Manufacturing tolerance

[granz]

First I'd like to say that I'm a complete newbie. I've read the MIT guide and some of the other guides here as well and I've had some hands on practice with my home lock and a purchase pick set.

The fundamental requirement for picking seems to be a lack of manufacturing tolerance in the pins. Setting a pin will allow a slight turn of the plug only if there is either a misalignment of the driver pins or some pin is wider in diameter than the others. All of the techniques seem to rely on one pin being the binding pin. So one at a time these slight misalignments are exploited to set driver pins and have them rest on top of the ever more rotated plug until they're all set.

But....what if the manufacturing tolerance is so precise that there is no single binding pin? What if seven pins are all the same diameter down to the fraction of a micrometre? And what if they're all in perfect alignment as well? They're effectively all binding at the same time and SPP is impossible.

So ...aren't there any "un-pickable" locks that are manufactured so precisely as to completely invalidate SPP methods? All of the guides make it seem as though SPP is guaranteed to work given enough skill. But this seems overly optimistic.

[Gordon Airporte]

It's actually harder than just that to manufacture a perfect lock. The chamber holes also have to be perfectly uniform in diameter and they must be drilled perfectly on axis of the plug. The plug has to fit as tightly as possible in the shell. These deficiencies combine, so a slightly small pin in a slightly large chamber makes the lock easier to pick.

As for imagining a mathematically perfect lock, I run into trouble with whether the plug diameter is allowed to be exactly the same as the diameter of the hole in shell that it fits into. Can your imaginary lock plug still turn?
If the plug is smaller then you can use that gap between the plug and the shell to twist the plug off axis, so the pin holes in the plug are on a different axis than the pin holes in the shell, and that allows you to pick the imaginary lock.

[criminalhate]

There would always be a tolerance variable. There may be one that would be extremely difficult to pick because the tolerance is so close but even a fraction of a mm would be enough to make a pin bind before the others plus the cost to create such a lock would probably run into the thousands or possibly even the millions to have a lock exactly the same.

[granz]

I hadn't thought about twisting the plug off axis. That seems like it might actually be a good technique to employ. I've read about axis misalignment due to manufacturing, but deliberately introducing an alignment change in the plug was never mentioned and it hadn't crossed my mind.

As far as cost goes, that's definitely a concern. Especially with my exaggerated "fraction of a micrometer" tolerance. One has to consider the average $30-60 household lock is probably never going to be manufactured to anywhere near that level.

I think I'm just having trouble with my lock, and it makes me feel better to imagine that it might be a very tough lock to pick. I've only been at it for two days. I think I might go out and buy a cheap lock to make myself feel better. The downside is if I can't solve that one either then I'll be really pissed.

[criminalhate]

buy a kwickset or a knock off of one, you'll be picking it in no time. Also the lock your picking may have security pins in it making it even more difficult.

[raimundo]

Some day you will find a 'Best' lock and see what very precise manufacturing is like, however, remember there must be some 'tolerance' space, or a tiny grain of sand could jam the cylinder or a pin like chock under a wheel. To work reliably, a lock needs a bit of looseness.

[elipriest]

In theory, if such a perfect lock existed SPP would be impossible. However it could still be vulnerable to a pick-gun or bumping (assuming we are only talking about inline pin-tumblers)

[andy77]

I'm with Raimundo. Once you get to the tighter precision locks, "Best" .. "Yale" ..etc, then the SPP process becomes much more difficult.

I thought I was doing pretty well until I started with these locks. Next thing I know I'm back to starting at 2 pins and reviewing past exercises.

Generally, I find that the tighter the lock tolerances, the lighter the tension I should use. But that might just be me. I'm far from expert.

Andy

[Legion303]

To re-visit one of the older comments, it is mathematically *impossible* to guarantee perfect physical tolerance. You can get arbitrarily close to perfect physical tolerance depending on how many billions you want to spend on facilities, but you'll never hit "perfect."

-steve

[straightpick]

You will never achieve perfect tolerance. For example, if you had .500" rod machined to +/- .000" and a .500" hole drilled +/- .000" you will not be able to put the rod into the hole. Tolerance is necessary. If locks were manufactures to extreme tolerances like +/- .0001", they would require so much maintenance and would wear out so quickly they would not be practical. Not to mention unaffordable!

[shadow11612]

Even if the cost was no object, a "zero" tolerance lock would not last very long. The tolerances are built in for wear and tear. Look at an old key and see how worn down it is and it still works. If the lock did not have any tolerance, an old key would not work.

[maxxx]

First you have to understand that locks are made in the millions. They are stamped out as fast as they can be made by machines. There is not some 60 year old German or Swiss Engineer hand fitting each part.

Locks are made by Robotic machines that work night and day. With each individual tiny part there is whats called a "tolerance". A plus or minus that will still function (hopefully) within the completed assembly.

The Engineers who design the parts have to take into consideration; production variances, lack of lubrication, thermal expansion (locks in Canada where its 20 below and in Texas where its 110 in the shade).

Then you also have production machine tolerances. The little cutting tools that shape the part also wear down. The stamps that make ten thousand parts per hour get dull and the parts get larger.

I challenge anyone reading my post to empty your pockets. There will not be a single thing there that that wasn't produced by the millions. Except the lint in the bottom but even that is very common.

So you are wondering why the tolerances arent tighter? Each part of a lock has a plus or minus measurement for acceptablility. One might be at the plus end, the next at the minus end. Welcome to the Industrial Age. It happened about 200 years ago.

[Legacy712]

It actually takes an incredibly minute amount of "slop" (tolerance) in a lock to be able to feel a single binding pin.

Statistically speaking, it is possible for a lock to be perfect, but -highly- unlikely (a quadrillion to one, probably). You may have two pins binding at the same time, but the odds that 4 or more pins will all bind at the same time is near impossible (unless you're putting a lot of force on the turning wrench maybe). And even if that happened, you could pick it in the opposite direction and get different results because of plug-to-shell tolerances.

Just my 2 cents...

[metalking00]

The machines that make the perfect locks would by necessity have to be perfect in every component as well. After any movement, there would be wear, the machines would no longer be perfect, and then neither would their product.

[datagram]

Granz,

You have to remember that "tolerance" as we talk about it is a combination of independent parts that form the lock as a whole. You'd need the pin chambers in both the plug and the cylinder the same size, you'd also need them aligned. On top of that you need to have the pins themselves also the same sizes, and so on.

Looking at all the parts necessary to prevent binding, you can see how hard it becomes. This is especially true because all pieces are manufactured seperately, and mixed and matched at random.

dg