My home made barrel key collection (& how to)

[Lauren]

I have Ilco's PDF. Their bit and barrel stuff is of poor quality.

[Lauren]

This Eagle padlock (shown below) was received without a key and in the locked position. Under close inspection through the keyhole, there appeared to be various levers to the left side. The lever edges did not seem to be flush with one another, indicating the hasp is locked by the levers directly.



This lock has a well defined patent date stamped on the back of the hasp. The hasp is comprised of two layers of laminated bronze plates, which is rather rare in antique lock design.



The locking levers are retained by a single lever in the rear of the lever stack. This unique design prevents picking until this lever is disengaged. A Peterson tension wrench was modified and then inserted in the back of the lock to displace the rear lever and to allow picking.





Ultimately, a thin shim was inserted between the hasp and the lock casing to retain each lever from the sway of the hasp as each lever was individually compressed with a needle pick. The next picture shows a view the 4 lever locking stack with a fifth obscured lever. The fifth lever is designated by a thin space to the right of the hasp hole.



A flat key was made from spring steel plate. The key blank was cut on a drill press using thin Dremel cut off discs. The most difficult element in preparing the blank was the precise cut for a "C" shaped ward plate that exist only on the left side of the key hole. This plate is directly underneath top half of the inside lock casing. The next two pictures show the some of the characteristics of this ward plate as viewed through the keyhole. This plate helps hold the key in precise alignment; an element needed to properly contact the rather thin levers.





The key bits were cut by reading the lever stack as the key was rotated counterclockwise. This padlock appears to open only when the key is rotated 90 degrees in this reverse direction. If the key is rotated clockwise, the key stops at 80 degrees. It would appear that the levers are each cut with two concentric surfaces. When the key is rotated just beyond 90 degrees in the counterclockwise direction the entire lever stack drops slightly about .100 inch, and the key stops when rotate back.



Here is a view of the finished product:

[MBI]

I always enjoy these updates.

[GWiens2001]

Lauren,

You are very gifted with these locks and making the keys. Thank you for sharing, and for explaining your methods.

Gordon

[dll932]

I have Ilco's PDF. Their bit and barrel stuff is of poor quality.

I have used them many years with few problems.

[PinPopper1]

This is incredible...I buy old keys on ebay and machine antique padlocks to fit them!!! We sholuld hang out!!! Just kidding, really really impressive. Great job!!

[RubberBanned]

Would you be willing to make, and sell custom keys?

[Lauren]

I get asked a few times a year about making custom keys. It can get pricey, since everything I make is machined from raw materials. You would really have to love the lock a lot. Some serious collectors don't want reproductions in their collections. I make keys for my own locks because it gives me great pleasure to see them work again, and for the simple fact that not many people can do what I do with the limited equipment that I have. Anyone willing to fork the change over, we can talk.

[Lauren]

I always find myself revisiting these old "Samson" Eight Lever padlocks by "Corbin". These locks have a great solid feel to them and are a pleasure to own. In the past, I discussed a method for shimming the lock open and producing a working key by reading the lever movement through the hasp hole. Although this technique was proven to work, the process is lengthy and totally unnecessary.

I began collecting a few of the original "Corbin" keys, and I noticed that each key was cut to 1 of 4 different depth cuts. I measured the total bit widths corresponding to each lever set on the key. I discovered that the cut will be .695 (no cut), .595, .495, or .395, measured in inches with a .005 +/- tolerance. In addition, at least one bit pair will always have a "no cut". Then I began looking at the series of possible combinations that a key could have, cut wise. I assigned a number to each total bit width.

1 = .400, 2 = .500, 3 = .600, 4 = .700

Since "Corbin" increased the possible key cuts by adding a ward plate in 1 of 3 locations between lever sets, I had to apply this information into the equation as well.

I discovered that the total possible combinations for the 1000 series containing 1 or more 4's in the number was 37. For example: 1114,1124, 1134,1141 .... 1442,1143,1444. The same was also true for the 2000 and 3000 series. If you are following the math, 3 x 37 = 111. But for the 4000 series, the total number of possibilities is 64. So, 111+64= 175. But since warding had to be considered in 1 of 3 places, 3 x 175 = 525 possible codes. I don't know if "Corbin" ever produced all 525 possibilities.

Now getting back to my mission statement, shimming the lock open in order to make a working key is not necessary. I recently obtained this "Samson" padlock without a key and in the locked position. All is needed to decode these locks is to have 4 keys. These keys could be machined or specially fashioned from original keys. If original keys are to be used, the top 3 layers of bit material must be removed, leaving just the bottom bits. This has to be done for a .700 bit, a .600 bit, a .500 bit, and a .400 bit key. Each lever set can be tested beginning with a .700 bit key and working toward smaller bit test keys. The proper code is obtained when the key turns 360 degrees in either direction using the largest bit test key.

My latest project lock was decoded using this method. The code was determined to be: 3412 beginning from the top lever pair to the tip of the key. This translated into the following total bit width: .600, .700, .400, .500. The ward plate is positioned between the first and second lever pairs. After doing some Ebay listing research, I discovered that my key is cut to a factory code of "P117".

I machined my key blank completely out of cold steel stock material. I did all the cuts to the bit area using a grinding disc and arbor set on my drill press. I rotated the key on a 11/64 drill bit, locked in a slide vise to maintain even cuts. I adjusted the arbor in the chuck of the drill press to cut at different bit heights. I highly recommend the "Dewalt" titanium split point drill bits. These bits drill more accurately with less walking.

[Lauren]

This latest homemade double bit barrel key was made without the use of small Dremel cut off discs in the production of the bit components. I used a carbide rotary file bit to form the bit and key head. This technique guaranteed that the bit and key head are of the same plane and thickness. A .250 inch bar was formed down the center of the work piece which was later drilled down the center to allow for the lock's center pin in the key hole. The square bar was then turned to form a shaft using a 11/64" drill bit as a turning axis against a rotating cut of disc and arbor in a drill press. The technique was proven to work well even with a one shot chance to drill accurately. Notice in the picture of the key how the shaft protrudes into the key head. And no, the key was not welded.

[Lauren]

Here is a sneak peak of a homemade steel barrel key that took 12 hours to make from raw material. The padlock that the key goes to was by far the most difficult lock that I have ever picked. And surprisingly, the lock has only has 3 levers. It took me well over a month of trial and error to come up with a process that picked the lock. Can anyone guess which lock it is and why the level of difficulty?

[GWiens2001]

I suspect an older Chubb. Didn't they use a middle lever to throw the bolt instead of the back lever? Maybe I am thinking of something else, but suspect that not having the rear lever to throw the bolt is the reason for the difficulty picking.

[Lauren]

No it's not a Chubb, but I wish I had re-visited one of my own threads to help avoid some of the difficulty in picking this lock. Here's another clue: "Leverage to the locking dog simulates lever and fence binding, but actually makes picking even more difficult".

[GWiens2001]

Have worked with a Yale safe lock that has a serrated post and levers, so if the bolt is tensioned, the serrations lock together. Does the lock in question work in that way? It would certainly explain the difficulty picking. Or are you referring to the Corbin that you wrote about that would lock the levers in place at the 90% point if the lock was not snapped shut?

Still trying to figure out the lock itself! This is a fun challenge, and educational for me.

Gordon

[Lauren]

Wow, Gordon! Very perceptive.

This lock is not a Cobin, but it is a Corbin clone. In fact, the lock was picked with hasp compressioning (while the hasp was in the locked state) in conjuction with constant rotational pressure of the rear cam behind the third lever....Final clue. Pictures to follow if you get it right.