HISTORY/DEVELOPMENT
Appendix 8 from the book "Betrayal At Pearl Harbor" by James Rusbridger and Eric Nave, provides and excellent account of the history and development of the Typex."As long ago as 1936, Lord Louis Mountbatten, then Fleet Wireless Officer with the Mediterranean Fleet, had recommended that the Royal Navy adopt a machine cryptograph for enciphering all its radio traffic as the German Navy had been doing since 1926 with the Enigma. The two Enigmas GCCS had bought in 1928 had lain idle while the Inter-Departmental Committee on Cipher Machines, formed in 1926, debated how best to use machine cryptography. But at the end of six years they had been unable to come to any decision. (Air 2/2720, Avia 8/355 & 8/356 and ADM 1/11770, PRO, Kew.)
In 1934, Group Captain O. G. Lywood, a Royal Air Force signals officer asked GCCS if he could borrow one of the Enigmas, and he took it to the RAF's Wireless Establishment at Kidbrooke, in southeast London. Together with Flight Lieutenant Coulson; Mr. E. W. Smith, the workshop foreman and Sergeant Albert Lemon, he set about building a copy of the Enigma using mainly parts from commercial teletypes then in service with the RAF.
The quartet busied themselves for three years and eventually produced a cumbersome machine they called the RAF Enigma with Type-X Attachments. Unlike the Enigma, which was the size and weight of a portable typewriter, Lywood's machine was huge and consisted of a standard Creed teletype with an Enigma rotor basket grafted on the front, which was machined from solid brass and alone weighed ten pounds. The entire machine weighed over 120 pounds and needed a 230 volt AC power source, unlike Enigma's battery system. The only advantage Typex had over Enigma was that it was able to print out the cipher and plain texts simultaneously on paper tape at fifty words per minute, whereas Enigma showed the text by means of lettered glow lamps that had to be recorded by the operator's assistant.
Lywood proudly demonstrated his brainchild to the Cipher Committee in 1937. They were unimpressed and refused to authorize any money to finance further development. Fortunately, the RAF decided to continue on their own and allowed Lywood to take his prototype to Creed & Company at Croydon, in south London, a small family firm run by the deeply religious Creed brothers, who led the work force in prayers before each day's work began.
Creed's main business was the production of electromechanical teletypes for the British Post Office (the forerunner of today's British Telecom). Since 1929 it had been owned by the ITT Corporation of America, but it seems that details of their work on Lywood's copy of Enigma did not get passed back to their American parent. With the help of Mr. F. E. Brake, Creed's managing director, and Mr. Kirk, their chief designer, twenty nine machines were built based on Lywood's prototype. These were called Type-X Mk I (the name Enigma having been quietly dropped) and were used to equip the main RAF headquarters. Creed then made a number of improvements and by 28 May 1937 had produced a much better machine they called Typex Mk II. This was shown to the Cipher Committee on 14 June, and they immediately approved an order for 350 Mk II machines at a cost of £107.8.0 each, which was enough to equip the entire RAF down to station level. A later attachment enabled Typex Mk II to produce punched tape using the standard five unit Baudot code (invented by the Frenchman Jean Baudot in the nineteenth century), but the Mk II could not work on-line with other Typex machines.
As orders grew, Creed transferred production to a new secret factory at Treforest in Wales. By September 1941, 3,232 Mk II machines had been produced at a cost of over £300,000. The Mk III Typex was hand operated and, although much smaller, still far more cumbersome than Enigma. Typex Mk IV was built around the Creed Model #7 teletype and printed out the text on rolls of message forms. Typex Mk VIII was the first model capable of interfacing with other Typex machines sending and receiving Morse code transmissions and automatically converting them into printed plain text. In 1942, after America entered the war, Typex Mk VIII was modified by Commander Don Seiler, USN, to interface on line with the American M-134 cryptograph known as ECM to the U.S. Navy and Sigaba to the U.S. Army. With the Typex converter, these became known as Communications Security Publications CSP-1700 or more usually as the Combined Cipher Machine (CCM). The reason for Seiler's converter was that U.S. regulations prohibited any foreign personnel from handling or seeing American cipher machines.
Bletchley Park display: This is a post-war Typex Mk22 was lovingly restored to functional order and with two types of plugboards. The original plugboard jumper wiring was black, not yellow. This excellent recreation using modern wires illustrates the plugboard wiring. A good operator could do 40 wpm on the keyboard. (Bletchley Park photo via Wikipedia) |
Typex Mk II used five active rotors plus one static rotor (similar to the Enigma's umkerwalze or reflector). The two entrance rotors, IV and V, were stators and, once initially set, did not move during encipherment. With its three moving rotors, two stators, and a reflector, Typex was plainly a copy of Enigma and infringed their patents. A long correspondence between the Whitehall bureaucrats began as to who would have to pay royalties to the Berlin factory. This was still going on when war began.What Typex did not have was, of course, the stecker or plugboard, because that had been added only to the military version of Enigma and no one at GCCS had ever heard of this improved model. GCCS had refused to cooperate with French and Polish codebreakers before the war, and it was not until 25 July 1939 that members of GCCS visited the Polish codebreaking center near Warsaw and for the first time saw their copy of the German military Enigma. The Army took thirty machines from the RAF order for field trials, most of which they left behind on the beaches at Dunkirk in 1940. Fortunately the Germans saw they were copies of the old commercial Enigma and simply took them into service, as they interfaced without difficulty with their older Enigmas; and as they believed Enigma was unbreakable, they made no attempt to see if Typex could be broken.
If properly used, the Enigma was unbreakable, and indeed several keys were never penetrated by GCCS throughout World War II; but the machine suffered from three basic problems. First, by the time World War II began its hardware technology was already ten years out of date, so that, for example, the most secure version used by the German Navy came with a set of eight rotors, four of which were used at a time. But by 1941, the U.S. Army's Sigaba cryptograph (called ECM by the U.S. Navy) used fifteen rotors, ten to create the electrical maze and five for the mechanical linkage. It was not until late into the war that the Germans introduced their ten rotor Geheimschreiber cryptograph. The second problem with Enigma was the software, particularly the initial keying procedure, which required a three letter group to be repeated twice, and it was this flaw that enabled Polish codebreakers to first enter the system. The third problem was that because some 10,000 Enigmas were in use, they became so commonplace that bored code clerks began taking shortcuts so that, for example, they would run their fingers diagonally down the keyboard producing repetitive three letter combinations like QAZ or WSX. All this was in addition to the fact that the rigid hierarchical command structure of Germany's armed forces produced endless stereotyped messages which were easily identifiable.
Although the history books have concentrated on showing how clever the British and American codebreakers were at penetrating the German and Japanese code systems, the truth is that the Royal Navy had just as insecure codes as did the Japanese and for several years into the war suffered badly from their refusal to accept modern cipher technology".
This directly quoted passage says "Typex Mks V, VI, and VII were all experimental models and never went into production". With respect to the Mk VI, this is incorrect. Rusbridger may have meant to say "Mk IV, V and VII", which are marks which Ralph Erskine states, "may have been experimental, or have had only limited production runs".
One other problem lies with the (omitted) direct quotation "The Royal Navy refused to even experiment with Typex, let alone use it in their ships, because it was beneath their dignity to use a machine designed by the RAF, which they considered the junior service, and they were also highly displeased that a junior officer like Mountbatten should have told them the Navy's cipher security was no good".
Erskine argues convincingly that the view that the Royal Navy rejected Typex is incorrect, and that the reason the Admiralty didn't put Typex on board ships was due to slow production. Certainly, the (omitted) statement that, "The Royal Navy refused to even experiment with Typex" is incorrect. Erskine cites Archive files to show the Admiralty had ordered:
* 630 machines by October 1939
* Further orders by July 1941
* Sent Typex to 50 ships and 36 shore authorities.
* In December 1942, the Navy wanted to receive 500 machines by 1943.
(Of course, the CCM came into use in 1943 instead).From 1943, the Americans and the British agreed upon a Combined Cipher Machine (CCM). The British Typex and American ECM Mark II could be adapted to become interoperable. While the British showed Typex to the Americans, the Americans never permitted the British to see the ECM, which was a more advanced design. Instead, attachments were built for both that allowed them to read messages created on the other.
Erskine (2002) estimates that around 12,000 Typex machines had been constructed by the end of World War II. Of these, about 8,200 were Mark II machines and perhaps 3,000 Mark VI Typex devices were made before August 1945.
Typex was used by the British armed forces and was also used in Commonwealth countries such as Canada and New Zealand. They continued to remain in service until the early 1970's.
From Wikipedia: "Another improvement the Typex had over the standard German Services Enigma were the use of of multiple notched rotors that would turn their neighbouring rotor.Some Typex rotors came in two parts — a slug containing the wiring was inserted into a metal casing. Different casings contained different numbers of notches around the rim, such as 5, 7 or 9 notches. Each slug could be inserted into a casing in two different ways by turning it over. In use, all the rotors of the machine would use casings with the same number of notches. Normally five slugs were chosen from a set of ten. On a Typex rotor, each electrical contact was doubled to improve reliability".
Typex rotors on display at the Communications and Electronics Museum, Kingston, Ontario. (Photo by Jerry Proc) |
This is a rather special set of test rotors that were used to check the functions of newly built Typex machines. (Photos by David Hamer, NCMF) |
Typex BID22 rotors. (Photo by Jerry Proc) |
Storage case for Typex rotors. (Photo by Richard Brisson) |
SOME TYPEX VARIANTS AND DESIGNATORS
Sometimes Typex is written as TypeX . The initial machines were called "Enigma type with X-attachment" . This was expressed as TypeX then later as Typex.Typex variants were assigned BID designations as : BID08/01 08/02 08/03 and 08/04
(Rockex was given a similar designations but higher in the series: BID08/05 08/06 08/07 08/08)
As higher versions of the Typex were produced, the BID08 designators were discontinued. Mark 22 was therefore designated BID 22 , Mk 23 became BID 23 etc.Typex 22 and Typex 23 were later variants that incorporated plugboards for improved security. The plugboards were located on either side of the keyboard. Both the Mk 22 and Mk 23 types were essentially Mk 6's with modifications.
Mark 23 was a Mark 22 modified for use with the CCM. In New Zealand, the Typex Mark 2 and Mark 3 were superseded by Mark 22 and Mark 23 on 1 January 1950.
Typex Mk II circa 1938. Based on the commercial Enigma design, the machine could achieve 300 operations per minute. Each Typex cost the government 107 pounds back then. Note that each print drum has 26 print solenoids, one for each letter of the alphabet. A switch on the front panel would also allow the operator to switch from letters to figures. Replacing a defective solenoid was a challenging task according to one Typex maintainer. (Photo by James Rusbridger)
Typex MkIII circa late 1930's. A message is typed with the left hand while operating the handle with the right. The encrypted text was printed on a paper tape. The machine had a maximum speed of 60 operations per minute. (Photo 10305551. Science and Society Picture Library)
Typex Mk 6 (Photo source unknown)
This machine (Mk 22 or 23) is on display at the Forsvarsmuseet (Trandum Annex) in Oslo, Norway. The Forsvarsmuseet is Norway's Armed Forces Museum whose main facility is in the Akershus Fortress in Oslo, an operational military base. The collection of cipher machines/radios/vehicles/etc. is outside of the city in Trandum, adjacent to the Oslo airport at Gardermoen about 40 or so km north. Admission to Trandum is by invitation only with an escort required due to its still-active military status. (Photos by David Hamer, NCMF)
This Mk 23 machine on display at Bletchley Park bears a nameplate with an Air Ministry BID/08/03 designation. (Bletchley Park photo via Wikipedia) MK III DESCRIPTION
Printing is by means of a wheel with the characters embossed on the periphery, continuously inked by a roller and against which the single strip of paper is pressed, hence there is no reference tape. The mechanism revolves once for each key press, the wheel being raised for upper case characters. Electrical power is produced by an impulse generator at the beginning of each revolution, stored in a capacitor and released when the appropriate character reaches the printing head to energize a small relay which trips the printing head to press the paper momentarily against the wheel. Because of this method of printing and the small amount of power available, all adjustments are very critical.PRINTING MODES
Cipher Mode : The left drum printed the plain language version of what was typed while the right drum printed the encrypted version By having a hardcopy of the plain text, it confirmed that input was correctly entered.
Decipher Mode: It was opposite to that of cipher mode. The left drum would print the cyphered text while the right drum would print the deciphered message in plain text.
In other words, the left drum always printed the keyboard input while the right drum printed the results of cypher or decipher modes.
Typex controls. (Bletchley Park photo via Wikipedia) OTHER OPERATIONS
This is an excerpt from Kruh and Deavours' Cryptologia article, "The Typex Cryptograph" (1983):
The space bar and the "X" key are connected so that word divisions are enciphered as X's. The Figures Shift is connected to the letter "Z" while the Letters Shift is also "V". Thus, the letters "Z" and "V" are entered via their shifted equivalents as visible on the keyboard while these two letters indicate the requires[sic] shift functions in the text.
In locations where commercial AC power mains were not available, Typex could be operated from a DC power by employing a rotary converter. Any rotary converter consists of an AC or DC motor driving a generator (DC output) or alternator (AC output).
The case dimensions are 14" x 14" x 11" and the converter weighs 27 kgs. |
The converter was made by E.D.C.C. |
This rotary converter unit No.2 (DC to AC) allowed the Typex to operate from 24 VDC power sources. Input : 24 VDC Output: 230 VAC; 50 Hz. 100VA. (Photos via E-bay) |
COMMENTS FROM TYPEX USERS Mike Simpson, a former Typex operator, relates some information. "At one time, our Washington office was located in a National Security Agency building, so messages were encrypted in Typex first, then typed up on 5-unit tape and re-encrypted in Rockex for transmission to the UK. Such messages had to be UKEYESONLY to justify such encryption. The rumour was that the Yanks could read the Rockex Tempest through the wall.
One problem we had with the Typex was the sticky tape the 5-letter code groups were typed on during encryption, or the plain text during decryption. If the weather got very cold, the glue became very brittle thus causing the paper to break".
********** An ex-Royal Navy sailor sent this Typex anecdote. "During WWII, I used the Typex many times and yes it had changeable rotors. Our instructions were to destroy the machine if it was in danger of falling into enemy hands. A heavy mallet was supplied with the machine. When Singapore fell to the Japanese all the rotors from the machine at the signal station were packed in a bag and taken to the evacuating ship. The machine itself was thrown from the top of a four story building (easier than using a hammer) but on examination all it had was a slight dent so it was ditched in the harbour".Bill Moffatt served in HMCS Iroquois in Korea in 1952. He recalls the Typex. "The British machine that we had on the Iroquois was a TYPEX. I recall that it had five the code wheels which I believe were much larger than those used in the American KL-7 machine. The Typex was a slow, clunky machine. I don't recall that we had anyone maintaining the Typex but I believe we had to wipe the code wheels with some kind of an oily cloth from time to time. The deciphered text was printed on a paper tape which was then stuck unto a message form and sent to the Message Center for typing and distribution. The classification levels that I recall were Unclassified, Restricted, Secret and Top Secret. They also had a priority classification. **********
As a 19 year old kid, just two years out of high school, I never thought much about security. We had been instructed in Comm School about the need to destroy the crypto machines and systems if we were about to be captured or sunk but there were no specific instruction or drills on the ship and I can't recall that security was ever mentioned. I don't believe that any background checks were ever made on operators and we were not warned not to discuss the machines or the cryptographic processes but we never did. The machines were on desks in a small compartment used by the ship's communicators but I don't recall any rules or signs barring access to the compartment by other crew members. When not in use, the code books and code wheels were kept locked in a safe in the compartment where the machines were kept. I don't recall that I needed to go to anyone else to get access to the safe. As I said, I didn't give it much thought at the time, but I left the navy in 1955 and in later years it occurred to me that there had been a very relaxed attitude about security during my time in the navy. I felt that the navy took more security measures in safeguarding the rum than with the crypto machines and codes.
We were taught the basics of book codes in Comm School but I don't think I ever saw a crypto machine before I went on the Iroquois. I was instructed in their use on the Iroquois by the yeoman who was a Chief Petty Officer and by a LS Signalman who had been in the Royal Navy before coming to Canada. After I left the Iroquois, I never used a crypto machine again. On my return to Canada from Korea, I took a teletype course at Cornwallis where I obtained a typing speed of 50 WPM. Subsequently, I was posted to the Naval Radio Station at Albro Lake near Dartmouth where I worked on teletype for the Canadian forces network".
********** Fred Jones, a Royal Navy Petty Officer Radio Mechanic during WWII, offers some background information on the Typex and one interesting experience with it while serving in HMS Lookout. "The various Typex "Marks" were developed over the years to cope with the German efforts to crack it's cyphering. Several of the earlier machines were left behind at the evacuation of Dunkirk by the R.A.F. These were found by the Germans, but since they were obviously copies of the commercial Enigma, they didn't give them much attention.
The major difference, apart from giving a printout of both plain-text and cipher text, was the requirement for an A.C. mains supply. It was also very heavy. It used to take two strong Marines to lift the beast off it's table and carry it to a secure area when we entered harbour. Typex was, of course, on the SECRET list, so it was mandatory to operate it behind a curtain in the Wireless Office. All messages decrypted on the machine were sealed by the operator and delivered by hand to the Captain. No one on board, except those authorized in writing from the Captain, were allowed to even see the machine. When in operation with the cover off, the Wireless Office remained locked.
I do remember an occasion where I was working on the Typex and there was a rattling and banging on the Office door. I stood up and asked who it was. " The Engineer Officer", was the reply. "Open this door, immediately!" I went back to the coding table, locked up the Typex, drew the curtain and then unlocked the door. The the Engineer Lieutenant stood in the doorway fuming with anger. "How dare you lock this door?" he snarled, trying to look past me into the Office. "What are you up to, out of my way " as he tried to push past me. I stood in his way and respectfully pointed out to him the notice on the door. "Have you written authority from the Captain to enter this compartment?" I asked. "Get out of my way " he said and tried to push past me into the Wireless Office. In response to this action, I slammed the door shut in his face, locked it and went back to my work. More thunderous knocking and imprecations from outside the door, which I ignored... then silence.
After a few minutes the office phone rang. "This is the Captain speaking. Why have you behaved disrespectfully to the Engineer Officer? He wants to charge you with a court-martial offence". I pointed out to the Captain that as the Engineer Officer did not have his written authority to enter the Wireless Office, I was merely carrying out the King's Regulations (K.R.'s) and Admiralty Instructions (A.I's) that were clearly written and affixed to the door. Also, I mentioned that I was working on the Typex machine at the time which the Engineer Officer was not authorized to see, so it was the Engineer Officer who was actually in default and culpable of a court-martial offence under the Official Secrets Act. A few more rumblings from above, then it all died down. Though that Engineer would never speak to me again, he was drafted to a shore base a few weeks later".
Another technician who serviced Typex machines while he was a technician with the Royal Air Force in the 1950's recalls the following. "They were normally very good and most problems were with the plugboards. Often they had been oiled, but one NEVER oils a Typex machine - it is lubricated with Vaseline only. When being sent to an RAF station we always took a spare Typex in our standard kit because it always took far too long to find out what was wrong with the faulty machine on station. Normally we gave them our spare and took the defective unit back to the service depot. *********
After finding many faults with the double plugboard wiring , I would test the machine by setting all the rotors to AAAAA and encypher mode, then type the plain language message so that it came out on the left hand drum in plain language, it should then come out in cypher on the right hand drum. Then I would reset all rotors to AAAAA and then switch to decypher, and then type the cypher message, which would then come out on the left hand drum in cypher. If I had repaired it properly it should come out in plain language on the right hand drum. If it didn't then I had to strip the machine down again - very boring".
**********
Ray Fortin, a former technician with Foreign Affairs Canada, also confirms that Typex plugboards were a maintainer's constant headache. " I recall that the plugboards were the biggest problems. By 1970, those wires on the plugboards had many years of wear and tear since the wires were changed on a daily basis and if the communicator doing the changeover would pull the wires by the middle instead of grabbing the connector body to pull them from the socket, the old and brittle wire would break or become intermittent. Vaseline was used on the rotor contacts.
David Smith , a former communicator with Foreign Affairs Canada says "If one was smart, they would clean the rotor contacts (both sides I believe), and then apply a small film of vaseline to all the contacts. This ensured the rotors would spin properly and seat well as well as making a good electrical contact. While in the service of Canada's Foreign Communicators, the machine acquired the sobriquet Iron Chinaman". **********
**********
From John Roy, former communicator with Foreign Affairs Canada."In 1962 I was posted to the UN force in the Congo and was informed that I, along with a Sargent would be escorting new Typex rotor inserts. I remember that on my crypto course, the instructor said how fragile they were. Eventually Ron Holding (another External Affairs ex communicator) and I were tasked to destroy some of the inserts. Believe me - the part about the rotors being fragile was simply not true. We tried sledge hammers and a piece of pipe with the sharpened ends and hammer and chisel, all to no avail. At this rate the descruction of the inserts was going to turn into a retirement project. In the end, we used a blow torch to melt each rotor into an recognizable blob of bakelite and bits of wire. Often wonder what the Congolese garbage man would have constructed out of the pile rubbish that we threw out?".
And how's this for improvisation? - John Roy describes a quick fix for Typex. "When we ran out of those springs used on the Typex tape reels, we used the Army issue condoms instead of the springs. A pair of scissors and a condom were kept close by. One a time, a piece of the condom was cut off and slipped over over the spring holders. The machine was placed in operation until the condom broke. The process was repeated until we could obtain proper springs. Army condoms were only slightly thinner than car inner tubes".
Additional Typex photos for this site Typex Inverter Power Supply - Royal Sigs Page Typex Photos - Royal Sigs Page Typex Simulator by Martin Gillow ADDITIONAL READING
1) The use of Typex in New Zealand, please select this link. Refer to page 8.
2) Restoring the Typex Machines at Bletchley Park.
Contributors and Credits:1) Bill Moffatt. E-mail: bmoff(at)iname.com
2) David Smith <drdee(at)sympatico.ca>
3) Matt Russell <matt_crypto(at)yahoo.co.uk>
4) Wikipedia http://en.wikipedia.org/wiki/Typex#References
http://commons.wikimedia.org/wiki/Typex
5) Fred Jones <g2ivsenoj(at)tiscali.co.uk>
6) Ray White <r.p.white(at)sympatico.ca>
7) Appendix 8 of the book “Betrayal at Pearl Harbor – How Churchill Lured Roosevelt into World War II” by James
Rusbridger and Captain Eric Nave. ISBN 0-671-70805-8 © 1991 Published by Summit Books, a subsidiary
of Simon & Schuster Inc.
8) Science and Society Picture Library - UK
http://www.scienceandsociety.co.uk/results.asp?image=10305551&wwwflag=2&imagepos=6
9) David Hamer, NCMF <dhhamer(at)comcast.net>
10) Raymond Fortin <raymondfortin(at)rogers.com>
11) John Roy <john.roy3(at)sympatico.ca>
12) David White <davidwhite400(at)hotmail.com>
13) John Kaesehagen <yendor3(at)bigpond.com>
14) Martin Gillow <martin(at)virtualcolossus.co.uk>
Mar 18/20