Paul Thurst – Engineering Radio

The Automatic Electric 3 slot pay phone

This does not have much to do with Broadcast Engineering.

Recently, I saw a video about an engineer in Vermont who was fixing and placing single slot pay phones at welcome centers around the state. These are functioning as courtesy phones for travelers to use, free of charge. That is an interesting project.

My father worked for New York Telephone Company in various capacities from 1954 until he retired in 1988. We had a number of old telephone sets around the house which were removed from homes, old farms, businesses, and the like. One such telephone was a three slot pay phone from The Hitchcock estate in Millbrook, NY. This was a very unique item, as it dated from the Timothy Leary period of residence. It had white, blue, pink, and purple swirls painted all over it. At some point, G. Gordon Liddy, who was the then assistant District Attorney for Dutchess County, rented a Cessna and flew over the house to obtain a search warrant. After Leary was arrested, the group of LSD experimenters living in the estate house soon dissipated. My father went in and removed the pay phone. He said that those folks were living in squalor, “like filthy animals.” In fairness to my father, his Korean War experience while serving in the Marine Corps may have slightly colored his opinions about such things.

We called it the “hippie phone.” I believe it was a Western Electric model. After I moved out of the house I grew up in, all of those things disappeared.

I would have liked to have kept that pay phone.

All of that got me thinking about pay phones. Could there possibly be a similar pay phone for sale on eBay? Yes, there could. But they all seemed terribly expensive. Except for one reasonably priced unit, which I purchased. This is an Automatic Electric pay station, which is almost identical looking to the Western Electric unit from the same era.

When it arrived, I discovered the reason why it was less expensive; it is missing all of the internal parts. I was a little irritated by this, but soon discovered that it was typical. The phone company often removed the coin collecting relay and vault door as soon as the phone was taken out of service. The hook switch, induction coil, condenser, terminal board, and wiring harness were all removed, probably to be used to restore another phone.

Fortunately the coin slot and coin validating mechanism are still in place. If you drop a dime you will hear a nice “ting ting.” A nickle will result in a single “ting,” a quarter will make a satisfying “bong” sound from the gong.

I found a manual on line for the Automatic Electric pay phones which had a wiring diagram with the wire colors.

Unfortunately, there are no part numbers. I had a difficult time trying to figure out the correct parts to search for. I finally stumbled upon a Youtube video where someone was tearing down one of these units. That revealed the OEM part numbers for the coil (D284476A) and condenser (D68850A). With that, I was able to find the original manufacture’s assembly which included the hook switch, induction coil, wiring harness and terminal strip for a reasonable price on The Old Phone Shop website. I also ordered the condenser and a ringer (these phone had no internal ringers as manufactured). There is an option to buy an entire modern network board, but I wanted to keep this as original as possible.

The wire harness from the rotary dial was also missing. I used some wire from Belden 9423 cable to make those connections, keeping the wire colors shown in the schematic.

Next, I purchase a Pulse Dial to Touch Tone Converter. On POTS lines, most CO switches still recognize pulse dialing. However, POTS lines are expensive and I do not want to pay for one at my house. The answer is an Analog Telephone Adapter (ATA) with this converter. I purchased a Cisco/Linksys 2102 unlocked ATA on eBay for $16.00. Lowest cost DID number I could find is about $8 – 9 per month. There are directions on how to DIY a pulse dial to touch tone converter, but I simply purchased one on Amazon.

A bit about Plain Old Telephone Service (POTS) lines. It is a circuit switched data system. These systems are very simple and consist of a two wire circuit between the Company Office (CO) and subscriber Point of Presence (POP) known as a local loop or subscriber loop. The circuit is made by closing the Hook Switch. When the circuit is open, there is nominally -48 VDC across the Tip and Ring. When the Hook Switch closes (goes off hook), that drops to about 7-9 VDC. When the circuit closes, the CO switch detects a current flow and puts a Dial Tone consisting of two tones, 350 Hz plus 440 Hz on the loop. The subscriber can then dial a number and be connected to another party through the switch and what ever local or long distance carriers are needed to complete the call. On the other end, the CO switch will put a 20 Hz 90 VAC ring voltage down the line to ring the receiving subscriber’s phone. If the distant station is already off the hook, a busy signal is sent to the call originator, consisting of 480 and 640 Hz tones.

Internal to every POTS phone is a network consisting of an inductive coil and a capacitor or two and a resistor or two. The purpose of this is for line equalization and to create a two wire to four wire hybrid. The four wire hybrid is what feeds audio to the ear piece and receives audio from the handset microphone. It cancels echo from the distant station and provides a small amount to “side tone” or audio from the microphone so the user can hear themselves when they talk. The audio pass band is from about 300 Hz to 3,300 Hz.

In addition to that, some type of dial mechanism is required; either pulse or DTMF (Dual Tone Multiple Frequency). Pulse dial interrupts the closed Hook Switch in rapid succession (39 ms on, 61 ms off) to send a number to the switch at the CO.

DTMF consists of the following tone pairs:

Number	Tones (Hz)	Number	Tones (Hz)	Number	Tones (Hz)
1	697, 1209	2	697, 1336	3	697, 1447
4	770, 1209	5	770, 1336	6	770, 1447
7	852, 1209	8	852, 1336	9	852, 1447
*	941, 1209	0	941, 1336	#	941, 1447

Provided the local loop stays intact, the POT system is extremely reliable.

The Analog Telephone Adapter (ATA) converts a POTS from a circuit switched system to a packet switched system, then uses the public internet to complete a call. ATAs do not accept pulse dialing, hence the Pulse Dial to Touch Tone Converter. Where ATAs often fail is the ring voltage and tip/ring polarity.

Back to the task at hand: When the parts arrived, they all matched the mounting holes in the back plate of the phone. That is a good sign. I took the rotary dial off and put the new wires on. I also took the time to de-gunk and clean the mechanical parts of the rotary dial, then re-oiled it. It works as well as the day it was manufactured. The old mechanical phone system parts were built to last a long, long time.

The Hook Switch, Terminal Board, harness and condenser all fit into the phone almost perfectly. The condenser was hitting against the coin validator which didn’t seem quite right. I moved the condenser down to where the coin relay was located so it didn’t get scratched up.

Automatic Electric 3 slot pay station with parts installed

I also bought a small ringer, which I installed where the coin relay was. I purchased a replacement vault door and lock. This is a plastic unit, not a OEM part, however, it does the job and I am not worried about anybody prying the vault door off and stealing all the change. Putting it all together, it looks great! Now I need to find a used phone booth to put it in. There are a few of those on eBay as well, but very pricey.

Automatic Electric 3 slot pay station, restoration complete

I took the phone to the office and plugged it into a POTS line and called my cellphone. The old carbon element microphone might need to be replaced, other than that, everything works great. It is now worth considerably more than I paid for it, which is also nice.

After 10 years, it is time to move

W277CJ, Pittsfield was originally a translator for WUPE-AM 1110 KHz. Since that license has been surrendered to the FCC, it is now a translator for WBEC-FM HD2 which is simulcast of WUPE-FM, North Adams.

Confused yet? Don’t worry, it is a reshuffling of signals because the land under the 1110 KHz transmitter site was sold and the license turned in to the FCC. Something that I think will happen many more times to many more stations in the coming years. This translator was first put on the air in June of 2015. This is the third part of a series, the first two parts are: More AM work, Part V and The Bext TFC2K broadband antenna.

The translator recently moved it off of the Holiday Inn (formerly Crown Plaza) in downtown Pittsfield, MA to the WBEC AM tower. In order to make that move, we needed to do several things;

“sufficient measurements shall be made to establish that the operation authorized in this construction permit is in compliance with the spurious emissions requirements of 47 C.F.R. Sections 73.317(b) through 73.317(d). All measurements must be made with all stations simultaneously utilizing the shared antenna.”

These are intermodulation products, or third order products, between the two signals being transmitted, in this case W277CJ on 103.3MHz and WUPE-FM 95.9 MHz. The antenna side mounted on the AM tower serves as a back facility for WUPE-FM.

WBEC-FM backup and W277CJ connected to Bext FDCSDC-2 combiner

Those measurements are as follows:

(F1-F2) + F1 or (103.3 MHz – 95.9 MHz) + 103.3 MHz = 110.7 MHz
F2 – (F1-F2) or 95.9 MHz – (103.3 MHz – 95.9 MHz) = 88.5 MHz
F2 +F1 or 95.5 MHz + 103.3 MHz = 199.2 MHz

In order to make those measurements, I used two Microwave Filter Company MFC-6367 notch filters to attenuate the carriers on 95.9 and 103.3 MHz. This keeps the spectrum analyzer from overloading, thus lowering the analyzer noise floor and giving better results.

Various tools for proofing FM installations

Over the years, I have collected various parts to assist in getting good measurements for FM proofing. Going clockwise and starting at the top, the Rhode Schwarz NRP-Z11 power sensor, the MFC-6367 FM notch filters, directional couplers with power extractor element, various attenuators including the HP 255C variable 0-12 dB unit, and in the middle are two Mini-Circuits NHP-200+ high pass filters. The high pass filters are great for measuring harmonics.

To measure the third order products noted above, I first measured the carrier without the filters and an appropriate pad to get a carrier reference level. Then installing the MFC-6367 filters to measure the third order products. In addition to that, harmonics of both FM transmitters out to the 10th harmonic. Of particular importance is anything in the cellular or mobile data bands. All of these measurements were well below the -80 dBc threshold required by the FCC.

All of these measurements were well within the limits established by FCC part 73.317.

Also, because this is mounted on an AM tower, there are some AM things that needed to be completed:

“The AM station identified below may be affected by the facilities authorized by this construction permit. Pursuant to Section 1.30004 of the Commission’s Rules, at least 30 days prior to commencement of construction of the facilities authorized herein, the permittee must provide notification of the construction to the AM station licensee. As part of this notification, the permittee must examine the potential impact of the construction of the authorized facilities on the AM station using a moment method analysis. The analysis shall consist of a model of the AM antenna together with the potential re-radiating tower in a lossless environment. The model shall employ the methodology specified in Section 73.151(c) of the Commission’s Rules, except that the AM antenna elements may be modeled as a series of thin wires driven to produce the required radiation pattern, without any requirement for measurement of tower impedances. If the AM station was authorized pursuant to a directional proof of performance based on field strength measurements, the permittee may, in lieu of the moment method analysis, demonstrate with measurements taken before and after construction that field strength values at the monitoring points do not exceed the licensed values.”

Since this station was proofed several times, we did about ten readings along the monitor point radials, both before and after.

The new isocoupler was properly mounted:

This is simply a large coil of 7/8 Coax wound inside of a PVC form.

These AM antenna systems are a regulatory nightmare. Although the Moment Method is an improvement over the system of field measurement proofs, it is still complicated. Part of the issue with AM in general is the expense of the the antenna systems, particularly anything that is directional.

No real research into Medium Frequency antennas and propagation has been done since the 1930’s. Perhaps we know all there is to know about it, then again, perhaps not. I am currently working on a project which will study Medium Frequency propagation, which I feel, is the first step into revisions of antenna design.

W2XMN Alpine, New Jersey

This historic site is where Edwin Armstrong continued developing FM radio starting in 1936 after his fallout with David Sarnoff of RCA. There is an active experimental station on 42.8 MHz, WA2XMN. It is usually on the air to commemorate Armstrong’s FM demonstration to the FCC, which happened on 17 June 1936.

Last Thursday, June 19th, I had a very enjoyable afternoon at SBE Chapter 15’s annual summer barbecue. Over the years, I have probably passed by hundreds of times, often thinking, someday I’d love to see what is there. I can say, I was not disappointed.

The site was constructed in 1936, after Edwin Armstrong was asked to remove his FM equipment from the Empire State Building by RCA. At that time, RCA was more interested in TV development and thought that FM had little place in broadcasting. Boy, were they wrong.

Armstrong had this 400 foot tower built and used this building as his test bed. Broadcasting on two frequencies; W2XMN on 42.8 (later 44.1) and W2XEA on 92.1 MHz (later KE2XCC on 93.1 MHz) with power of 50 KW.

W2XMN building, Edwin Armstrong’s laboratory

Today, the WA2XMN broadcasts are produced by a replica GE BT-1-B phastron transmitter built by Steve Hemphill, with a power output of 250 watts. Unfortunately, there is a large hill between that site and my house, so I cannot hear it very well from my radio room. However, during the broadcast, there were several reception reports from around the area, including out on the eastern end of Long Island.

WA2XMN transmitter on 48.2 MHz, Alpine, NJ

The story of FM and Edwin Armstrong is a mixed tale. He was a professor of Electrical Engineering at Columbia University in New York. A creative engineer, having several patents to his name including regenerative receiver, super hetrodyne receiver, wide band FM, and early FM radar, he had a close relationship with David Sarnoff, the president of RCA. Unfortunately, that relationship soured over patent infringement lawsuits and Armstrong eventually committed suicide in 1954.

After which, the site ownership went to Columbia University. It sat, mostly unused, until it was purchased by the current owner (K2 Communications) for use as a land mobile transmitter site. Thankfully, they knew what they had and preserved the historical features. They are currently working on turning the W2XMN building into a museum.

Columbia University has an extensive archive of information about Edwin Armstrong. More information can be found here: Alpine NJ, Alpine, NJ Part II, Closing of Station KE2XCC

There is also this excellent article in the Short Wave Listening Post: WA2XMN revives Armstrong’s Legacy with 90th Anniversary Broadcast

If you have the time, This Week In Radio Tech (TWiRT) has a Youtube Video: Armstrong FM at Alpine (it is about an hour long).

I always recommend, for those interested in the history of FM broadcasting; Empire of the Air by Tom Lewis.

And a final thought; there is nothing in this job that is worth your life. Suicide is a permanent solution to a temporary problem. Most engineers I know (myself included) are introverts. We tend to take responsibility for things which do not really belong to us. One of the things that I learned, over time, is to enforce boundaries. It can be difficult to do with the electronic connectivity today; the ability for people to interrupt your life via text, email or phone call. Sometimes those things are important, but too often, not. When this happens, use your words. Respectfully tell the other person that you will deal with it during business hours. If that doesn’t work, remember; there are many, many open positions for broadcast engineers these days.

The Alford Antenna

I have been doing a bit of work in this building recently.

5th Avenue Lobby, Empire State Building, New York, NY

Therefore, when I saw the opportunity to acquire a piece of NYC radio history, I took it. The SBE Ennes Workshops are designed to bring affordable education to the Broadcast Engineering community. One major problem in the Broadcast Engineering field is the aging workforce. Any resource that can spur interest by younger people is important. Named after Harold E. Ennes, the Ennes Educational Foundation Trust funds scholarships, workshops, seminars, and helps underwrite the costs of producing text books. Making a donation by bidding on an element of the Alford antenna was an easy decision.

The Alford antenna was manufactured and installed in 1965 on the 102 story observation deck of the Empire State Building. It was designed by Andrew Alford, an electrical engineer, who designed antennas for ILS and VOR aviation navigation systems. It is the first purpose build combined FM master antenna system in the world. Both the antenna and combiner system were novel ideas at the time.

The master antenna was becoming a necessity because the spire and tower on top of the Empire State building had become over crowded with FM and TV antennas. In the 1931, RCA/NBC leased the 85th floor of the ESB for their development laboratory.

Circa 1936, RCA television transmitter on 85th floor, from the Early Television Museum

Once both services became commercial, the FCC forced RCA/NBC to allow other FM and TV stations to locate on the building.

According to the IEEE paper written about it, the idea of a master antenna came about in 1959. It was not until the mid 1960’s that the first three stations agreed to the idea.

The requirements for the new antenna included:

1) System should accept the output of seventeen different FM station transmitters at a power level of 10 kW each.
2) The antenna should radiate an essentially omnidirectional signal in both horizontal and vertical polarizations in approximately equal amounts.
3) The antenna gain and multiplexer losses should be such that each station can achieve an ERP of approximately 5.5 kW in both polarizations (the maximum power presently authorized by the FCC for that height above ground and that location).
4) The antenna system should be well matched over the entire frequency band from at least 92 to 108 MHz and preferably from 88to 108 MHz.
5) The external portions of the antenna should be deiced.
6) System VSWR should be 1.1 or less at each station’s input over a 200 kHz band centered at each station’s carrier frequency.
7) Isolation between antenna and other antennas on building should be at least 40 dB.
8) Isolation between FM transmitters connected to the system should be at least 40 dB for adjacent FM carrier frequencies and at least 36 dB for non-adjacent FM carrier frequencies.
9) The frequency-phase characteristic of the system as measured at each station’s input should be linear within ±50 throughout a 200 kHz band centered at each station’s carrier frequency.
10) System should allow for any number of stations to participate at the beginning and should allow for additional stations to be added from time to time as required up to a maximum of seventeen in any arbitrary order of carrier frequencies.

IEEE TRANSACTIONS ON BROADCASTING, VOL. BC-13, NO. 3, JULY 1967, The Empire State Building Master FM Antenna.

That article is an interesting read.

This is one of 32 elements of that antenna. It was in position 10L, according to the name plate. Date of manufacture was 7-65. It is 81.28 x 139.7 x 13.9 cm (32 x 55 x 5.5 inches) and weighs about 45 KG (100 lbs). The RF input is 1 5/8 inch EIA flange, located in the middle of the mounting plate.

Of course, the covers had to come off so I could see what was inside. The antenna is cast from non-magnetic (AKA stainless) steel. The elements were installed with the covers down, as they are not water proof and have weep holes to allow condensation to drain out of the element.

Inside contains the resistance heating element. Interestingly, the return wire for the element is tied to ground and does not have an insulated path to neutral. Each return wire was burned open, meaning that at least this antenna element had no heaters. In the northeast, that will certainly lead to problems.

I wonder if that was one of the reasons to replace it with a new master antenna. Other reasons would be that newer antennas have better circular polarization, more even signal pattern giving better reception, better gain characteristics, better combining systems with more isolation and the ability to pass HD Radio side bands, etc.

No doubt, this antenna section will end up in my radio room, but I am still working on how I will display this interesting piece.