DR CRYPTO "DR" <darwin_maring(at)hotmail.com> provides an opening submission for this device.
"The B-2 machine, known as Python, was an enormous square box, filled with relays was used with one-time tape. On this tape were a series of keys. Keys were composed of groups of random Baudot characters. At any given time, there were only two of these one-time tapes in existence, worldwide. The transmit tape and the receive tape had a number printed with an arrow every so many characters. That would delineate the beginning of a new key.
The transmitting operator would inform the receiving operator which key to use. As an example, the operator would type "Set 32" in plain text, then both ends would put the arrow associated with the number 32 on the set mark on the tape reader head. After a pause of several minutes the transmitting operator would start typing the plain text to be encrypted. If another circuit was available, it would be reserved for plain text order messages. In this scenario, after the tapes were mounted in the right spot, the receiving operator would send "32 Set" and communications would commence on the secure line. Besides using the keyboard, messages could also be provided in the form of punched paper tape thus resulting in faster transmission time.
The key tape was provided on a really large reel akin to a movie reel, and there was no limit on the length of the message. If the tape ran out, operators would simply SET an new tape. When plain text messages were provided on punched paper tape, it was mounted on a reader with a dual head. The plain text and key tape would be read simultaneously to produce the encrypted traffic.
To encrypt, modulus 2 was added without the carry to the character which was typed and it incremented each time a new letter was typed or a character read by a plain text tape reader. At the far end, modulus 2 was added without the carry and the result was plain text. It was very simple encryption and robust because there are no prime numbers or any other set of numbers to generate it. It was strictly random key and only 2 one-time tapes.
Because this machine was designed before the days of TEMPEST, no consideration was given to the suppression of electrical noise emitted by the relays in the B-2 and the taut heads (tape readers). Technology eventually progressed to the point where the B-2's electrical emissions could be read at some considerable distance thus hastening its service life".
George Mace recalls his experiences with the B2 Table. "The first and only B-2 Table I ever saw and worked on as a maintenance technician was in the Joint Services Communications Center on "FRED" Island, Eniwetok Atoll during the 1958 Nuclear Weapons Tests. It was in fact a large oak desk, similar to those used by executives during WW-II. The Teletype printers and transmitter/distributors (TDs) sit on the desktop. One TD for the clear text traffic and one TD for the Crypto One Time tape.
Operational controls were on a front panel of the desk, where normally the feet would go. Under the desk were numerous electro-mechanical relays, inter-connected by all BLACK colored wire! I helped make point-to-point ohmmeter checks of every wire in that beast. As far as I know, it was never used for traffic during the 9 months I was there".
Read George's account of his posting to Eniwetok Atoll.
Sid Carmean <sid_carmean(at)yahoo.com> also used the B-2. " I'll throw in my memories of the B-2 table from the standpoint of an operator in the USAF during my one-year assignment to Goose AB, Labrador in 1954-1955. We had three patchable full duplex circuits and we held tape systems with Thule, Narsarssuak and Sondrestrom ABs in Greenland as well as Pepperell and Harmon ABs in Newfoundland and Andrews AFB MD, Loring AFB ME and Offutt AFB NE. These latter two systems were for SAC operations as they rotated tankers and bombers on TDY up north to train in the Arctic environment, just in case. When they were operating, we were busier than usual.
I remember 'B-2' as being shorthand for the manufacturer's model number '131B2' and the Army/Navy nomenclature as being AN/FGQ-1 (Army) and CSP-2599(Navy). And, as often as not, we referred to it as the "TAUT Table".
Curiously, or maybe not so, this device was absolutely unclassified, as was the tape reader that read the cipher tape. Only the cipher tapes were classified, at the SECRET level".
Sigtot variations from manual TBSIG207. (Submitted by George Mace)
General manual for SIGTOT equipment . (Submitted by George Mace)
PP748/UM-1 power supply for use with "132B2" table which dates back to 1946. The B2 would take two teletype tapes, one clear text and the other random letters, exclusive OR (XOR) them together and send the resultant output as secure traffic. (E-bay photo)
Power supply nameplate (E-bay photo)
Manual TM 11-2209 is applicable to the 131B2 Table.
Manual TM 11-5810-201-15P applies to the TSEC/HW-10 which was the walnut coloured transmitter/distributor which read the One-Time-Tape for this system as well as other one time tape systems.
While on the topic of long paper tapes, Doug Eyre offers this glimpse into the 1960's world of punched paper tape.
"Known as nodes or hubs, various major sites around the world routed data via secure teletype circuits. These sites were known as P55(?) relay sites. Traffic was received and punched on Baudot paper tape. This produced huge amounts of chad which was disposed of in 55 gallon drums! One of the famous tricks of the day was to fill a fellow worker's tool box with chad.
Operators would take reels and reels of encrypted punched tape, read the holes to determine the start of header then string them over to other tape readers to be resent to outlying legs. Tension was kept on the tape through spring loaded wheels. It kept the tape taut so that it wouldn't hang up when it came out of the tape punch or when read by a tape reader. Hence it became known as a taut head or taut tape relay.
Operators used to wrap the tape on their hands in a figure eight between their thumb and little finger. Sometimes these could get pretty big and were referred to as "footballs". You put the beginning of the tape inside by your thumb and you wove it back and forth around your thumb and little finger, crossing over in a figure 8 in the middle. Then you could lay the football on the floor and pull the beginning of the tape from the center and feed it into a tape reader and it would generally flow flawlessly without getting tangled.
A relay site was known for bells and lights. When "FLASH" traffic was received, operators had tape strung across the room literally from the output of one tape punch to the tape reader relaying to another node. The operator would guide the tape by hand to ensure that the message was handled in super fast (for its day) time. An operator was expected to be able to process a FLASH message in and out ot the relay site in less than 5 minutes. Some operators colloquially known as tape apes could take up to 1/2 hour to relay a message.
All of these messages were manually logged and there was no doubt that it inflicted a great deal of stress on the operators. At Hickam, we had between 80 and 100 different lines that were secured by the KW-26. Many of those lines terminated in the P55 relay! Those men and women that worked that relay site earned their money many times over. In the chauvinist days, there were very few women in the crypto, teletype, or radio (career) fields. However, they were great typists so the majority of them worked in communications".
Taken in 1980, this is an example of a small Communications Center somewhere in Australia. On the right side of the room are Model 28KSR teletypes and readers. Tape perforators, which produced a large amount of noise, were fitted into sound-proof cabinets. At the left bottom is a Model 28KTR
Teletype. (Photo courtesy Syd Nagle)
|The CSP-2699 is a Teletype Transmitter-Distributor and was used with the Python cryptosystem. Click here for the data sheet. (Courtesy of NCM)|
Contributors and Credits:
1) Darwin Maring <darwin_maring(at)hotmail.com>
2) Don Robert House, N.S.E., Bell System Retired. Curator Emeritus NADCOMM. email: k9tty(at)dls.net
3) Syd Nagle. <syd_tracy_nagle(at)iprimus.com.au>