Electronic Relays in Pipe Organs

Buddy Boyd

Foreword

Recent discussions on the list have prompted me to write on the subject of electronic relays. While this material may well become dated in just a few years I hope that it provides some insight into this area for the organist and the hobbyist on the workings of the instrument.

History

I have in my collection George L. Miller's "The Recent Revolution in Organ Building", c 1913, reprinted by the Vestal Press in 1969. This would seem a good place to start. Electricity had found its way into organ building and for the first time you could now build a movable console. Up until this point the console was connected to the organ via the mechanical linkage of a "tracker" or lead tubing for a pneumatic system.

These pneumatic organs were later converted to electro pneumatic by replacing the tube connection between the console and the chest with a "primary" chest action and then fitted with an electric console. Many Mollers were built with chests that were made for a pneumatic organ and then fitted with the primary valve set. They didn't have to redesign the chest or change their drawings.

Now these organs are for the most part pretty much straight organs. I mean that each rank was 61 notes and the manual chest contained all the pipes of the rank. While the use of a "Pitman" chest would allow the division to play from another manual, borrowing or unification was pretty much not done. All the action for any couplers was located in the console. Offset chests for large pipes were dealt with in what ever manner the builder devised. I remember one Skinner that had contacts over the pipe holes on the bass pipes on the manual chest. When the note was played, the contacts closed and provided the necessary signal to the offset chest.

So here's what was going on in the console: for a completely straight organ there would be one contact wire under each key and this would go to the manual chest in that division.

If this was the great manual and we wanted to add a solo to great coupler it would be necessary to add another contact under the key. Why not just wire it to the existing contact? The first reason is that the contact is now supplying twice the current as before and would soon burn out. The second is for isolation. If it were wired to the same contact as used for the great manual then when you played a note on the solo division the note on the great would also play. This problem is known as a "run" and will occurs when two contacts cross under a key, when there is a short in the cable, or if there is some trash on the key contacts that shorts them together.

I need to add at this point there is also in the cable between the key contact and the chest a switch with 61 contacts that close all at the same time. This turns the division or coupler on and off.

Well when you start adding all those divisional and manual couplers you have a number of contacts under each key.

Now comes the idea of the unit chest. This chest will have one magnet per note on the chest and can act like a division. For example, if you wanted to have the trumpet in the swell play in the pedal it is now possible by putting the trumpet on a unit chest and adding a contact under the pedal keys and the swell manual key. Can you see what will happen? The more ranks you have that you can borrow the more contacts you have to have under the keys. Pretty soon you run out of room down there. Think about a theatre organ with all those contacts. The result was the key relay. In the early organs starting at the turn of the century these were pneumatics in a chest with a large number of contact before each one that acted when each key was played. By the mid 1920's large electromagnets started replacing the pneumatic actions. These "direct" key actions are still available today. And remember that 61-contact switch. You are going to have one of these in line between the pedal and swell manual each and the trumpet chest in the swell. So now when you pull a stop instead of it moving a slider or chest ventil it closes the appropriate switch.

So we have a few contacts under each key with wires leading from the console to a Key Relay. Each key relay has many wires that fan out to 61 note (contact) switches. The wires from these switches go to the magnets on the manual chest and offsets.

The Diode and the Transistor

If you have ever seen the working of a large original theatre organ, whole rooms the size of a medium chamber are given over to a relay. And besides the combination action, much of the console is filled with the coupler action. With the invention of the semiconductor came the promise reduction in size.

A DIODE is a device that allows current to flow in one direction and blocks its flow in the opposite direction. With a diode we can reduce the number of contacts under a key and fan the signal out from each key to hundreds of places. The problem comes when the current necessary to drive many chest magnets is more than the key contact can deal with. We need a way to drive a larger amount of current.

A TRANSISTOR is a three-terminal device. When a small amount of current is allowed to flow into the "base" this allows a large amount of current to flow between the other two terminals. Factors of 30 to 100 times are very common. So now we have a way to produce enough current to drive a chest magnet with very little demand on the key contact - a relay driver.

Both of these devices are available in single components or integrated circuits (chips).

Diode Resistor Logic

One of the early relays developed used two diodes and a resistor to replace the multiple contacts under the keys and the 61 note switches. These were wired together in the same configuration as the Electro-pneumatic relays and were developed first for electronic organs. By replacing the sound producing circuits with transistors or relay drives that could power the chest magnets these systems became available for pipe organ use. This greatly reduced the size of the relay but takes a lot of work to construct one even with the purchase of circuit cards. For the amateur organ builder who is trying to save funds on his home installation this is one of the cheapest systems to build yourself. You can build it completely from scrap or any level of production materials you want to purchase or make.

I have noted a few problems with this design: All the parts need to be kept close to the key action. Long wires will pick up radio signals and cause notes to play. A transistor takes only 0.7 volts on the input to turn it on. Long power supply runs can fool transistors into turning on. But with proper installation this will work for years and the parts are available and user maintainable.

Scanning Systems

With all those contacts and stop tabs the cable between the console and chamber/relay was getting pretty large. Many organs not only had a relay located away from the console but also the combination action. With the advent of the integrated circuit a lot of operations could be placed in a small space. Two of these early systems are still on the market almost 30 years later. These systems do not require a computer. Their cost is much lower than other systems. But as with all things in nature there are certain tradeoffs. The systems once installed are not simple to change. Any changes and additions require additional hardware and/or wiring more than their computer counterparts. Each board in the system performs a function. A number of each type of board are connected together to produce a relay that controls the chest. The more ranks you add the more boards you need. Other functions may require the purchase of additional special function boards. Pay as you go. These systems reduce the cable between the console to the chamber down to about two wires per division or rank if it is all unit chests.

Computer Systems

There are a number of designs that start with something like the system just described to more complex software dependent systems.

At one end of the scale is a system that uses a computer to set up the functions of the cards. You may have a set of console boards that will need to be set up for the particular couplers you desire or to store combinations. Once these things are loaded into the relay the setup computer is then removed. But there are smaller computers on the boards that run their own little programs to carry out your instructions. So we still have a computer(s) running things.

This frees up your PC to do other things. You just have to bring it out whenever there is a change to be made.

At the other end of the scale is a fully dependent control system. The input boards in the console are all the same. Wires from the keys, pistons, and stops are all connected to the same board. A cable from the board goes to a local PC where you have told the software what each wire goes to. The PC then scans the inputs and determines what action it has to carry out. It then sends a command out to the chamber where a similar board is located with the chest, shutters, and other devices are connected. Again the computer has been told what wire is connected where and sends the proper signal to the proper output. These system will act as a relay, combination action, organ recorder and playback. You can even set it up to dim the lights if that is your thing. Because of the PC it will even interface with the latest electronic gear. Changes to the specification can be made in minutes. By storing the data for recall you can change organ specifications just like you can select different memories in some combination actions. It will do all that as well. These are complicated systems and are not always simple to fix. In most cases they will do much more than the average organist might ever demand of them. PC's and operating systems advance and as long as the builder supports his product then there is not much of a problem with the system going out of date.

In Conclusion

In general, on systems I have seen when the manufacturer has gone out of business another company has stepped up to continue to support the product. This speaks well for the systems performance and desirability.

I have tried to speak in general terms and I hope I have not offended any manufacturer. I expect this to become dated and yet I hope that those system I have had in mind when writing this will be around in another 10 years for the next edition.


About the Author

Buddy Boyd is a 52-year-old Senior Systems Engineer for Northrop Grumman. As a child he took more interest in classical music than the Beatles. The performance of transcriptions of orchestral works introduced him to the pipe organ. With an interest in mechanical things at 15 he started working with a friend rebuilding player pianos on weekends. He later apprenticed himself to an organ builder and was able to pay his beer tab at his fraternity by tuning and repairing organs while at university. He has disgraced himself at the console both in the United States and Great Britain. Today he happiest playing for his own amazement and spending most of his time behind the console.

Buddy Boyd