Part I of a two part series...
It's all well and good to build a coast station for marine or point to point service. But that work has little practical application if the transmit and receive sites of the station can't communicate with each other and with the outside world via company's central office in the city nearest the station. The early practitioners of the art had to design their way through the problems of land line communications and remote control just as they did when addressing the design of the station itself.
Now comes Mr. Bill Ruck of the MRHS Maintenance Department with the first of his two part series on the methods used from the beginning to the present day.
Gentlemen, we give you... Mr. Ruck...
Of these innovations, a remote controlled transmitter took time to develop. Originally a radio station had the transmitter and receiver in the same place and used the same antenna for both receiving and transmitting. There was a big switch that would be manually operated to switch from transmit mode to receive mode. For example, in the Hillcrest picture below the switch is in the ceiling and operated with ropes and pull handles. Pull one to connect the transmitter to the antenna and pull the other handle to connect the receiver to the antenna.
 |
| KPH Hillcrest ca. 1919 showing antenna switches and .44 revolver at the ready. |
By modern standards receivers used in the early days of radio communications were insensitive. The first generation receivers were passive - the vacuum tube, let alone the transistor - had not been invented. Various detectors were used, many of them mechanical.
By World War I the vacuum tube had been invented and applied to radio receivers which greatly increased their sensitivity. It took many generations of receiver technology improvement to get to the "modern" receivers we use today.
At the same time, improved transmitter technology allowed the increase in power from a few hundred watts to kilowatts to hundreds of kilowatts.
One side effect was that the high power transmitters were literally damaging the receivers. For this reason the trend in "modern" high-power installations (circa 1910) was to separate the transmitter from the receiver. This reduced the power received from the transmitters and allowed simultaneous reception and transmission.
With two frequencies the station could operate full duplex - simultaneous receive and transmit. This is also known as "full break in". In a conversational mode like marine communications this was useful but in the point-to-point service the goal was a constant flow of messages in both directions all the time, full duplex allowed more messages to be passed which improved revenues.
The American Marconi Company knew this when they built their West Marin facilities in 1913/14. The transmitter was located in Bolinas and the receiver was located in Marshall, about 25 miles away. The Bolinas transmitter location was chosen because it was at the edge of the Pacific, reasonably close to San Francisco and had access to commercial power, eliminating the need to construct a power station on site. In the event, lines from two separate sub-stations, at Woodacre and Alto were brought to the transmit site. American Marconi used their standard trans-oceanic 250 kW rotary spark gap transmitter so being able to purchase electrical power rather than generating it themselves on site made economic sense.
We have no documentation on how the American Marconi Company connected the Bolinas transmitter site to the Marshall receive site. Our best guess is that they used a telegraph circuit either installed by Marconi or purchased from the telephone company. Wireline telegraphy was a mature technology at that point. With a slave relay in a telegraph circuit ultimately keying the transmitter a telegraph key at Marshall could send messages from the Marshall receive location. But the actual keying probably took place from the company's central office in San Francisco, as was the case in the later RCA era.
This allowed one transmitter to be operated on the telegraph circuit. This was not an issue then since there was only one transmitter in Bolinas, the rotary gap. There were existing methods of "multiplexing" telegraph traffic on a single telegraph circuit which could have been used to key more than one transmitter on the same telegraph circuit. For example, Alexander Graham Bell was working on a method to send multiple telegraph circuits over a single telegraph line using different tones when he stumbled on what we call today a "telephone".
An interesting side note is the use of the words "mark" and "space". These words came from the days of wireline telegraphy when a high speed tape reader was used to send and a paper tape with a siphon pen was used to receive. The marks on the paper tape made by the pen were called "mark" and "space". Today modern state-of-the-art digital transmission uses the same words to describe a "1" and a "0"!
Although the principles were well known, technology at the turn of the century was limited to mechanical means. Various methods were invented and tried but none worked well. It took the invention of the vacuum tube and improvements in that technology to allow active circuits to make stable and reliable methods of remotely keying transmitters using audio tones.
By the time RCA chose to build the shortwave transmitters in Bolinas in Building 2 and to build the new receive site in Point Reyes there were methods of transmitting Morse code over distance using telephone lines. One method used an audio tone that was keyed on and off. The presence of this tone could be detected and used to remotely key the transmitter from the central office. Incoming tones from the receive site were received by ear or on inked paper tape. With careful design and sharp filters several different transmitters could be keyed over the same telephone circuit using different audio tones. This technique had the disadvantage that noise and distortion could also key the transmitter.
 |
| Operators at an early, unknown central office copying Morse from a remote receive site by ear and inked paper tape |
An improvement was to use a continuous tone that was shifted in frequency to key the transmitter. A discriminator was used to demodulate the difference between a "mark" frequency and a "space" frequency. Non-harmonically related frequencies were used to minimize interaction between different circuits on the same telephone line due to harmonic and intermodulation distortion.
Although little documentation and other evidence remains of the first generation equipment used to control transmitters, we do have a couple of Northern Radio Company "Dual F.S. Tone Keyer" units. The panel below the keyer has labels that say "Monitor 11 3825 BL-5" and "Monitor 12 4675 337VV". We believe that this is an artifact of being used to monitor transmitters located in Building 1 from the control room in Building 2 or Building 2A. "BL-5" and "337VV" are the names of transmitters once located in Building 1. The Northern tone sets likely are the second or third generation tone set used at the station.
 |
| Northern Radio Type 153 dual tone keyer formerly employed to monitor transmitters 338VV and 361VV |
The "VV" transmitters used military BC-339 exciters, referred to as "V" sets. The overall transmitter, BC-339, BC-340 water cooled PA and RA-22 HV supply were referred to as the "VV" sets. Power panel switches remaining at Building 1 in Bolinas still carry labels for the VV transmitters.
 |
| Frank Spicer, Technical Supervisor at Bolinas, checking VV transmitters in Building 1 |
We do know that RCA installed their own 12-pair telephone lines between the receive site in Point Reyes and the Bolinas transmitter site and a second 12-pair telephone line between the Bolinas transmitter site and Sausalito where those circuits were handed off to the Pacific Telephone Company for transmission under the San Francisco Bay to RCA's office on California Street in San Francisco. Those circuits were used to key the transmitters in Bolinas and to transmit received circuits from overseas to the San Francisco California Street office for the point-to-point service and to directly key transmitters in Bolinas from Point Reyes for the maritime service. RCA maintained those private telephone lines until the 1970's when the cost to maintain the old telephone lines forced the replacement with a DS1 digital service.
 |
| Small portion of the circuit diagram showing dedicated circuits to San Francisco |
Because those were copper pairs RCA used common telephone techniques to increase the number of circuits. The center tap of two circuits could be used to "phantom" a third circuit. RCA also used the technique of isolating DC from AC to use DC circuits for on and off switching on the same copper pair that was used for audio circuits. We have evidence of how this was done in some of the older documentation that survived.
The top row of jacks are "LINE" and the left 12 jacks are noted "RS" and the right 12 jacks are noted "SF". Below the top two rows of jacks are repeat coils (transformers). Note that the center taps of the 12 left repeat coils are connected and feed transformers to phantom 6 more circuits on the 12 pairs. This allowed RCA to have 18 audio circuits between RS and BL in only 12 copper pairs.
 |
| Back of rack showing 111C coils and phantoms used to terminate dedicated lines |
When RCA expanded the Bolinas transmitter site by adding Building 2A they also built a new, larger control room. At that time they added state-of-the-art tone sets to control the new transmitters. We use the same equipment today to key the transmitters from Point Reyes. Details of that equipment will be covered in Part II of this series.
