Silly Exercise in Master Keying Desks

I have been looking back at notes from jobs I have done years ago and this post derives from that job. It was an office with file cabinets using IN8 keying which in these parts is most all of them. Some technician had set this up before me. I just had to add one more cabinet to one of the changes. While not a robust solution to the problem it was kind of neat to see.

First the background on this keying system. They are five wafer locks with each wafer being one of three possible sizes, namely 1, 3, or 5. When you move to the keys, you can use the powers of 3 to get the number of key combinations. The MACS (Maximum Allowable Cut Separation) is 4 which allows a 1 and 5 to be adjacent. The only 'rule' which strikes out key codes is the 'flats'. You can not use the key 11111 or 33333 or 55555.

So how many keys are allowed? 3^5 - 3 = 243 - 3 = 240

In practice, you should exclude the pullout codes too. These are code which step down or stay flat on successive cuts from the shoulder. After a bit of wear on the key and the lock, the key may pull out in any position whether the lock is relocked or not. The list is a bit long, but it would include such key cuts as 11335 or 13355 or 11115. The flats above are also pullouts. If you write out the full list you should get away from the computer more! However, be that as it may there are 21 pullout keys which should not be cut.

This leaves 222 good practical keys.

Now to master key with just this simple system. (While I wrote notes on this trick when I saw it, I did not record the real key codes. So I get to invent some here.)

The Master Key is A with cuts 31135.
There are two simple change keys.
A1 is 35135
A2 is 31535

The cabinets locks have a wafer dropped out which I will show with an X.
Cabinet 1 has wafers 3X135. This allows both the master and A1 keys to work. The keys share 4 cuts and in the space which is different there is no wafer at all.
Cabinet 2 has wafers 31X35. This allows both the master and A2 keys to work. The keys share 4 different cuts than above and again in the space where they differ there is no wafer. If you insert key A1, it will not turn due to having a 5 cut in the second position when the lock has a number 1 wafer.

Cool huh? Not really. This decreases the security of the lock by decreasing the number of wafers in the lock. (This is a bit minor from a picking point of view as I can pick this group of lock in under a minute in most cases.) Also, of the 222 keys above, 3 of them will open one of these cabinets. One is the master key, the second is the change key in use and there exists a third key which will also open it up. This is trivial in this situation really.

How far can I push this?
There is a 6 wafer version of cabinet locks and you could drop out one wafer at a time. This could be charted as follows with the wafers given in brackets after each key.

Master A is 131353
A1 is 331353 (X31353)
A2 is 151353 (1X1353)
A3 is 135353 (13X353)
A4 is 131553 (131X53)
A5 is 131513 (1313X3)
A6 is 131351 (13135X)

Gosh, was that tough to type! Imagine writing code tables by hand all day!

Can we push this further? As I show it here, it looks like you can give unique keys to only as many locks as there are wafers in the lock or cuts on the key. The mathematics gets more messy, but you can drop out half the wafers at any time and cycle through all the possibilities and generate more keys.

This is pushed even further in one line of high security lock you may never have seen. It is factory controlled and used in places like prisons. By using a system very similar to this, they can set up a master key with over 1 000 000 keys under it. (I was just now trying to find the actual numbers on the web and could not. I may update this post if I find it.)


--- --- --- --- ---
The contents of this post are released for non-profit or educational use in whole or in part provided this statement and the attribution below are kept attached.

Laux Myth ... Thoughts From a Locksmith
By MartinB, Found @ http://lauxmyth.blogspot.com/