Mechanical Drawing Mess

Recently, I was tasked with installing some hardware which I had not installed before.  Well, not this model but to some degree a door closer is a door closer.  I am choosing to not name the culprit but if you recognize this diagram, so be it.  Basically, the instructions ticked me off and so they get some time on the net.

This is the main diagram to measure off the drill and tap points on the frame header and the door.  Being a diagram and not a template, it is not expected to be at 1:1 scale.  However, a quick look will find measurements which are not even in proportion.  For instance, compare the line for 4 1/8 (105) to that for 4 15/16 (125) to that for 11 3/8 (289).  For the best contrast, there are two places labelled 4 (102) and they are not even close to the same.

Another problem is that of reference.  A door closer swings a door and for this mechanical reason, the centre line of the hinge is the natural reference point. Every measurement should be from it. This is a counter example for the engineering drawing class I once took on positioning dimension lines. At site, it also drives you to either add up the numbers or serially measure. Either method introduces errors.  The five mounting holes never repeat spacing and they really could.

Am I done commenting on the instuctions from the box. Not even close.

• The fastener hardware supplied posed a few mysteries. There are clearly items to be used. Some are for applications I can tell are not needed here.  And then there are other screws. 
• The cover is held by 3 screws when 5 holes are available. The cover sits square if you put the top screws in and would be less subject to vandals taking the two lower bolts or them being lost from vibration. Yet, without the screws in the bottom, it looks unfinished.
• The arm is put on with a big bolt over a washer. Two with the right threading are provided and yet the diagram shows a third type and a very clearly different washer. 
• Like many closers, this has a closing force adjustment nut on the end of the barrel. The instructions say it is shipped at a middle setting of the power range. There is no table given for turns should you have a wider or heavier door. 
• You have to dismount the closer from the mounting plate to get access to one of the screws. This step is never said.
• They claim to provide hex keys to adjust two different items. Only one was there. However, if planned better, both could have been that one size.  The needed hex wrenches were side by side on my tool kit.
• The instructions given refer you to do the wiring as per another document which was not in the box. 

Some will read this and think the locksmith is grumbling again. If that were all it was, I should not be writing. Poor quality instructions slowed me down. You recheck measurements. You ponder over which bolt to use where.  You test out the mounting locations with the hardware as template before you drill.  One thing I learned is that I should be careful turning this item over to a less experienced coworker.

Another thought also came back to me.  Ease of installation affects product sales at times.  If two competing products step into the marketplace and one takes more time to install, that can quickly eat up the price difference. I am reminded of some of the stand-alone push button locks which had multiple wires running thru the door so were hard to install and hard to service. Human hands had to hold peices on both sides of the door at once as you found where the bolts connected them and you could never see then if you were pinching a wire. When one company found a way to do this without wires, it came in at a higher price and still sold well.

Some products will not sell if the right person does not name them to the client. If you knew how hard some piece of hardware was to install and how much reading between the lines you had to do, you skip it.  Very few products can not be substituted by other makers.

I am pushing a phrase from computers at this point. User interface. The instructions are the user interface for the product. Are the instructions as carefully planned as the product?


--- --- --- --- ---
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. And remember, keep your follower on the plug.

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

Diving Deep Into Pi

There are many creative ways to generate key, safe and alarm codes.  Some are trivial and generate repetitive or patterned codes.  I am going to cover a few of these to avoid followed by a few to get you stronger codes.

Birthdays & Other Dates
During setting safe codes, you learn from the client the code you just set the safe to was derivative of some birth date or other anniversary. In one case, the safe code was always pulled from one date or other.  The problem is this generates a very small set of safe codes.  A past employee who knows the last code was derivative of a date, can guess the new code is too. (If you tell me, then you probably tell some of the staff too.) Another problem is this code is drawn from a small set since it will over choose numbers under 12, under 30 and even the year is not truly random.

Phone Numbers
See everything I said above. I should add that if you were to derive a code from two different phone numbers of people who most of the staff do not know AND never tell anybody you were pulling the code from phone numbers, you could do this sometimes.  If you repeat it often, you will fall into patterns regardless.  A fact of life, we run out of friends at times.

Address Numbers
See everything I said above.

Time for some better methods.

Internet Random Number Generators
Many of these give pseudorandom numbers in that they repeat -- eventually. Some give truly random numbers by sampling real world noise.  Either way, to make this work for you, get the web page to print a few hundred at a time. When you need the alarm or safe code, you can pick six digits from the page and you could read down or at some angle too.

This link will give you 1000 numbers between one and 999 999 into ten columns. (It does not pad zeros to the front of shorter numbers so 15346 is 015346.  To a minor degree, you are decreasing the randomness to add the zero at the end.)

https://www.random.org/integers/?num=1000&min=1&max=999999&col=10&base=10&format=html&rnd=new

Let's assume you are the head cashier and you told even one staff this was your method to get safe codes. Or somebody may have snooped at the computer, the printer or the connection. You want the code you finally use to be well hidden.  Part of this is pulling 1000 codes from the server but you could also pull 5000 codes and print all of them too.  Any reload of the page will do no good as it will give a different set of codes.  However, the computer and printer may cache the pages you are using.  Again, print a page or two and then pick one code.

Diving Into Pi
Modern mathematics gives up several decimal numbers which never repeat nor terminate.  One such number is the number Pi from geometry as the ratio of a circle circumference to diameter. It is now calculated to billions of digit but you only need find a web pages with a few million digits and dive down.  Like the random numbers above, you can get a printer to spit out pages and pages of digits.  You print and then pick the six digits you need either by standard reading or backwards or vertically or some other sampling.

Again, if you are a had cashier, save these pages in a secure place.  Provided you did not circle the code you used nobody could find the code from them.  However, a week later you could since you picked the sample the first time.

To get you started, here is a link to get the first million digits of Pi.
http://newton.ex.ac.uk/research/qsystems/collabs/pi/
There are many other similar sites on the web.

There are also similar numbers to use in much the same way.
Square root of 2 or the roots of many other numbers.
Looking on the web, I found this great link page from NASA.  It gives several numbers to many decimal digits.
http://apod.nasa.gov/htmltest/rjn_dig.html

Something Approximating a Summary
One can use simple methods and faithfully do two things to get secured codes.  Mix up your methods and do not tell anybody how you derived a code.

The better system is to use a method which does not matter if they know.  Find a source for thousands of random codes and pick one. Next time around, you generate a few thousand more. I could learn your method and you still not 'guess' the code.

--- --- --- --- ---
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. And remember, keep your follower on the plug.

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

Master Key Exceptions

I have been building master key systems lately and mostly they have been simple tree structures.  There are times when you should vary from this. Some are statutory and some simply more practical. (I am not going to cite laws of my province or nation, since the net is bigger than that.  Get to know your local laws.)

A room or group of rooms should be outside the basic master key system if any of the following apply.

1. Cash storage or high value inventory.  If you ask staff to sign for the contents of a room or safe or till drawer, then you should isolate key access very strongly. If many staff travel about casually with keys, then cash or product shortages can not be traced.

2. Explosives. Lets keep these in the hands of only the licensed people.

3. Drugs. You may need to distinguish between OTC, prescription and narcotic type drugs. These are often treated differently in the laws which govern who has control for these.  Often the drug storage in a hostipal ward is under the control of only one pharmacist or a senior nurse.  Some drugs have street value and then during the day issues of armed robbery need be addressed and at night issues of forced entry need be addressed.

4. Forensic evidence. Police agencies collect objects which may have to later be used in court. However, accounting firms and others gather information which they may later pass to the police to generate charges.  For this to be useable in court, you need to show you have had continuous custody of the artifacts.  If not, the police can not lever your information into a search warrant and certainly not as evidence in court. Hence, private investigators need to have control of the keys to rooms holding records and artifacts.

5. Radioactive material. Workplace safety issues requires you keep untrained staff safely away from radiation.  Additionally, there are often federal regualation.

6. Personnel records. Employment records are best isolated from general access. Since you often have a janitor with a partial masterkey, this room will often get cleaned during the days. (Now that I get to it, many of the rooms above get cleaned during the day and often by the staff responsible and not the general janitorial staff.)

7. Accounting records in a server room. Your financial files are expected to be secured at the same level as paper records. The server room needs keying restricting it to only those who need access. (Regulations which derive from the Sarbanes–Oxley Act of 2002 in the USA may affect you and it is your responsibility to know it.)

8. Fire arms and munitions. Much like explosives and radioactive material, if you have these to store in a room, you know you have regulations governing their storage.

There are several ways to key rooms outside the master key system.  The worst is to simply pick a random key and pin the cylinder to that. Now you have to keep that as a record should you need to cut more keys. Often these rooms have very small sets of people with keys even having only ONE key in circulation which passes at shift change to the person on duty. If you pick a random key, you have to manually read through the pages of key records to assure there is no cross-over with any other room.

The better way is to plan for a small number of rooms to have master key exceptions and block them. All the keys are close together and the relationship to other key groups is clear.

These exceptions can be done in one of two ways.  The first is to have the key below the Top Master Key (TMK) with no intermediate master keys at all.  In a crisis, only the most senior staff can get access to the room. This might be how personnel records are kept. The other way is to set the lock as Single Key Different (SKD) with no master keys functioning at all. This is possibly better for drug and narcotic storage.

No list like this could be complete and the keys have to react to how the organization is structured. If you think I should add more categories, feel free to toss me a line.


And remember, keep your follower on the plug.

--- --- --- --- ---
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/

Errors in ITL 9700A Flat Width Settings

I am gradually setting up a shop within my company to do locksmith work and most of that is master keying new construction projects.  I have a new ITL 9700A and ran a key to test calibration early on.  The specified cut depths on an SC4 came out fine both at the bow and the tip as measured by calibers.  I ran a depth key and put pins into an OEM plug and everything was flat and felt nice.  So it seemed the machine was good to go.

This last few weeks, I was tasked with setting master keys for two projects each of about 100 cylinders of mixed types:  mostly KIL but some rim and mortise too.  The first was on Yale GB and was going well until I did the block of mortise.  The cylinders were in two slightly different lots as a pair had one cam and one part number while about 8 others had a different cam and so a different part number.  These final 8 were all dragging as you turned the keys.  Some more and some less, but all dragging.  I took it the millings were holding up the key rather than letting the key rest on the bottom of the keyway at the shell.  I redid those few with bottom pins 0.005 inch lower and all was fine. This project was shipped.

I was then given the next project of about the same size based on Yale GC keying and started with the KIL cylinders and they started dragging at times for some samples of the MK or CK from the start. I then started looking deeper and found I could shift the key in the ITL vise and widen the cuts to stop this.  Also, when I pulled out the plug and looked at the pins at the shear line, the deep pins were rising while the 0 pins were not because they did not have a steep to hit.  I looked about and found the spacing numbers for the Yale G series and the ITL numbers matched.  It is a very tough item to measure on a key, but best I could the centres of the cuts did match the factory specification. 

I continued to pin but knew I had not found the problem much less the solution.  Some worked fine but some were dragging to turn.  This was heavy pressure drag and yet did not leave me impression marks even when I went to the sunshine. [My shop has poor light for this. We are working on it as it is killing me checking the third driver in on each end of a Yale plug.]  This was enough drag, some clients would think this is not even the right key.

I then realized I had a copy of the Yale spec which I had used my Professional Locksmith Association of Alberta (PLAA) presentation in October. A quick look at this says the flat should be 0.054 but the ITL setting used its flat #4 which is 0.048. A small amount but looking at the keys which I had cut which were at the MACS limit told me I could move up the flats and not risk anything.  I then went into the manual and defined a new flat width for #9 as 0.056 and recut keys.  After that, all worked fine.

I redid the flat width by setting up flat #9 and using a correction factor to get it to apply to my Yale keys.  Doing this means it disappears every time the machine is reset. I have to reprogrammed the Yale setting to make this permanent. I am thinking of moving back to 0.054 at that point.  Also is flat #4 of 0.048 used for other keys system?  I will have to check the factory specifications for Sargent and Corbin and Schlage.

This tells me that the first batch were not riding up on the milling at all.  A few pins were riding up since they were hitting the steeps closest to the shoulder.  I hope dropping the pins by 0.005 will not be a problem if somebody widens the flats to the factory numbers.  I doubt it but 0.006 was the error by ITL in the flat.

If you know of other problems in the ITL manual, let me know. I have worked one for years and found them usually robust but doubt this is the only problem.

Pin Usage Frequency in LAB Smart Wedge LSW005

I have taken on the role of setting up a locksmithing operation within a larger company which does many other related building tasks.  As such, I am the local expert on the insides of the locks.  There are many within the hardware specification department who know about lock functions, fire rating and finishes. It is a once in a lifetime chance as I buy tools to do the tasks which are on the horizon without too many errors and sitting on unused stock and tools.

To that end, I have purchased a universal pin kit -- the LAB Smart Wedge LSW005. I like the size shape and stock of this kid and it has reasonably appropriate tables on the lid for most of my work.  However, when first purchased each cell contains 72 pins and I have started to run thin on some.  I was thinking of doing a pin order based on the locks I will most likely see and wondered what sizes will run dry first.  Being the geek  I am, I gave it to a spreadsheet and analysed for Yale GA, WR5, Sargent LA, and Schlage. (I will run into Corbin System 70 in time too but this gets me started and then you have to consider both plug sizes.)

So what sizes won?  Not so fast.
Not all the pins were even used by these four key systems. From 165 thou to 355 thou the 11 sizes were never used once.  Although in the real world, they will get used for other less common brands or to fit a badly cut key to only one lock.

One size was hit 3 times [drum roll] it was 240 thou [cymbal crash].  The following sizes are used by two of the above systems:  170, 180, 210, 220, 260, 270, 275, 295, 310, and 330 thou.

 For those as nerdy as I, here is the table.

And remember, keep your follower on the plug.

--- --- --- --- ---
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/