / Classic Organ Works, as of September 1, 2026, labels 1948 as the beginning of a transistor era in its published chronology. The 1948 date is the manufacturer’s chronology, not evidence that transistors were installed in pipe organs in that year, and the about-1960 date is one practitioner’s dating. Both are given here as stated; neither is presented as the settled boundary.

What the current figures do show is the physical distinction the reader is looking for. In a contact-switched organ the key contact carries the working current, with the sparking the practitioner describes; in a transistor-switched organ it carries a signal current roughly two orders of magnitude smaller. How transistor and diode circuits are arranged, and what their presence does and does not establish, belong to the diode-matrix and transistor page.

One institutional record shows the transition happening on a specific instrument, with dates. As of September 1, 2026, the society that maintains the three-manual, sixteen-rank M.P. Möller theatre organ at the Trenton War Memorial records that the instrument’s original electro-pneumatic relay was replaced between 1974 and 1976 by a solid-state relay containing 1,600 transistors, designed by W. McKissock, Jr. and built by society members, and that during renovations from 2002 to 2005 a modern electronic relay was installed to make the console movable. It is one instrument, and it illustrates possibility rather than prevalence: an electromechanical relay in service into the mid-1970s, replaced by transistor switching, and then replaced again a generation later.