Cold War Machine Cryptography

Written by Godfrey Dykes

© RN Communications Branch Museum/Library

  It is generally agreed that the Cold War started very soon after the end of WW2 and finished {has it?} in the early 1990's. Given that definition, I joined the Royal Navy as a 'communicator' when the Cold War was well and truly established being fifteen years old in 1953! I left in 1983 having spent the whole of my career under the cloud of the Cold War {but with many small 'hot wars' added in for good measure}, and during that thirty year period, for the most part, held a PV [positively vetted] security clearance, and dealt with the coding and decoding of signal traffic up to an including 'top secret' with added special-handling caveats. Whilst we were trained to use high-grade book [manual] coding procedures like OTP [One Time Pads], and emergency low-grade [manual] codes like SEASCOUT, they were very rarely used operationally, and the bulk of our work was done using cryptographic machines.

  Because we were operating in a NATO environment plus other Organisations like SEATO and CENTO against the WARSAW PACT countries, each NATO, SEATO etc navy had the same 'crypto machines, which were used to process signals [traffic] for NATO purposes inter alia or for NATIONAL purposes, the crypto-train being dictated by separate [to the machine that is] 'keying material' in the form of highly secret keycards. The cards gave the settings for those parts integral to the machine which were operator programmable, a function/operation normally performed on a 24-hour basis or as ordered for operational reasons.

  More of that in a moment. In the meantime - also to be explained separately - we were aware, obviously, that there were many others with a need to code their sensitive information, and clearly there must have been a commercial company manufacturing cryptographic machines for sale to international clients and governments not directly involved in the Cold War, or were they? Equally clearly, our machine's, and quite separately the keycards, were designed, developed and manufactured by processes controlled from beginning to end by National Security Agencies and were not available commercially! The keycards in particular were of course checked and mustered at least on a daily basis, and destroyed under strict supervision when time expired or when compromised. In one of my ship's, the frigate Rothesay, I took part in the musters as well as in the use of these keycards, only to be informed that shortly after leaving the vessel a new officer joined the ship, a Sub Lieutenant Bingham, and part of his duties were these musters and checks. Imagine my surprise when the officer was declared a Russian spy and was jailed for a long period in 1972 - 21 years I believe!

  Always interested in the technicalities of my 'tools of trade' I had read several articles about cryptography which means secret writing, a manual written transcription. In 1889 a man called WIER invented the first commercial coding machine as we in the navy had a 'crypto machine in use by 1912 - see this file The RN's first operational cryptographic machine I was au fait with the name Boris Hagelin, a Swedish mathematician of note, and from his name, the sign-posts led to Switzerland and to his Company 'Crypto AG' engaged in the production of esoteric devices. Because he was wholly commercial, and I thought a scientist outside of the defence remit, I took little notice totally unaware that he was perhaps more important to the USA, the West and to NATO/SEATO/CENTO than all the navies in the world put together in communications secrecy terms.

  Quite recently the CIA/NSA of the USA inter alia, have released now declassified papers covering Mr Hagelin and his machine's sold to and used by many countries and organisations around the world during the Cold War years, and even today, for his original company is still in business. At the end of this page you will find an MP3 audio file which I have copied and reprocessed from the public domain for you to use and learn from. Please note that the copyright belongs to the BBC.

  Before we view a picture of the famous Hagelin CX-52 system and you go on to listen to the MP3 file telling of the vast difference between it and the C-52 which I can assure you will really open your eyes as to the goings on during the Cold War, let's return to my R.N., days specifically to our use of cryptographic machines.

  My first ship, a castle-class frigate, with a relatively small wireless office, had an enormous 'crypto machine called a TYPEX, in this case a Mark 23 model. It was a British machine designed and used before WW2 prior to the many Mark's [modifications] which were made to it. In pre-war years the USN had its own machine, the ECM. When the USA joined the war after the Pearl Harbour attack of December 1941 bringing about the immediate need for all allied forces to communicate by coded signals, we offered our Typex machine to the Allied high command, but the Americans wanted to keep their own ECM and moreover did not want to share their machine with others however friendly. I never saw an ECM but the Typex was a truly big monster and I can understand the American's not wanting it. A compromise was sought and a device called a CCM [Combined Cipher Machine] was designed and fitted to each machine, Typex/CCM, which made them compatible. For all that, it was an irksome task both to code and to present a copy of the groups to be transmitted by Morse Code to an operator and to decode to present a plain language copy for the commanding officer to read. I can remember senior operators feigning other jobs-in-hand to try and avoid operating the Typex machine. The CCM device could be bypassed when using the machine for National traffic and the programmable drums were easy to reset from keycards. The term electromechanical described its operating mode added to which was the clunking noise it made. It was called an off-line equipment as was its successor machine, the KL7. All Typex machines were identical in build terms and all parts interchangeable. All CCM's were also identical.

  For my next ships and submarines* [see bottom of page for asterisk properties] from 1956 onwards, for the Cyprus EOKA crisis and the Egypt Suez Canal War until the very late 1960's, the Typex/CCM's successor, the KL7, ruled the roost. Its size was tiny compared with the Typex/CCM, and it could easily be packed away and stowed in its purpose-made carrying security case. As mentioned it was an off-line machine, and each KL7 machine was identical, using the electromechanical method of operating via moving drums/rotors which had several settings per drum, once again, taken from keycards changing at pre determined times. The electrical contacts on each drum [fixed brass studs one side and spring-loaded brass contacts on the other] were regularly cleaned with a special rubber [a non-abrasive eraser avoiding wear] and then lubricated to avoid sticking when rotating/stepping. Each set of rotors complete - fixed and variable programmable parts - were housed in a red fabricated box, and the norm was to set-up drums/rotors for a given day/date ready to receive or send signals coded or to be coded in different codes, using a British code and/or a NATO/SEATO etc codes. For special handling signals of a high classification/sensitivity and usually a high precedence, an appointed commissioned officer was the custodian of a special rotor-setting keycard locked into a safe external to the wireless office/crypto office all other operational keycards were locked into the operators safe's. On arrival in the vessel of these special signals, the custodian officer would be alerted and he would bring his rotor red box and the key-setting card to the office to have an operator set the rotors for the required date under his supervision. Once set-up, the rotors and keying material were never out of sight of the officer. The rotors were then offered-up to the KL7, the settings checked for accuracy and alignment, and the signal code was typed into the KL7 machine, usually by the radio operator, with the officer carefully masking the printed gummed-tape bringing out of the machine the plain language decryption. Once out, it was moistened [licked] and then stuck down on a signal pad, edited, and then with everything else, including the original signal in coded group form, taken out of and away from the office for the officer to show the commanding officer. If necessary, the plain language version was then typed [on the office typewriter or on the wardroom/captain's office typewriter] for the CO's files, and an acquaint-chit was the only record of the signal having arrived in the ship was kept in the wireless office.

  The great advantage of the KL7-system [known as Adonis] and all other systems to follow, was that one basic machine could be used for many purposes, and all that was required was that each holder of a machine be given only the rotor settings on a need-to-know basis, with senior fleet commanders having all levels. As a working example of such a distribution, NATO countries, paradoxically, were not always friends, indeed Greece and Turkey were positive enemies and had to be kept apart - see their war of 1897 here Greco-Turkish War: we used to say that we were fighting the USSR but the Greeks and Turks were still fighting one another. Both countries had KL7's but different keying material and neither had 'executive status'. I remember well that during the Cyprus EOKA problems, British warships would carry a Turkish soldier or policeman as the authority in case a Greek terrorist was arrested during the vessel's patrol of the southern coastline from Paphos to Famagusta, with Paphos being only 217 nautical miles from a major Greek island! Also, subsequent to Archbishop Makarios' deportation by the British to the Seychelles in 1956 as a trouble-maker, only to return to Cyprus later as its first President, the Turks invaded the northern territory of the Island and they are still there in large numbers today.

  During the War against Egypt's Colonel Nasser in 1956 I was the operator of a 5UCO [5 Unit Code] system sending traffic by short wave radio from Port Said in Egypt to Cyprus and to Malta. This system, when working properly, was designed to carry huge traffic loads, but was designed for high powered transmitters connected to efficient directional aerial systems, and accurate receivers displaced from the transmitters and its aerials by several miles. That can only be achieved between terra firma stations e.g., between Malta and Admiralty UK, where, in the case of the Admiralty, the communication centre using the 5UCO is in London, the transmitters nearly 200 miles away in Inskip Lancashire and the receivers, also 200 miles from London in Forest Moor Harrogate Yorkshire, 60 miles due east of Inskip - an ideal set up! The traffic intelligence [the AF - audio frequencies converted to DC pulses] were routed/interconnected over leased secure telephone cables. In our case afloat, everything was on top of the 5UCO operating position, with omnidirectional aerials so close to one another on the ship causing mutual interference. It was not a great success? See this file Suez - a Communications Overview and whilst on this page note the speed of the Morse Code used!

  In the mid to late 1960's the off-line machines were relegated to be used only for very special signals of great sensitivity, by the on-line machines which used digital systems instead of electromechanical systems, using noisy-diodes having a repeat cycle of approximately 200 years to encrypt and decrypt ASCI-Code 'active' conditions - 1khz tone and flip-flop circuitry to avoid selective fading. These were systems called Raleigh [BID580 - strategic area broadcasts inward bound] and Orestes [BID660 - tactical working between two or more stations/ships] working in synchronous or asynchronous simplex or duplex modes usually at 100wpm [75 bauds]. This meant real time encryption with signal-handling in plain language throughout. It brought immediate operational advantages in terms of legibility, instantaneous interpretation, distribution and filing with an increase in the speed of communicating not thought possible at the beginning of the 1960's. Just like the KL7 there were various levels of keying material as well as theatre/area operating differences, meaning that all the BID's were identical. Computers carrying high speed data used to fight the ships could now be transmitted between ships over the new highly accurate TX/RX systems to effect the best possible use of weapons and 'killing ability'.

  Then, and inevitable, in the early 1970's, the off-line systems were eventually put to bed with everything sent on-line, and the KL7 had no further role. To do the job of the off-line KL7 a device called the TLA was introduced into the BID660 [Orestes] system where classifications transcending cosmic top secret, were handled overtly in plain language. This involved literalisation/deliteralisations for on-line coding and decoding respectively. One of my last recollections was being almost dressed in paper tapes of 5-Unit Code from the TLA, sending high precedence, mega classification signals to the erstwhile Admiralty [defunct from 1964] and now [MOD[N] - Ministry of Defence [Navy] - when one of our submarines heading for the far east was stopped from using the Suez Canal by the Egyptian authorities.

  Of course we had other crypto systems which covered analogue to digital voice channels et al, BID 150/250, BID820, TSEC KG13 for example, and as technology goes, we are one of the leading navies of the world especially when one adds our vast experience acquired by being a world-wide blue-water service, communicating world-wide.

  Today, in the middle of the second decade of the 21st century, 'things' are magic and all done over various satellites, high and low level, geostationary and orbiting, at great speeds, with almost instantaneous reactions from shore-based commanders.

  Well, finally [and the reason for this page anyway] let's return to the 1950's and to cryptography as used by?.......well, let's say non Royal Navy and USN services. That leaves a vast number of other people, and so sit back and listen to this fantastic story of intrigue, by those, who by sheer cunning design, got the best systems available, and those who didn't or couldn't because they were not aware of what was possible or for sale, and so got a system which was so easily undermined and broken-into by crypto analysist's employed by the manufacturer and its agents/sponsors!

CX-52 HAGELIN (copyright BBC)

How to break and unbreakable code?

copyright BBC (28 minutes)

  *For those of you who understood the Cold War in terms of 'Check Point Charlie', the 'Berlin Wall', the 'Berlin Air Lift', etc which adversely affected the lives of Germans both East and West, I would ask you to remember how adversely our lives were affected especially those of our men serving in submarines. To that end, a new book has been released based on secret papers now declassified and recently released, telling the story of our one-to-one upfront incidences with Russian submarines. It mentions my submarines involved TURPIN and AURIGA, and is written by Iain Ballantyne a well known naval man, writer and journalist, and called 'Hunter Killers'. Published by Orion Books Limited 2013 [paperback 2014] of London and produced, printed and distributed by various English companies, it is assigned ISBN 978-1-4091-3901-0. I commend this authoritative, exciting and well penned book for you to read. Iain says on his page immediately preceding the Contents page.

  "To the Royal Navy submariners who fought and won the Cold War and their families, not forgetting the shipbuilders, the support staff and good friends in the United States Navy submarine force."

  For a more detailed account of all Royal Naval cryptography see Cryptography

  For further details a recommended link is Cipher Machines and Cryptology