How did transatlantic telephony of AT&T work in 1942? Where was the radio station? In New York?

I am studying the A3 scramble machine Roosevelt and Churchill used, which was replaced with the SIGSALY machine in 1943. Both had to transmit the call via rhombus antennas across the ocean.

Where was the radio station? But where were the rhombus antennas positioned? Not on top of the Bell building on West Street in NYC.

  • By "work" are you looking for schematics or something else. Yes, it was in NY as you could have found out using google quite easily (as I did).
    – Andy aka
    Commented May 4, 2023 at 13:48
  • 2
    Not by cable: the first transatlantic telephone cable was placed in 1956. The first phone 2-way "phone call", via radio, was in 1927.
    – Davide Andrea
    Commented May 4, 2023 at 13:54
  • Of course not by cable. But where were the rhombus antennas positioned? Not on top of the Bell building on West Street in NYC. Any clues?
    – Chuck
    Commented May 4, 2023 at 13:59
  • 1
    It's not really an EE question suitable for this site IMHO.
    – Andy aka
    Commented May 4, 2023 at 14:04
  • 3
    Okay Andy but you found someting using google. Why not share it?
    – Chuck
    Commented May 4, 2023 at 14:07

2 Answers 2


The A3-scrambled signals were carried by the ordinary civilian telephone network (AT&T and GPO in the US and UK), and the SIGSALY messages were carried by the military communications network (in particular, the US Army Signal Corps), which used different physical facilities. But of course some of these facilities might have been leased from the phone company.

For the telephone network: the transmitting rhombic antenna array was in Lawrenceville, NJ and the receiving array was in Netcong, NJ and later Manahawkin, NJ. I believe the corresponding UK antennas were in Baldock. The ground links were carried by AT&T and GPO long distance land lines. Thus: White House to the Washington DC long distance operator, who connected to 24 Walker Street in Manhattan where the transatlantic switching took place, then to Lawrenceville, then via radio to Baldock, then wire to a switch in London, and so to 10 Downing Street, and similarly back to the White House, but via Netcong.

A detailed summary of the AT&T side of things, from the pre-rhombic era, is in a paper Transoceanic Telephone Service-Short-Wave Stations Planning and Construction of a Short-Wave Radio System by F.A. Cowan, an AT&T big-wig.

I will dig up my references and supply sources in the near future.

  • 1
    Interesting that the transmit and receive were so far apart (~50 miles, 80 km). Certainly makes it easier to separate the signals. Similarly on the UK side, Baldock to Cooling is a bit further, but not by much.
    – Jon Custer
    Commented May 4, 2023 at 15:02
  • The AT&T antennas were sited before the development of the rhomic antenna, and care was taken to put them in rural areas with few automobile spark plugs. I am not 100% certain about the UK antenna sites, as there was rapid technology change and so on. In the pre-rhombic period the UK receivers were at Cupar, and maybe at Rugby later in the 1930's. Commented May 4, 2023 at 15:14
  • Thanks. Great info. It is my understanding that they used rhombic antennas in London to set up the A3 scramble call with Washington. @kimchi do you think they used rhombic antennas in the US as well? I found somewhere else the following: ”SIGSALY was transmitted as a multiplexed data stream via an independent sideband HF circuit, between Scotland and Rocky Point if my memory does not fail me". Do you think this is not correct.
    – user61079
    Commented May 4, 2023 at 16:20
  • @Chuck Rhombics were used for both transmitting and receiving in the USA on the AT&T network. Cupar in Scotland and Rocky Point were termini for an earlier (1927) AT&T/GPO radio telephone link, and might well have been also used by the US Army Signal Corps to carry HF traffic, but I don't know. Commented May 4, 2023 at 17:03
  • 1
    Details of the design of the rhombic antennas are in ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=6772753 - no mention of installation at the sites described here. Details of the amplifiers at Lawrenceville were published in ieeexplore.ieee.org/document/1696703
    – Jon Custer
    Commented May 4, 2023 at 18:16

A Fourmilab web page (not Fermilab mind you) has information on the system. They quote from a page at the Antique Wireless Association (although I cannot locate that page):

Cooling Radio Station was at the UK end of a point-to-point, shortwave signal beamed from Lawrenceville, New Jersey. The site of the station was carefully selected as the antenna, MUSA (Multiple Unit Steerable Antenna), upon which it depended to receive the incoming transmission, had to be: directly aligned with Lawrenceville NJ, USA; two miles long; comprised of an array of 16 individual rhombic antenna; and have an area of three miles in front of the MUSA that would be free from radio interference

They also link to a Bell System Technical Journal article about the system if you are interested in the technical details.

So, the antennas were in Lawrenceville, NJ (about midway between Trenton and Princeton) and most certainly would not have fit in New York City. A Princeton Online link has some information about the Lawrenceville site, as does the Hopewell History Project.

  • 1
    Thanks Jon, any more information would be highly appreciated
    – user61079
    Commented May 4, 2023 at 16:44
  • @Jon Do you know of any good photographs of rhombic antennas? I like the one of the AT&T antenna in Dixon, CA, from about 1930, taken in a kind of weather that makes the wires more visible than they they usually are, but few reach that quality. Commented May 5, 2023 at 0:49
  • Hey Jon, watched the video of the Antique Wireless Museum. Wow, what a rich historical source. Extremely well showing the great effort it took to make transatlantic calls possible. Thanks a lot!
    – user61079
    Commented May 5, 2023 at 7:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.