A linear rake gun.
I have brain damage and a very dodgy short term memory. I get intensely interested in a subject and wake up one day with no memory of it at all. The sad thing is that all of the passion for the subject goes too !. So this web site is a record of what I am thinking about and doing in my workshop, then if I have a memory wipe, I can read it and I will then know what I was doing. The danger of thinking out loud is that you can end up looking like an idiot, but that does not bother me as much as investigating every possibility. As I go along I reserve the right to change my mind.
As you may know I have a fixation at the moment that swinging arm pick guns are not the best solution for opening locks fast. I also think that all rakes are wrong as well, because I believe that the pitch of a rake should be exactly the same as the pitch of the lock pins. In short I think a pick gun should have a reciprocating action and that the only rake to use is the equivalent of a bump key !.
The mechanical resistance offered to a key entering the keyway varies from lock to lock and this means that the conditions required to open a lock will also vary from lock to lock. I suspect that for each lock there is a narrow band of conditions that will cause the lock to open. It means that we need to program the gun to operate in the center of all the different bands, that would encompass all locks. For example, different locks may need different rake velocities, or different stroke lengths. The simple way of controlling these parameters is by using a micro-processor, that can step the excitation over the operating range until the lock opens.
My next thought was that I do not have any idea about parameter spread on a real device and that it would have to be found experimentally. Part of the experiment apparatus would be the electronic controller, so let's have a look at that first. A measure of success of a give set of conditions, ie rake velocity is the lock opening time. If we think back to the bump key lock opening times can be measured in milli-seconds and so one function on our box of tricks is going to have to be an electronic means of measuring lock opening time, in other words, a stop watch function.
Then comes the question of the prime mover that will give a reciprocating action. Possible devices are a DC motor, a stepper motor or a solenoid. Lets have a look at the qualities of each.
Simple and cheap. It takes a finite time for the motor to get up to speed, during which time the lock pins are being excited, the pin pack gets out of phase and pin bounce is random. On the other hand because it takes a finite time to reach the maximun speed, during that time it is ramping through the RPM range. This also works with swing needle EPG's where the experienced user operates the gun in short bursts, to cover a wider range of frequencies. However, the pins very quickly get out of phase and bounce in a random manner. I therefore came to the conclusion that the only time the pin sets would 'fly' together was immediately after the first strike, which means unless the needle is touching ALL of the pins at that instant the pin pack will not clear the shear line together and the lock open. On top of that we have the problem that all pins have to be struck with the same force and that can never happen with a swinging needle, because it describes an arc and the back pins will always be struck harder than the front pins.
If we consider the velocity required for a strike, then stepper motors operate too slowly to open pin tumbler locks. The strike acceleration must exceed 1G.
A solenoid looks ideal from a first glance as it is simple and has the required velocity. We can change the stroke frequency over a range and also the stroke length. Varying the velocity of the stroke is a bit more difficult and would probably require a dash pot. On the other hand the more power we feed into the solenoid the faster it operates, so we could have a small amount of control by varying the width (PWM) of each pulse of current. So I think I will start off thinking about using a solenoid.
A solenoid operated reciprocating rake gun.
The key element in such a gun is the solenoid armature return spring. This has to be powerful enough to drag the mass of the armature forward with enough force for the rake to overcome the mechanical resistance of the lock. On the negative side the solenoid now has to be powerful enough to drag it's own mass back into the solenoid, overcome the lock resistance AND compress the return spring.
The first objection was that the armature and return spring now make up a fairly powerful hammer !. I wasn't sure that I would want to hit a Bogotá rake with that amount of energy !. The solution is to only use that power on the stroke out of the cylinder plug, which is the way most people rake anyway. The rake goes into the plug only under influence of the return spring. There is another advantage in doing it this way. If we assume that the gun is switched off when the rake is inserted into the plug, the rake is automatically at full extension before activating the gun, which reduces the danger of trying to smash the rake through the back of the lock. It would also make sense to put a depth stop device on the gun, similar to those used on electric drills.
We could of course simple bolt the rake onto the end of the solenoid armature BUT ..... the armature is actually touching the inside of the coil housing and that means friction. The return spring also produces friction and unwanted torsion ... and all of these add up to far more torsion than we require to apply to the lock plug. Just think how little is required when bumping. So somehow we need to mechanically decouple the solenoid armature and return spring from the rake, so that there is very little friction operating on the lock plug. This is important because torsion (or the lack of) is so important in getting a lock to open. On the experimental side any test needs to be made over a range of torsion loadings to find the ideal.
The electronics box of tricks.
A PIC microprocessor would be programmed to switch a power transistor to pulse the gun, from an internal PSU. The pulse frequency and pulse width would be independently variable. The device would contain a 16 x 2 line LCD display to show pulse frequency and pulse width percentage. In addition it will also measure and display the lock opening time, in milli - seconds. Yes I am optimistic aren't I hi!. The electronic box is actually very simple to make and program.
First we would establish the nature of the 'unlock bands' I mentioned earlier. Once established the micro - processor can be programmed to automatically step thought the centers of each band, until an unlock condition is achieved.
It would also be possible to program the micro-processor to sweep over a wide range and by measuring the unlock time, obtain a figure of merit for the relative operating efficiency, for each type of rake !.
If anyone wants to discuss this further, please email me at email@example.com
15th February 2010.
Thought I would do a preliminary check to see if the solenoid I have chosen is powerful enough to withdraw the bump key from the keyway, at speed, against the resistance of the lock and the return spring.
About life size.
OK, a similar set up as before, the bump key is bolted to the solenoid core. When energised the core smartly withdraws into the coil, stopped only by the spring when at full compression. Now I estimate that it will only need to move the distance equal to the distance between two peaks on the bump key. The distance the core is allowed to fly back out, under the influence of the return spring, is set by the knurled know on the left hand end of the solenoid. So far I have done nothing to decouple the key from the unwanted torsion applied by the mass of the core, core friction and the spring. In the above set up the fact that the key is actually bolted to the end of the core, means that any sideways pressure applied will cause the key to bind in the keyway, so I may have to think about mechanically decoupling the key before continuing with the key withdrawal power test.
It became obvious that I wasn't going to get very far without starting to bolt things down. So I have mounted the lock on a bracket and next have to make some sort of mount for the solenoid. A jury rig test showed that I will have enough power for key withdrawal. So what I need next is some sort of decoupling device between the key and the solenoid.
16th February 2010.
Found an excellent article by Tommy Tyler on the development of pick guns, very interesting, Did some thinking about decoupling the bump key from the solenoid, to remove unwanted torsion. Looks like a minimal bearing area problem. I did start to wonder about neodymium magnets !. In fact I got a little side tracked into wondering about rotating magnets to power a swinging needle type gun ... and also in a rake gun. As the magnet rotates like poles attract, unlike poles repel. I am not sure if it would solve the decoupling problem, so I need to do some experiments. If the rotating magnet is fixed then it would cause the free magnet to move laterally.
A variation on the theme would allow an electromagnet instead of the rotating magnet. As the current in the electromagnet changes polarity, so will the magnetic polarity of the electro magnets core, causing the left hand magnet to be attracted and repelled in a sinusoidal manner.
Also started to wonder about the resonant frequencies of the pin sets. The springs and driven pins are approximately the same size and weight, but the key pins are of different sizes and weight, which means each pin pair will have a different resonant frequency, which will vary from lock to lock, depending upon the cut of the key.
12th March 2010
13th March 2010
I can see a conflict coming up, so I want to separate any bump key/bump key gun element from this page and let it have it's own page. So I have opened a 'Spring bump gun' page. So what does it leave us with here, on the rake gun question ?. Well, I have suddenly realised that from an engineering point of view I do not know exactly how rake keys work, they just do not seem to make sense. If anyone out there can correct me, I will willingly include your comments here !. Whereas the bump key has a scientific explanation of the logic behind them, rake keys seem to have none .... other than any irregular shaped piece of metal inserted into a lock and moved around in an infinitely variable, random fashion, will eventually open the lock. That is the way jigglers work. By speeding up the raking action we would increase the chance of the lock opening at an earlier time, so if we fitted the rake key to a reciprocating pick gun, it would open the lock eventually, but at what cost of damage to the lock ?. Invariably when the lock opens, rake the key jams in the keyway and applying any further force can only cause damage to the lock. Look at video's on youtube of manual raking and you will see what I mean.
Another point with rake keys I have seen is that they can have a longer length of bitting than a normal key. With manual raking it does not matter because the user would automatically shorten the stroke so as not to strike the 'tail' at the back of the lock and the key never engages the lock fully. Making allowance for this on a motorised rake gun, may restrict the length of stroke to the point where the key cannot open the lock, because it cannot set the rear pins to shear.
29th March 2010
Got lost !. I now have a powerful DC motor with a pinion that matches the 57 tooth brass Meccano gear. I think it is the motor I removed from the air compressor .... pity I threw the other parts away !. That will give me a 3:1 speed step down and a x3 increase in power, that should sort out any lock resistance problems, but it will also mask any jamming, so I will have to work up slowly in the experiments. I feel sure now that I will not be able to use rake keys, but there is a possibility of torque loaded clutch, using the decoupling (or auto-stop ?). Have mounted the motor on an aluminum plate and the next job will be to mount the 57 tooth brass gear. I can then consider the con rod and the linear drive. The decoupling solution, I have already done and will retain that method.
Just thinking out loud, but I am considering getting a Klom Electric gun, but also agonising over whether to design another needle gun, with more power and a parallel motion, with a very small needle displacement. The body of the motor I am going to use on the 'rake' gun is 53 x 36 mm and a similar sized one would be interesting on a needle gun. Buying a Klom might nip this in the bud and allow me to move onto other things like the sweep frequency generator. Roll on Spring !. I suppose I could do all of the software on the Mk 1 development board and see what comes out of it.
How about an el-cheapo manual frequency controller for the Klom, Southord, Wendt, HPC etc guns ?.