The Long Lines are a remarkable engineering achievement in two senses:
1. They are a milestone of microwave RF transmission coming to maturity, and being widely and economically used. This includes such aspects as analog transceiver front-end design, antenna design, reliability know-how etc.
2. Long Lines managed to relay analog voice calls over dozens of hops with enough quality to allow low-distortion human and early fax/modem communication over thousand of miles overall. This is an amazing achievement for the American society and economy at large.
Just to highlight how far we've come in terms of data capacity, a single TH-2 Long Line link could carry 900 calls. STM-1, the slowest standardized carrier grade payload framing currently in use (predominantly over fiber), carries 63 VC-12 containers, each with an E1, for a total capacity of 1890 calls.
Of course, the above is not an apples to oranges comparison as they use different transmission media. I was not able to locate information about the bandwidth of specific Long Line link types to calculate RF spectral efficiency, but the ballpark figure is an improvement of ~100x when compared to modern digital microwave links of similar bandwidth (accounting for both modulation and digital voice compression).
For people in the Bay Area, there's one of those towers/installations in Redwood City, it's a large concrete tower there. It's now a data center building. It still has the microwave horns in it. [1]
The Bay Area features several unusual and distinctive microwave sites, far different from the modular construction normally used by AT&T. Part of this is because some of these sites were relatively early ones, part is because of local zoning requirements, and part is because some of them were originally built by GTE. Perhaps the most interesting is the Oak Hill site, which I believe is nearly unique, with just one other of a similar design.
The Redwood City site is interesting in its own way, as I believe the building below contained the Redwood City tandem at the time prior to its elimination in favor of a single, very large 4ESS tandem at Oakland. It was not very typical for tandem switches to be colocated with large microwave sites, usually being end offices on the microwave network with only two antennas (tandem offices are typically in large cities so AT&T often used a "city and junction" pattern where the tandem office, named for the city, was an end office on a short leg from a large microwave site on the outskirts of town named city junction). The exceptions tended to be in highly urbanized areas where space was difficult to obtain, or a few odd cases like Mojave where an end office grew over time into a gateway site.
You see something similar around the UK, and particularly in Scotland where the terrain wasn't conducive to long coaxial cable links - mostly squat squarish towers that used to hold either microwave horns or (later) dishes.
You can see one in the link below, the tower on the left. The one on the right was (when I used to maintain kit up on this site) run by Scottish Power as part of their communications network, with tower space leased to other customers. Just visible at the top of the left tower are two 156MHz folded dipoles used for VHF marine communications in the Clyde estuary, which went to two old Philips VHF transceivers in the wooden shed you can see in the enclosure behind the fence. That was replaced by a metal shed about ten years ago when I went to site one day and found that a combination of rotten wood and the depredations of rabbits along the bottom of the timbers and a particularly severe gale had made it "sink" by about 18"!
Some of the sites on the west coast were quite interesting, with Mallaig having UHF and VHF links stretching 30 miles or so to remote islands with no real line-of-site. The VHF link to Canna just south of Skye carried a single call on around 80MHz (I think it was 84MHz/76MHz split, can't remember) which fed a tiny ten-line telephone exchange. There was a six-channel link to Rhum, a little to the east, which was a much larger island and had a more conventional exchange.
All gone now, and done using IP over 6 or 24GHz microwave links to little 1U servers in the exchanges.
I climbed one of these towers with a friend a couple of years ago. The antennas had been removed, or had fallen off, but the tower itself was still in good shape. Up on a ridge line, it offered an amazing view; we got to look down on a couple of military jets, out flying practice maneuvers in the valley below us.
The service building at the base of the tower was unlocked and empty. With a couple hours of cleaning and mopping, it would have made a fabulous spot for an all-night rave. Alas, the pandemic put all such plans on hold; in the meantime, the site was demolished. There's nothing there now but a gravel patch.
I'm glad to have had the chance to appreciate this bit of infrastructure up close before it vanished.
Ha, I was just about to post the Internet Archive's link to Long Distance and you beat me to it.
Even though an AT&T promo, this is a great documentary. It's an excellent exposé of the forerunner to AT&T's post-War Long Lines—the cable and wireless systems—which was pretty much state of the art before the Long Lines microwave system superseded it.
The doco is particularly useful for historians of tech/communications because it provides an excellent overview of how the US's telephone comms worked by the outbreak of WWII—from the women who ran the network to images of line, terminal and transmitting equipment. It's and excellent snapshot in time.
Two notable things are the huge number of women employed to do the manual routing—Strowger switching being essentially limited to towns and crossbar/long distance dialing yet to be deployed—and the methods used to lay underground cables. Women's employment 80 years ago was very different to today and the film illustrates the fact; on the other hand, there's a remarkable resemblance between the drums of cable and cable laying techniques of 1941 and the laying of fiber optic cable of today—those 1941 cable drums could equally have carried fiber.
Some things never change.
Incidentally, I used to work on microwave television links.
I’d love to know more of the blow-by-blow history of the internet’s built environment. I feel like the story must be something like “long lines terminated at telephone exchanges, internet exchanges naturally located near telephone exchanges, and colos like One Wilshire located near internet exchanges?
Another fun example of this is 111 8th Avenue. It’s all Google territory now but it was a pretty interesting free for all of various tech adjacent things in the late 90s dotcom era.
In 1962 I had a summer job in engineering with Bell of PA (well before the breakup of AT&T) so I have some knowledge about the long lines network.
Interestingly, at that time AT&T owned many of the local operating companies (e.g., Bell of PA which actually owned Diamond State Telephone Company, the operating company that covered Delaware). There were many other operating companies that were not AT&T owned. IT&T was an example of one that owned lots of local operating companies. How they interfaced, both technically and financially, with AT&T is beyond my knowledge.
Long distance phone calls were routed through a hierarchy of several levels of aggregation, ultimately to something like as few as ten (recall is hazy) top-level centers. These had long distance microwave communication from one to the next, where the call would then be routed downward through its hierarchy to the local destination. I don't remember the specific nomenclature for the top-level centers. One is in downtown Wayne. You can see the tower on Google Maps or Google Earth at West Wayne Avenue and South Bellevue Avenue. It used to have horn antennas, but now has various other dishes and antenna structures.
As I recall, the horn antenna can be thought of as being a sector of a full dish. At the frequencies then in use, the full dish would have been huge, probably too big to put up on a tower. However, for a reasonable signal to noise ratio, the directionality you get with a dish would have been important. So they used part of a dish.
I don't know whether in general, they didn't have hazard lights. I'll have to check some night when I'm out, but I'm pretty sure the one in Wayne does have lights. Both the FCC and the FAA have regulations regarding when lights are required. If I remember from my ham radio days, one of the criteria had to do with the height of the antenna and the distance from an airport.
> The first coast-to-coast telephone call was made August 17, 1951, using the Long Lines — sometimes known as the “Skyway” or “Telephone Skyway” — network.
It's odd to think about how a long-distance phone call was limited before then. We take it now for granted that this is possible, even cell coverage works well in most rural areas in the US now.
That’s not strictly true. That was the first coast-to-coast call on the long-lines system. The first cross country call was made by Alexander Graham Bell to his old assistant Thomas Watson in January 1915. From then until November 1951, this was an operator-assisted process (and much longer than that for much of the country).
You would call the operator, tell them who you wanted to talk to, the city they’re in, and their phone number (if you knew it). If the call is close enough, you’d stay on the line while they connected you. Otherwise, you would be instructed to hang up and wait for an operator to call you back and let you know they’d gotten your party on the line (or not). This could take 5 minutes, or a few hours, depending on the distance, complexity of the route, availability of trunks, etc.
By the 1920s you could even call London from the US, but the radio link only had capacity for like 12 calls and it worked better in the evening. There was also a network of ship-to-shore radiotelephone links that operated on HF and VHF.
It would have been entirely possible, but extraordinarily expensive to have a multi-party conference call between someone in London, someone on a ship in the Pacific, a third person riding a train between New York and Washington DC, and a fourth on a mobile phone in a car in Chicago at least as early as 1947.[1]
I’d ballpark the cost of that call at around $100/min in 1947 dollars.
The Wikipedia article on long distance has an excerpt from Dragnet of a call from LA to a rural party line in Utah. It’s worth a listen.
[1] Transatlantic telephone service began in 1927 between NYC and London, VHF mobile service began in 1946 in St. Louis and then Chicago, telephone service on trains began in 1947 (at least as provided by the Bell system, independents and train companies may have gotten there first), ship-to-shore via AT&T High Seas Service was available by the mid 1930s.
I recently read this article while searching for how these towers are used in 2022. It's truly a work of engineering genius for the era. There are still many of these towers around the place where I live if you drive up and down the interstates. It's why I started searching for them to begin with.
It's a lot of fun driving around and spotting these once you know what to look for. There's a continuous line of them heading out on I90 from Seattle through to Idaho. Some have been repurposed for cell phone use. Occasionally the towers and bunkers come up for sale.
Quite a few of the early concrete towers are still standing in northern Illinois. There are also old tower sites not on the article's map in my area. I suspect many towers around here get new leases on life for data links, and the old long-distance switching center outside Norway, IL still has a mast with an impressive number of antennas.
My grandfather Fred executive produced the see it now broadcast mentioned in the article. It was noteworthy at the time because it was the first time someone showed both coasts of the US at the same time on the same screen, live.
I don't know about these specifically, but similar systems the military used are interesting in that they have components you didn't see much in other systems. Like "traveling wave tubes" and klystron based amplifiers.
The systems I'm familiar with were at temporary "forward air control" posts, where a radar system was installed relatively close to the front lines. The portable microwave transmitters were used to relay the radar data and voice calls back to the generals further back from the front lines.
Reminds me of this video I came across a few months ago. Its a tour of an old AT&T bunker in the process of renovations. Pretty interesting to see how well setup and large they are.
Microwave links are still heavily used, especially for cell tower back haul. Here's a map of them in western Switzerland: https://s.geo.admin.ch/9c26f359f1
There's one of these in the Hillcrest neighborhood of San Diego. Was always fascinated by it when we went there for dinner or whatever. Haven't seen any in NC but there's a few locations on the map. Need to make a trip I think.
> One of the new innovations was the use of fiber optics. Fiber optic lines are typically located underground, eliminating the vulnerabilities of earlier coaxial-based cable systems.
1. They are a milestone of microwave RF transmission coming to maturity, and being widely and economically used. This includes such aspects as analog transceiver front-end design, antenna design, reliability know-how etc.
2. Long Lines managed to relay analog voice calls over dozens of hops with enough quality to allow low-distortion human and early fax/modem communication over thousand of miles overall. This is an amazing achievement for the American society and economy at large.
Just to highlight how far we've come in terms of data capacity, a single TH-2 Long Line link could carry 900 calls. STM-1, the slowest standardized carrier grade payload framing currently in use (predominantly over fiber), carries 63 VC-12 containers, each with an E1, for a total capacity of 1890 calls.
Of course, the above is not an apples to oranges comparison as they use different transmission media. I was not able to locate information about the bandwidth of specific Long Line link types to calculate RF spectral efficiency, but the ballpark figure is an improvement of ~100x when compared to modern digital microwave links of similar bandwidth (accounting for both modulation and digital voice compression).