|
|
|
|
|
|
|
USSR Rotor Fialka
The internal wiring of the M-125 is extremely complex and is
difficult to deduce. Almost all wires have the same
colour and are bundled together with sticky tape, which doesn't make
the process of tracing the wires any easier. The wiring of the
later M-125-3 machines is even more complex, due to the addition
of the NumLock switch (30 ↔ 10). Fortunately, we were donated a
partly disassembled rotor mechanism, which eventually allowed us to trace
each individual wire [1].
On this page we will try to explain the working principle of the
two Fialka models, and show how they compare to the well-known
German Naval Enigma machine.
For this we will use the block diagrams from our manual.
Should you be interested to examine the Fialka in more detail,
the full circuit diagrams are available in the
Fialka Reference Manual, that can be downloaded below.
|
|
|
Block Diagram of the Enigma
|
|
|
First, consider the block diagram of an M4 Enigma.
At the heart of the Enigma is a set of rotating cipher wheels – the drum –
between an entry disc (right) and a reflector (left).
It has a keyboard for input and a lamp panel for its output.
The keyboard has 26 keys (one for each letter of the alphabet).
When pressing a key, an electric current is sent from the keyboard
through a series of coding devices. First the signal passes the plug
board which allows pairs of letters to be swapped.
From the plug board,
the signal is sent to the entry disc, which passes it on to the rightmost
wheel (4), which in turn passes it to wheel 3, etc. until it hits the
reflector at the left. Inside the reflector, each contact is connected
to one of the other contacts (i.e. 13 pairs of contacts) which effectively
'reflects' the current back into the set of wheels.
The current passes all 4 wheels (in reverse direction this time),
the entry disc and the plug board, until it arrives at the lamp panel
where one of the 26 lamps will light up.
The advantage of using a reflector is that it is a reciprocal
operation. In other words: the whole process is reversible or symmetrical.
To decode an Enigma message, one would use exactly the same
settings as for coding it.
The main disadvantage of this method however, is that a letter can
never be encoded into itself, as the return path is always different
from the entry path. This is a property of the reflector, and
turned out to be one of the
weaknesses of the Enigma
durng WWII.
The Fialka design is similar to Enigma, but has a number
of powerful advantages. First of all the plug board is
replaced by a card reader (German: Kommutator) which greatly
simplifies the setup procedure. In addition, the card reader
provides more possible permutations due to the fact that the
letters are not swapped in pairs. This means that each letter
can be mapped to any other one.
Next, Fialka has 10 cipher wheels (instead of just 3 or 4)
and an irregular stepping mechanism.
Adjacent wheels rotate in the opposite direction.
Like on the Enigma, a reflector is used to send the current back
in to the drum, until it finally arrives at the keyboard again.
But unlike Enigma, the result is printed directly to paper
instead of being shown on a lamp panel.
But that is not all. The Russians clearly learned from the
weaknesses in the Enigma design
that were discovered in WWII,
and developed a clever solution to the problem that avoids
a letter from being encoded into itself.
The next paragraphs will show how this was achieved.
|
|
|
Block Diagram of the M-125
|
|
|
In order to understand the circuit diagram, it's important
that you know how the various parts of the machine are
connected together. Consider the block diagram
of the 'simpler' M-125 machine:
We'll start at the keyboard (centre right), which has 30 keys
in order to accomodate (a subset of) the Cyrillic alphabet.
When pressing a key, an electric current is sent to the Card Reader,
which behaves exactly as the Enigma's plug board.
From the card reader, the signal goes to the entry disc,
which passes it on to the drum. At the end (on the left)
the signal is reflected back into the drum, through the entry
disc and the card reader, until it arrives at the keyboard again.
There it is fed to a diode matrix which converts it into a unique
5-bit pattern, similar to (but different from) the 5-bit Baudot
code used on teleprinter machines. The 5-bit data from this
encoder is used to drive both the printer and the paper tape puncher.
In addition to driving one of the 30 switches, the keyboard
also contains a mechanical 5-bit encoder, which produces a
digital code (identical to the one above) of the original
plain-text letter. This code is used when the Fialka is
configured as a 'standard' teleprinter device (i.e. in plain-text mode).
However, the plain-text 5-bit code can also be used to override
the 5-bit data of the enciphered letter. This is controlled by
a single contact from the reflector. When the current hits that
particular contact on the reflector, no signal is sent back
into the drum and the 5-bit code of the original letter is used
instead.
As a result, there is a chance of 1:30 that a letter
is enciphered into itself.
In addition to this, 3 other wires from the reflector are
fed into the so-called Magic Circuit, which uses a clever
rotational principle that make the Fialka partly lose its reciprocity.
This circuit will be discussed in more detail later.
The remaining 26 contacts of the reflector are interconnected
in pairs, just like on the Enigma.
|
|
|
Block Diagram of the M-125-3
|
|
|
The circuitry of the later M-125-3 machines
is much more complex.
Although the machine is backward compatible with the earlier model,
some powerful features were added, which greatly enhance
the possibilities and the security of the machine.
In its basic setting, the machine behaves exactly as the
old M-125 shown above.
However, the introduction of a reduction circuit,
adds the possibility of encoding messages consisting of numerical
data only (i.e. pre-coded messages). This is called numbers-only
mode. The reduction circuit is operated by a switch at the bottom
left of the machine, which is marked 30 ↔ 10.
When set to '30' (i.e. using the full set of 30 characters) it behaves
as before. When set to '10',
the keyboard is reduced mechanically and electrically to just the 10
coloured keys.
In this mode, only 10 of the contacts of the reflector are used
(4 magic wires and 3 fixed loops). The remaining 20 contacts
are fed back into the card reader, to the contacts of the 20 unused keys.
This way the signal is looped around until it finally produces a usable
numerical result. Precise details of how this is achieved can be found
in the Fialka Reference Manual.
This principle is known as re-entry, and is also used with the
American KL-7 rotor machine,
for which the US held a secret patent.
Both Fialka models have a paper-tape reader that is
integrated with the keyboard. The tape reader generates 5-bit data
from a paper-tape and feeds it to the mechanics of the keyboard,
which 'translates' it into a signal on
one of the 30 switches,
just as if it was typed manually.
The same mechanics are used to generate the plain-text 5-bit data
when typing on the keyboard.
|
- Tom Perera, Partly disassembled Fialka mechanism - THANKS !
Received June 2004.
|
|
|
|
Any links shown in red are currently unavailable.
If you like the information on this website, why not make a donation?
© Crypto Museum. Created: Monday 07 July 2014. Last changed: Thursday, 07 January 2021 - 11:37 CET.
|
 |
|
|
|
