The cabinet was made as a wedding gift for my son and daughter in law. The cabinet itself follows conventional design and construction techniques - what makes the anniversary cabinet different is the combination locking mechanism. The four drawers require you to know a code (combination) in order to open them. Each drawer has a different combination. The idea then, is to give the combination for one drawer at a time (on the wedding anniversary). To get things going though the combination for the first drawer (the bottom one) is given "free" at the time the gift is presented so that the general operation can be explained and tested. In this case the guests at the wedding reception were given a sheet of paper and an envelope so that they could write a note to the bride and groom. The envelopes were sealed and placed at random in the top three drawers. So on the anniversary when the combination is revealed the contents (envelopes or other gifts) can be accessed. Of course this may not work as intended, as the level of bribery re getting the combination early could come into play.

The **video below** shows how the combination locking system works and the **Steps** following the video give details on the construction of the cabinet and the locking system.

## Step 1: Video Showing How the Combination Locking System Works on the Anniversary Cabinet

## Step 2: Some Stages of Cabinet Construction

Mortise and tenon joints, pocket holes, and dovetails are the primary joinery methods used in the construction of the cabinet and drawers. But any kind of joinery can be used to make a cabinet/drawer for the purposes of including a combination lock like the one described in this instructable.

## Step 3: Making the Lock Rod Plates

Each drawer has a "lock rod plate" that mounts with two screws to the drawer. The drill press and the table saw (with a sled) make this operation easy. I used scrap wood spacers to keep the spacing between openings consistent.

## Step 4: Making the Lock Rods

Mark the dowels where you need the flats machined after the lock rod plates are made and mounted on the drawers. The flats on the lock rods (1/2 inch dowels) are made with the router. An mdf block with a setscrew secures the rod for the first flat as the block runs in the router table groove. Flipping the block over and routing the second flat ensures a perfectly parallel set of flats. A little bit of sanding at the cuts completes the procedure. Seven lock rods were needed for this cabinet. The angular position of the flats for any given lock rod were selected arbitrarily - I just mixed it up a little by eye. Light springs and brass pins keep the rods in place and makes for smooth turning.

The combination of the lock rods and the lock rod plates on the drawers should, in theory, complete the locking system and present so many combinations that guessing the correct one (or two) would be pretty well impossible. However, I should mention that this is Generation 2 of this wooden combination locking system. The first cabinet had pretty well an identical system to this one in terms of the lock rods and the lock rod plates - **but** it didn't take long for son number one to figure out how to beat the system. By applying a light tugging pressure on the drawer handle as you slowly twist each lock rod, you can eventually "feel" as each of the rod flats line up with the lock rod plate openings. To overcome this deficiency in the current version of the cabinet, a hinged "lock rod cover" with a linkage to a pins that fit in a right angle groove on both sides of each drawer securely locks the drawer in place when the lock rod cover is open. With this setup you can't twist the rods and tug on the drawer to beat the system because when you can see the rods to twist them (cover Open) the drawers are locked in place by the pins. The video above or here: http://youtu.be/jXR1rI32I5A makes this easier to understand.

## Step 5: Lock Rod Cover and Linkage Details

I put together a couple of mock-ups as part of the design process to get the lock rod cover and the linkage working smoothly before i did the final construction. One mock-up is shown here. I went with plastic flat bars to give a friction free sliding action and to insure dimensional stability in the system over time. The brass pins that are press fitted into the plastic flat bars are kitchen cabinet shelf pins.

## Step 6: Some Other Details and a Question

The lock rod brass "pointers" are made from one-half of 1/2 inch brass grommets (the kind used for sails, tents, etc.). The points of the compass are the heads of brass brads. The final step to make the locking system work even more effectively is the lock rod shield plate that is shown being inserted in the drawer cavity. This shield helps prevent visual observation of the lock rod positions when one or more drawers are removed. **I'm hoping that some Instructable viewer can figure out how many lock rod combinations are possible, besides the correct one (or two), for any given drawer. (I don't know)**

First Prize in the

Woodworking Challenge

## 34 Discussions

3 years ago

WOW!!!

Truly Great Craftsmanship, Design & Purpose. MUCH RESPECT!

I'M GOING TO BUILD THIS WITH MY DAD AS A FATHER / SON PROJECT!

THANK YOU. BLESS

5 years ago on Step 6

There are 128 possible combinations for each drawer based on a little simple binary math. Because each rod has a position where it is correct (1), and its opposite (0), each rod's unlocked position can be said to represent a binary bit. The total number of working combinations can be worked out by determining what the largest number is that could be represented by the number of binary bits corresponding to the number of dowels. The range here is 0000000 to 1111111, with the possibilities in between being variants such as 1011010 or 1000001, and so-on, with each rod being aligned either as intended or in the reversed position.

*Since the locked positions of the dowel can be thought of as continuous representations between 0.0 and 1.0, we can ignore them for this calculation, but being able to divide them into discrete chunks will be useful later on.

1111111 is the largest number that can be represented in 7 bits, and that number in decimal is 127. If we include 0000000 (0), which represents all of the dowels aligned as intended, that brings the total up to 128 possible arrangements per drawer. It's not terrible security, actually, considering that an apartment door that has been set up to work on a master key system might accept as many as 35 other keys.

If we assume a fair level of precision and no angling of the cuts that might help to align dowels that are close but not on the money, we get 16 possible positions for each of the seven dials, of which eight are duplicates of their opposite position. That means 8^7 possible positions or 2,097,152 total positions if an attacker knows how the chest is made and is smart enough to only move the rod on an arbitrary 180 degree slice to prevent duplicating their efforts.

In a six drawer system, that means that as long as no two drawers share the exact same dowel position to open, there are 128 x 6 combinations that will open at least one drawer, half of which can be thrown out by an attacker, leaving 64 x 6, or (Remember that 1111111 will open a drawer just as easily as 0000000, so why try both?), or 384 of the 2,097,152 total possible combinations. Assuming this, that means that there are still 2,096,768 combinations that will not open any drawer. If I've carried all my ones and put my decimal points in the right places, that makes the odds of putting in a code at random that opens one of the six drawers 3:16,384 or roughly 1 in 5,461.

If the machining is extremely tight, you could increase the number of possible positions to 32 or even 64, though you might need to use steel or brass rods to achieve this level of precision, and at 64 positions per dial you would increase the number of possible positions to 4,398,046,511,104, or roughly 4.4 trillion total combinations while still only allowing 768 total combinations that would open a drawer at random, or only a single position per drawer if you only notch one side of the rod so that it doesn't operate in opposing positions. Friction locking the rods when the cover is in place would effectively prevent an attacker approximating the correct position and coaxing it in to place with brute force.

Anyone who does mathematics or cryptography for a living is invited to double-check my work, but I think I got it right. A motivated attacker could feel their way from right to left, or left to right, or outside to middle by turning the knobs a half step at a time in pairs until they got some movement, significantly reducing the number of guesses required to get in. This attack could be thwarted by putting each drawer on a guide track with gliders or bearing slides to reduce lateral movement, but might be overkill depending on the actual level of security desired.

One could simply smash the chest or drill out the dowels, but that's not really the point, is it? I hope I answered your question, and I think I got my numbers right. Again, if you're reading this and you're a cryptographer or professional mathematician, feel free to correct anything I've gotten wrong!

Reply 5 years ago on Introduction

Hi churusaa - Wow, thanks for the work on figuring this out! As you can see from others there is disagreement in the final calculated outcome and approach to solving the problem. Not being into this kind of determination (not a clue) I appreciate all input that results in some answer as to how many possible combinations are involved to open one of the drawers. If you have access to others with like abilities perhaps you might pass the page link on to them to see if your particular calculations can be verified. If you do get a chance to do this please let me know. Thanks again.

5 years ago on Introduction

Thanks for sharing certainly inspirational. Enter in a contest here.

Possible improvement a hutch or a fire box, (Lord forbid that someone would ever need it.)

Reply 5 years ago on Introduction

Thanks tseay - I entered an Instructibles contest some time ago and won first place. Got real nice drill and impact driver as a prize.

6 years ago on Introduction

This is ingenious. I love the design and the thought that went into it. The lock mechanism is a work of art. Did you design the lock yourself or is it based off of a model? Either way, the details you put into it are incredible.

Reply 6 years ago on Introduction

Haunted Spider - Yes I designed the locking system myself and this is the second generation of the idea. The first cabinet (a totally different cabinet design as it was a jewellry cabinet) worked fine but it was easier to "break the code" as it didn't have the secondary locking system (closing the lock rod cover to enable opening the drawers). Thanks for the comments!

Reply 5 years ago on Introduction

I was thinking this would be a good jewelry cabinet. It seems like this might be a historical idea and I love it. But if this is an original idea then filing a patent is fairly simple to do. Since you likely already have the drawings that define your design. Just stamp a patent pending on your documents and make copies and file. This is awesome.

6 years ago on Introduction

Love it, so many little details, very skilled and thoughtful

Reply 6 years ago on Introduction

gtramp - Thanks!

6 years ago on Introduction

Great ideas and nice design. If you have numbers instead of little dots on each node, it would be easier for you to remember the combination.

Reply 6 years ago on Introduction

Right now I refer to points of the compass when I am thinking or talking about the different settings. But numbers would definitely be easier.

6 years ago on Introduction

Assuming 8 positions for each "flat", there would be 8 possible combinations if you had only one locking bar.

If you had two, you'd have 8 * 8 combinations (8 positions for the second bar for each of the 8 positions for the first bar), or 64 combinations.

if you had three locking bars, you'd have 8 * 8 * 8 combinations (all of the combinations you had for the first two bars would be repeated for each possible spot on the third bar), or 512 combinations.

Continuing on, each bar adds a factor of 8 to the number of combinations, for a total of 8 * 8 * 8 * 8 * 8 * 8 * 8 = 2,097,152 possible combinations.

But, if you take into account that your "flats" don't have to be exactly on the eights, and really wherever you felt like putting them, then you have an infinite number of combinations.

Reply 6 years ago on Introduction

1048576, because the exact opposite position also opens it. Divide by 2 :)

Reply 6 years ago on Introduction

actually it would be 823,543

if there are 8 positions on each spindle (not using any positions in between pips)

2 of them will open the lock and 6 will not.

so if we combine the 2 opens into one permutation (since if it is in either position it will open) and take the 6 locked positions that will give us 7 per spindle

7 * 7 * 7 * 7 * 7 * 7 * 7 = 823,543

Reply 6 years ago on Introduction

You're treating the two separate opens into one. You could do the same with the ones that don't open, saying that they are just.

Let's do a simplified example of two knobs. Assume the first knob opens on 1. That means it would also open on 5, but wouldn't on 2, 3, 4, 6, 7 or 8.

Now the second knob opens on 2, and therefore also 6, but not on 1, 3, 4, 5, 7, or 8.

You have 8 * 8=64 possible combinations of numbers. But only the combinations 1,2; 1,6; 5,2; 5;6 will open it.

We still have 64 combinations. The fact that you have 4 solutions instead of just one, doesn't change the number of combinations.

Reply 6 years ago on Introduction

so that would give you 1,048,576 combinations and 128 possible solutions?

your gunna need more drawers Nlinventor :)

Reply 6 years ago on Introduction

Interesting math discussion - will wait in case there is a new number before I setup to make a zillion drawers :)

Reply 6 years ago on Introduction

Good catch. If the cutouts were offset, you would have the 2+ million, but with the offsets centered, you do have to divide by two, and you end up with your number.

Reply 6 years ago on Introduction

Like the 2 million plus but like infinity even better! Thanks!