Cypher Machines Maintenance And Restoration Spanning Sixty Years
From Cryptologia,  July 2003 by Clarkson, Dorothy

ABSTRACT: We describe the life and times of a servicing mechanic for cypher equipment for the Royal Air Force during World War II and an opportunity to serve again as a cypher machine restorer for Bletchley Park Museum.


I was born on 25th March 1922. My love affair with aeroplanes started around the age of seven. We moved to a house close to the perimeter of the old Croydon Airport and I soon found that by sitting on the top of a stepladder in the bathroom I could see over the fence and watch the aeroplanes. When Amy Johnson returned from her flight to Australia they put a rope all round the airfield, inside the fence and let the public in. She was driven round, sitting on the folded hood of a large open tourer, followed by 'Jason' (her aircraft) on a trailer - and I was there!

At about the same time I became fascinated by small mechanical devices such as clocks and musical boxes and was encouraged by my father. He had wanted to be an engineer but that was not acceptable at that time for the son of a professional person (his father was a solicitor) and he had to choose between the law and medicine. He took medicine and after completing the general part of the course opted for dentistry. He was a very good model maker, working in brass and silver, making small, delicate models of cars, ships, aeroplanes, etc. He was born left-handed but encouraged by an enlightened tutor to use both hands equally. He taught me in the same way and I cannot remember a time when I was not using tools and making models, mainly in wood. At an early age I was given a medium-size Meccano construction set which was added to year by year. During the twenties my father started building his own wireless sets and I became interested in this although understanding very little, if any, of the principles involved.In the thirties we visited the Hendon Air Displays a number of times and on one holiday visited Southampton to see the flying boats. I also saw the airship R101 as she set off on her ill-fated final voyage.


With such a background it was almost inevitable that I should join the Royal Air Force so in 1937, at the age of fifteen and a half, I became an Aircraft Apprentice, enrolled for a three-year course to become a Wireless Operator Mechanic. On the outbreak of the second war the course was shortened and the trade amended to Wireless Electrical Mechanic. In April 1940 I became a fully-fledged Airman, theoretically capable of servicing, maintaining and if necessary repairing any item of R. A. F. wireless and associated equipment, including vehicles and petrol and diesel driven generating sets. Towards the end of the main course we were given a short but intensive 'hands-on' course on the Typex cypher machine, with no manuals, no drawings and definitely no note-taking. I believe the machine used was a Mark I as it had solid drums, of which only three were driven and of course no plug boards.

By June 1944 I had attained the rank of sergeant, the trade had again been redesignated - as Wireless Fitter - and I was serving with a Mobile Signals Servicing Unit, responsible for the maintenance (and modification as required) of mobile radar units and their associated radio equipment, the unit being staffed by equal numbers of N. C. O. Wireless and Radar Fitters, equipped with vehicles fitted for one or the other function.

Shortly after 'D' Day the Unit landed on the beaches at Arromanche and set up in a small valley a mile or two inland. Following the break-out through Caen we progressed via Amiens to Brussels where we occupied a large Citroen garage and showroom on the outskirts, plus the next-door house for accommodation.

Sometime thereafter I was accompanying one of the radar fitters on a routine visit to a mobile unit when we were asked if either of us knew anything about cypher machines as theirs had developed a fault which they could not fix. I was able to effect a repair and this was automatically reported to Command (as the fault had been), the upshot of which was that thenceforth I was on call for any Typex servicing required, this taking priority over my normal duties. During this time I came in contact with both a Mark III and a Mark VI although not required to work on either.

Following the end of hostilities my position was regularised and I was designated 'Senior Servicing Mechanic, Electro-mechanical Cypher Equipment, B. A. F. O. Command' which sounds very grand and implies at least some staff but in fact I was not only the Senior Mechanic but also the Junior Mechanic and all those in between - in other words, the one and only! I was supplied with a vehicle equipped with a complete Mark II, a set of sub-assemblies, a selection of components, the necessary tools and test equipment and most important, a set of straight-wired test drums and given the task of visiting every Typex-equipped unit once a month to service, clean and check their machines. From then on eight out of every ten days were spent 'on the road', working, eating and sleeping in my vehicle, except for the period of the airlift when I was transported to and from Berlin in noisy, doorless (therefore very draughty) Dakotas (C47's) which on a previous flight may have been carrying flour or coal. For these journeys I had only a minimal toolkit, any replacement parts requested and the test drums, in their heavy steel case, chained to my wrist. As time passed and units were disbanded the workload decreased but I continued servicing Typex machines until my return to UK in October 1948, becoming so familiar with the Mark II that I could probably have serviced them in my sleep!

In the autumn of 1952, after a prolonged illness, I was invalided from the RAF and a year later joined a company manufacturing flight simulators and trainers, later to become Rediffusion Simulation Limited. Thirty-three happy years later I finally retired, having reached the exalted position of Senior Systems Design Engineer.


In 1999, in the TV programme 'Station X', I saw for the first time an 'Enigma' machine, which I had not previously even heard of and was astonished by the similarity of its encoding mechanism to that of the Typex that I had known so many years before. To see this in the flesh as it were, I paid a visit to Bletchley Park, saw both three and four rotor machines and discovered that they possessed three Typex machines, a Mark III portable, a Mark 22 and a Mark 23, the latter two being variants of the Mark II with the Mark 23 being fitted to accept the CCM Adaptor. I spoke to one of the guides, was told that none of these machines were operational, introduced myself and was then introduced to Margaret Sale, wife of the then Director Tony Sale. She then introduced me to Brian Oakley, one of the Committee members and we had a long conversation, followed by some correspondence and a meeting with Tony Sale was arranged, for a day on which the Mark 22 would be out of its display cabinet for filming purposes so that I was able to get a close look at it. I noticed that a small part of the right-hand shift mechanism had at some time been incorrectly assembled, locking the right-hand printer in the lower case position. This could have been deliberate but I see no reason to do it if the mechanism is operating correctly. Whether deliberate or accidental I think that spotting this convinced Tony that my claims were genuine. After a further meeting later in the year at which I met the Trust Director Christine Large and the Museum Manager John Gallehawk, I started work on March 9th 2000, working six hours on one day a week.

Of the three machines the Mark 22 was clearly in the best condition and having been kept in a display cabinet appeared to be reasonably clean but I quickly discovered that this cleanliness was only 'skin deep' so to speak and did not extend to areas out of view of the cameras, nevertheless I was able to clean both printer units, the keyboard, the DC power supply and the base unit as complete assemblies but the drive unit and drum box were a different story. On the drive unit the main shaft, sleeve and shift-cam drums were so clogged with dried lubricant that they could barely be rotated and the whole assembly had to be stripped. The biggest problem was with the drum drive pawls: these are crimped or spun to one end of light alloy bushes which rotate on a steel shaft but three of the five were immovable, necessitating complete dismantling and it took a lot of persuasion, assisted by various penetrating fluids, to remove the bushes from the shaft which then required burnishing.A few small parts were missing and I replaced these from the Mark 23 and from my own resources I manufactured a set of links for one plugboard. After lubrication and adjustment and some forty-three hours of work the machine was again operational.

For the Mark III I had the assistance of another Volunteer whose main interest was teleprinters but who, in the fifties, had had a course on the machine, had never worked on one but had a set of notes which proved invaluable, especially when it came to setting up. Earlier I referred to this machine as "portable" -I think a more accurate description would be "transportable" as it is quite large and heavy. For those who are not familiar with this machine the following may be helpful: it uses the same drums as the Mark II series but is hand-cranked although there is provision for the attachment of a motor drive. The speed is governed to a maximum 60 operations per minute against the 300 of the Mark II series. 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 energise 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.

Superficially the machine appeared to be in a similar condition to the 22, having been kept in the same cabinet, both having appeared in the TV programme but far more dismantling was required to enable the removal of dried-up lubrication. There were several wiring faults where tight lacing had precipitated fractures and careless assembly had crushed two wires, causing short-circuits. The counter required a new internal non-return spring, which I was able to manufacture but again the main problem was the drum drive, in this case all five pawls being immovable. Fortunately there were no parts missing and although the capacitor was below its nominal value it was still good enough to operate the relay satisfactorily. Partly because of my unfamiliarity with its mechanism but mainly because of the critical nature of the adjustments, this machine consumed some 140-150 man hours, spread over twelve weeks but was eventually fully operational.

Finally to the Mark 23 and that was a different story altogether: I was told that it had been used for instructional purposes and frequently dismantled and had obviously been left uncovered for long periods. It was very dirty, clogged with dried oil and grease and with a number of parts missing, in addition to those that I had removed. The drum feed mechanism had the same seizures as the Mark III and several of the printer solenoids had broken cheeks and connections. On one printer, the casting carrying the paper and ribbon feed mechanism was distorted, with two corners broken off, leaving two bearings floating in mid-air. It-is virtually impossible to straighten castings of this type (so-called 'white metal') as they are very brittle and I didn't even try, in fact I was surprised that this one was still basically in one piece after the knock that it must have received. The general appearance indicated that the assembly had been dropped while in the 'open' position. I managed to build up the casting, positioning the bearings so that the shafts were in the correct alignment and manufactured replacements for all the missing parts, with the exception of the paper reels and guards which were beyond my capacity as I did not have the necessary materials. A large number of screws were missing, of various sizes and lengths and all to the old B. A. (British Association) standard. I was told by every merchant that I visited that these were available to order but only in commercial quantities and none had any in stock. (In some places younger members of staff didn't even know what I was talking about!) Fortunately I had some of the larger sizes in my scrap box, left over from various projects carried out many years ago and finally found the smaller sizes in a local model shop. Eventually, after around 130 hours of work, much of it in my own workshop, the machine was serviceable but not usable because of the lack of the tape reels.

A number of specially modified Typex machines were used at Bletchley Park during the war for the routine decoding of German messages once the initial settings had been determined by the code breakers.

Following completion of the work on the Typex machines I was invited to participate in the restoration of one of the Museum's most valuable exhibits, the Lorenz SZ42 cypher machine, in collaboration with Dr. Craig Sawyers, an electronics engineer with a keen interest in and some knowledge of the machine.

We began work in the second week of October 2000 and initial inspection revealed the following:

1. The cryptographic unit, in its gasket-sealed box, was in excellent condition, the lubricating grease still soft and when the drive gear was operated manually appeared to be working perfectly.

2. The top layer of the main chassis appeared to be complete and in good condition but rather dirty.

3. Under the top layer all components and wiring were as new.

4. The power supplies, under the base casting, were a disaster area. The cover plate had obviously been missing for a long time and there was damage, both mechanical and as a result of exposure to damp resulting in the chassis being covered with a white powdery substance. Some of the wiring appeared suspect.

As the only information available was the article by Donald Davies (Cryptologia, Volume XIX Number 1, January 1995) it was decided that the first priority was a full survey of the wiring. This was carried out and between us we produced wiring data sheets (point-to-point connections) and circuit diagrams and found that in fact only a few connections on the power supplies required replacing, all other wiring being fully serviceable.

The main transformer, selenium rectifiers and capacitors were unserviceable but after much searching Craig discovered companies that were able and willing to manufacture exact replicas of the transformer and rectifiers and what's more, to do some of the work for free as a contribution to the work of the Museum. Craig himself restored the capacitors by fitting modern components inside the original cans.

Unfortunately, from the beginning of November I was incapacitated for twelve weeks following surgery on my left foot, leaving Craig to do all the hard work of stripping, cleaning and re-assembly so that by the time I returned at the beginning of February 2001 he was in the last stages of setting up.

We then located, in one of the Museum's store rooms, a dirty and dusty Lorenz teletype machine (a Springschreiber Type T32 Lo) which, despite its initial unpromising appearance, after a thorough clean came up looking almost new. Craig then built a small interconnecting unit which enabled the T32 to control the SZ42 and print cut whatever was typed in and the combination was successfully demonstrated to a group of visitors on 8th March 2001.

The following week I started work on the Siemens-Halske T52E cypher machine. I suspect that this machine had been stored in the same location as the Lorenz, the general condition being very similar - dirty, no base cover plate and a similar level of corrosion. The hinged cover for the relay unit was missing. Fortunately there was no sign of mechanical damage other than to the plug on the flying lead which was shattered.

As with the Lorenz the only information available was that found in Cryptologia. I cannot quote the reference as I no longer have a copy of the article. The machine has an integral D. C. power supply for the relays but requires additional external supplies and a connection unit but as I had no information on any of these was unable to do more than clean, re-lubricate and ensure that all mechanical parts moved freely and smoothly. Finally I produced wiring data sheets similar to those that I had written for the Lorenz. I don't really know how long I spent on this machine as all the wiring checks were done in my own workshop but it was eventually returned to its display cabinet on 12th July 2001.

That completed my work on Bletchley Park's cypher machines but I was then asked if I would like to 'have a go' at refurbishing a Hollerith Type 313/0 Card Sorter. Similar machines were used at the Park during the war for sorting and collating the mass of information collected but this machine is post-war, probably early to mid-fifties. I had never seen one before and there was no information available but it is an electro-mechanical machine and I agreed to see what I could do. I have since discovered that there is a manual in the archives but this was not known at the time.

I started by making sketches then piece by piece dismantled and cleaned the mechanical components, labelling each item and its associated fixings and marking their positions on the appropriate sketch. Next step was to clean and check the electrical components and wiring and produce wiring sheets and circuit diagrams. From all this I was able to deduce how the machine operated and after re-assembly and some adjustments persuaded it to run but it would not feed or sort cards. I was then put in touch with two ex-IBM/Hollerith service engineers who were persuaded to visit the Park, apply their expertise to the machine and impart some of their knowledge to me.

A small number of old cards were found elsewhere in the Museum and the machine was set up and ran successfully using these but unfortunately most were damaged or destroyed in the process. We were then presented with a box of new cards designed for naval use but the machine did not like these, trying to feed two at a time so we were more or less back to square one and had to repeat the whole setting up procedure. This done, the machine developed an electrical fault which I was left to sort out and was eventually able to demonstrate the machine to a number of volunteers and visitors on 19th September 2002.

I have since discovered that the new cards are approximately one thousandth of an inch thinner than the old ones which demonstrates the accuracy with which the machine has to be set up.

Dorothy Clarkson
ADDRESS: 2 Furlong Crescent, Bishopstone, Aylesbury Bucks HP17 8SG ENGLAND.
Written 1 October 2002.


I served in the Royal Air Force from 1937-1952, first as an apprentice and then as a Wireless Electrical Mechanic. From 1953-1986 the author served the company, Rediffusion Simulation Limited as a Senior System Design Engineer. The work for Bletchley Park came about quite fortuitously and has now been completed.

Copyright Cryptologia Jul 2003
Provided by ProQuest Information and Learning Company. All rights Reserved


Back To Typex

Dec 3/05