This work is licensed under aCreative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
The material described on these pages is created, collected, edited and published
by Frode Weierud, © September 2023
| The Hagelin HR/HX Machines |
|---|
The idea of enlarging the Hagelin Cryptos to include other machine types came to Boris Hagelin around and after 1948. As he says in his 1961 memorandum "The Hagelin Cryptographers — a review and a program": "It should be realised that in 1948 when I began to plan for the reactivation of my ciphering machine business, the Hagelin machines then in use were based on a design which was years old. These machines had stood the brunt on many fronts during World War II, and the need for improvements as well as for new constructions became obvious." The project, started in 1948 with Dr Edgar Gretener to build teleprinter cipher machines, was already leading Boris Hagelin away from purely mechanical cipher machines towards more electromechanical devices. Nevertheless, the first models were still based on the Hagelin pin-wheel crypto converters known from his prewar machines.
The collaboration with Dr. Gretener lasted only one year; the two gentlemen seemed professionally and personally incompatible. Boris Hagelin was asked by the Swiss authorities to leave Switzerland, but he finally managed to obtain residency as a private person. He decided to continue the development of the teleprinter cipher machines after his ideas, which resulted in several models that lasted until about 1974. Boris Hagelin was still actively involved with developing new devices in this period, with his first employee and later chief engineer Oskar Stürzinger, who joined the new company Crypto AG in 1952.
The first Hagelin teleprinter cipher machine was the T-52, developed in 1951–52. Exactly when the idea for using permutation wheels, often called rotors, in these machines occurred is not entirely clear. The rotor concept was not new to Boris Hagelin because in his design from 1926, the B-21, he had used rotors. Because he was in direct competition with the German Enigma firm Chiffriermaschinen Aktiengesellschaft (ChiMaAG) for a contract for machines to the Swedish armed forces, he knew their commercial rotor machines well.
Hagelin's notes show that he started rotor pilot tests in 1952, which were done in Oskar Stüzinger's "hobby room". Small local workshops provided the parts. Hagelin called the rotors Durchgangsrad or D-Rad in German, which directly translates as a wheel with through connections. The idea to make these rotors easily reprogrammable seems to have occurred early in the development. Another problem Hagelin wanted to avoid or improve upon was the rotor contact problems that almost all the rotor machines of the time experienced to some degree. Even the relatively new US cipher machine, AFSAM 7, later called KL-7, brought into service in all the NATO countries in the early 1950s, had contact problems. All available information indicates that the re-entry or re-injection principle was part of the early rotor designs. This principle was conceived in the 1940s by the US cryptographer Albert Small while trying to discover how the Japanese had constructed their cipher machine Angooki Taipu B, codenamed Purple by the Americans. The idea was patented in 1944 but kept in secrecy for 17 years until it was released as US patent 2,984,700 on 16 May 1961. Therefore, in 1952, when Hagelin started his rotor developments, the re-injection principle was not public knowledge. However, on 16 January 1953, Boris Hagelin applied for a Swiss patent for the same principle under the title "Schaltungsvorrichtung für Verschlüsselungsgeräte — Switching device for cipher equipment." The patent was published as CH312,238 on 29 February 1956. Hagelin also applied for a US patent on 16 October 1953. Due to some error by the US patent office, the patent claim was not checked against their secret patents, and therefore, the unlike thing happened: Boris Hagelin was issued US patent 2,802,047 on 6 August 1957. The whole story is explained in great detail by William F. Friedman in a memorandum for the record on the subject "Hagelin Negotiations", initiated on 18 December 1957 and finalised on 10 January 1958. Here, we also learn that Boris Hagelin discovered the re-injection principle from a comment made by the German cryptographer Dr. Erich Hüttenhain during a visit by Hagelin sometime in 1952.
Unfortunately, the Hagelin Negotiation document has been heavily redacted by an overzealous declassification person who does not seem to have known that the re-injection principle was no longer secret and had been in the public domain for over 50 years. Some of the finer details about Hagelin's rediscovery of the re-injection principle and his use of the principle in his new rotor design are, therefore, still hidden from the historian 70 years after the event took place. From the unredacted part, it appears that Dr. Hüttenhain did not mention the re-injection principle explicitly or that he explained how it worked. Boris Hagelin left the meeting with Dr. Hüttenhain without understanding how the principle worked or could be implemented. When he was back at Crypto AG in Zug, he started turning things over in his head, making sketches and drawing without arriving at a solution. When later discussing the matter with Oskar Sturzinger, Mr. Sturzing immediately explained how it could be done. However, Hagelin explains to Mr. Friedman that the "trick" mentioned by Oskar Sturzinger he had already thought of by himself. Yet again, the redactions, which most likely refer to rotor re-injection, do not allow us to judge if Oskar Sturzinger or Boris Hagelin reinvented the principle. From the redacted document, one is left with the impression that Oskar Sturzinger's name also should have been on the patent application for this invention.
In 1953, Crypto AG started preparations for a prototype of a cipher teleprinter machine using rotors for the crypto converter. In contrast to the other Hagelin teleciphering machines, such as the T-52 and later T-55, which were stand-alone online machines connected between a teleprinter and the telegraph line or radio equipment, the new machine was a teleprinter with a built-in crypto converter. As such, it was similar to the Siemens & Halske Schlüsselfernschreibmaschine (SFM) T-52 used by the Germans during the Second World War. However, the year after, when the prototype of the new machine, TMX-53, was finished, Boris Hagelin decided not to launch a production. He realised he had misjudged the market and how quickly the two large European teleprinter manufacturers, Lorenz AG and Siemens, got back into the teleprinter business after the war. He realised he could not compete against these large competitors in the teleprinter market. Hagelin was also establishing closer relations with Siemens during this time, which a few years later resulted in an agreement between the two firms to stay out of each other's markets. Siemens would leave off-line cipher equipment to Crypto AG, while Crypto AG would stay out of the teleprinter market. Siemens would also sell Crypto AG equipment to their customers, such as the Hagelin machines for making random OTT (One-Time-Tape) keys on paper tape.
We know very little about the TMX-53 because only one prototype was developed. From the few available photos, we can see that it is equipped with five rather large rotors and that the movement of the rotors is controlled by the C-type pin-wheel and lug cage system. The rotors appear large, formed like a cylinder with a pin-wheel attached, unlike the disk-shaped rotors in similar machines. Boris Hagelin mentions that the rotor design was inspired by the B-21 rotors he designed in Stockholm in 1926. The prototype also revealed problems with the teleprinter sending and receiving contacts due to faulty construction. Due to the sparse information about this machine, we don't know whether the rotors used the re-injection principle. It likely did not, but further research is needed to verify this.
When the TMX-53 was abandoned in 1954, Crypto AG decided to start a project study for an off-line rotor machine, a study they named HX. In parallel, they had the TX study for online cipher machines to be used for enciphering teleprinter traffic. In the autumn, one prototype, the HX-54, was in development. The machine used five rotors with 26 contact points — 26-point rotors. Hagelin showed the prototype to William Friedman when he visited Crypto AG in September 1954, as explained in Friedman's Trip Report from 28 March 1955. In 1955, they had two HX machines in prototype development. The new machines, HX-55, were supposed to use ten rotors of a new 32-point design. The machines were designed to encipher the standard 26-letter alphabet, and the six extra wires in the rotors were being used for re-injection. This appears to have been the first machine to use rotors with re-injection. However, the prototype shown to Mr. Friedman during his visit to Crypto AG on 21–28 February 1955 was not equipped with ten rotors, only five like the previous prototype HX-54. On the other hand, it was fitted with the new 32-point rotors measuring 38 mm in diameter. Furthermore, it was equipped with a paper tape puncher and reader and a second reader for random tape (Tarnstreifen). How the OTT function was supposed to work is not mentioned, but it could have controlled the rotor movement.
The reason for making two and not only one prototype was because as soon as they were ready, they were going to be shipped to Paris for tests by the French Army, as explained in Hagelin's letter to Friedman dated 27 September 1955. The two prototypes would be ready by the end of October 1955.
Source: The three first Hagelin documents above are from Boris Hagelins Privatarkiv, Vol. 6:3, Krigsarkivet (The Military Archives of Sweden), Riksarkivet, Stockholm, Sweden.
Source: William F. Friedman Collection of Official Papers(), National Security Agency, Ft. Meade, USA.
Copyright:
This work is licensed under a
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
The material described on these pages is created, collected, edited and published
by Frode Weierud, © September 2023
|
Site visitors: |
