The
B-21 was the first cipher machine, invented by
Boris Hagelin, that was commercialized by A.B
Cryptograph. When in 1925 the Swedish General
Staff was searching for an alternative to the
German Enigma machine, Boris Hagelin told them
that he was experience in the design of cipher
machines and that he would produce, within six
months, a machine that was superior to the
Enigma. This was actually quite a display of
bluff poker, as Hagelin had nothing to start
from. They agreed and Boris managed to get the
machine finished in time. The only resemblance
with the enigma is the keyboard and the lamps.
All images courtesy John Alexander, Leicester.
Technical Description
The B-21 was the breakthrough design that started
the commercial success of Hagelin Cryptos.
Hagelin started with an existing design of Arvid
Damm, the 'simplified rotors' in a five by five
grid. These rotors were completely different in
design and functioning as the rotors of the
German Enigma machine. The Enigma had three
rotors, each with 26 contacts, one for each
letter of the alphabet, and a reflector to return
the signal back into the rotors. The Damm design
required only two rotors with five input contacts
and ten output contacts.
Beneath the keyboard there are 10 horizontal bars
[1] that can be divided into 2 groups of 5 bars.
If a key [2] on the keyboard is depressed, one
bar of the first group and one bar of the second
group is shifted horizontally. All 25 keys have a
different combination of 2 bars, one in each
group of five. The bars control 10 switches [3]
which are also divided into 2 groups of 5
contacts. On the first group of five contacts,
one side leads to the positive of the power
supply and the other side leads to the first
rotor [4]. On the second group of five contacts,
one side leads to the negative of the power
supply and the other side leads to the second
rotor [5].
Each rotor has 5 ring contacts and 10 output
contacts. The 5 input contacts are scrambled to
the odd output contacts in one manner and
scrambled to the even output contacts in another
manner. With each of the 10 steps of the rotor
there is switched between the odd and even output
contacts to pass the signal to the static output
rotor [6]. The output of both rotors leads to a
five by five matrix [7] with 25 lamps (one letter
is omitted or combined with another). The 10
rotor positions are labelled A to K (the letter J
is omitted)
The Pin-wheels
The design of these rotors wasn't new,
but it did a good job to scramble the electric
signals. However, the innovation in the design
was how these two rotors were stepped, and this
completely new design by Boris Hagelin would
determine the unique Hagelin machines for
decades. The B-21 is the first machine to use the
famous pin-wheels, also called key-wheels. The
B-21 has four of these pin-wheels with 17 (A to
R), 19 (A to T), 21 (A to V) and 23 (A to X)
pins, which all step each time a keyboard key is
depressed (the letter J is omitted at the
wheels). The combination of these primes,
together with the 10 possible positions of the
ciphering rotor, produces a maximum periodicity
of 1,560,090.
The two left side pin-wheels control the left
side rotor and the two on the right side the
right rotor. If a pin occurs on one pin-wheel, on
the other pin-wheel or on both wheels of such a
pair, the controlled rotor is advanced one step.
This produces a very irregular stepping of the
two ciphering rotors. The number of possible pin
combinations for all pin-wheel together is an
enormous 1024.
The key setting of the machine consists of two
parts. The first part is the setting of the pins
on the four pin-wheels (active or passive). This
setting was generally changed once every 24
hours. The second part is the initial position of
the pin-wheels and the ciphering rotors at the
beginning of the message, the so-called message
key. This message key had to be unique for every
message. To manually change the pin-wheel
position, the pin-wheels are disengaged from the
gear system with a half turn of an eccentric
handle..
When
the machine cover is removed, we can see in the
top left part the ciphering mechanism. The four
big pin-wheels are located at the bottom of that
compartment. Right above these pin-wheels is the
axle with cogs to transfer the position of the
pins into a movement of the ciphering rotors.
These are the two small rotors at the top, with
five slide contacts at its axle on one side and
five pin contacts on the other side. Although
they don’t look like ordinary rotors they
also have an input and output side and the wiring
is scramble.
The use of these pin
wheels would return in many later Hagelin machines with
huge commercial success for the C-36, M-209 and C(X)-52.
The prototype, build in 1925, was approved by the Swedish
Army and later on sold also to some other countries. The
commercial breakthrough however was an adapted version
for the French Army. This portable machine, designated
B-211, had a printing mechanism and could be operated
with a crank in case of power failure. The rotor
ciphering electric circuit was powered by a battery.
In the B-21 and B211, the
four pin-wheels are used only to control the stepping of
the ciphering rotors in a complex fashion. In later
Hagelin machines such as the C-36, C-38/M-209 and C-52,
the pins of these pin-wheels are used only to offset a
reciprocal alphabet print drum. The CX-52 uses the pins
for both offsetting the alphabet drum and to control the
very irregular pin-wheel movements, making finally full
use of the ingenious pin-wheel design.
