Introduction: K'Nex Coin Pusher
This machine has been built almost entirely from the construction toy K'Nex. The only non-K'Nex components are the signs. None of the K'Nex pieces has been modified in any way.
This Instructable cover various issues concerning the design rather than being a step-by-step guide on how to build it. If anyone is as wacky as me and wants to build something similar, further details can be given as required.
General Look of Machine
It was decided that the machine should be substantial and not a small toy, and that it should resemble a booth, rather like an arcade game. There would be no glass - children would be warned that a monster would appear if they attempted to cheat, and that if the monster was asleep they would be pummelled to a pulp instead.
The width of the machine was scientifically determined by seeing how long it was before tedium set in when constructing the beds. The final width turned out to be about 19" (49cm).
If the coins were thin there would be a chance of them getting wedged underneath the leading edge of the moving bed, so the British 1p coin was ruled out (they were 1.52mm thick (made of bronze) until 1992, when they became 1.65mm thick (made of copper-plated steel) - yes, the British 'coppers' are of two thicknesses).
The British 5p coin would be far too small to handle (it's only 18mm in diameter), and the obvious choice was the British 2p coin. Pre-1992 they were 1.85mm thick when new, and the modern ones are 2.03mm thick. That little bit of difference in the thickness over the 1p coin makes all the difference. This nice thick coin would be ideal.
Some thought was given to using the British pre-decimal 3d coin, but it was only just over 2cm in diameter and about 2½mm thick and the weight on the beds would have been much higher than for the 2p coin.
It was decided to have six slots in view of the width of the machine. Conveniently, three coin-slides could be joined together by red rods.
Some attempt needed to be made to reject invalid coins. The size of the slot restricts oversized coins, and there is a trap at the top of each slide which gobbles up most undersized coins (but there is no guarantee!). Have a look at the fourth image.
By this is meant the white section at the back of the machine which the coins bounce down. This provides some randomness. It consists of two arrays of white connectors, the connectors being joined with green rods. The two arrays are separated by white rods, with two blue spacers on each being used to maintain the distance between the two arrays. There is a white rod on one side of each of the white connectors, making them 53mm apart horizontally.
The white rods are staggered in alternate rows so that coins bounce down randomly.
The two beds which were to hold the coins would be supporting a lot of weight and so they had to be rigid and strong and they had to move with as little friction as possible.
It was decided to make the surfaces of the beds from yellow connectors only for aesthetic reasons, the connectors' length being towards the front and back of the machine so that the coins slid easily along them.
This produced the first problem: how would they all be joined together given that transverse rods which held the connectors together would have ends sticking out, thus limiting the potential width to that of a grey rod?
The solution was to use the underside of the bed to clamp together the yellow connectors on the top. The eighth image shows a partially-built sample. Some of the yellow connectors aren't supported underneath, but it doesn't matter because the coins are 1" (25mm) wide and so most of each coin is resting on adjacent (supported) connectors.
The underside of the bed had enough solid areas for it to move over wheels, thus minimising the friction.
This took a lot of thought. The top moving bed was going to be heavy even without any coins on it, and when fully loaded would weigh a few pounds (2 or 3Kg).
This meant that traditional methods for achieving reciprocal motion would not handle the potential weight - whatever method was used would have to be substantial.
Around two seconds was deemed appropriate for each direction's movement. The range needed to be just large enough to avoid too many overlapping coins.
The fifth image shows the reciprocating engine, but the video shows more detail.
One of the problems during experimentation was that at the beginning of the stroke there was a lot of leverage required to start the bed moving, making it jerk, and so in the final design green rods (which are quite short) are used to start the movement, then blue clips take over, and finally the corners of red connectors finish the stroke. Have a look at the video to see it in action (at 3m 30s). The final result is that the movement is a little jerky, but it's not really noticeable when playing the game.
Just one motor would have been powerful enough to provide the reciprocal motion, but a second was added just in case there was some strain. The motors work on the same rod so that there are no synchronisation problems. Interestingly, when construction the engine, the two sides were getting out of synch., even though the motors were working on the same (worm) gear. It turned out that an 84-tooth yellow gear was inadvertently being used (yes - they do exist) instead of the usual 82-toothed one!
Coins not being Pushed / Coins Piling Up
These were expected problems, and indeed they manifested themselves very quickly.
When the coins fell from the moving bed onto the lower one, they passed over the curved edge of the red connectors.. This made them fall about an inch or so onto the second bed and often landed on top of other coins, meaning that they were not able to push coins forwards towards the pay-out tray. The coins needed to fall straight down into the empty area of the bed, and if the edge of the top bed had been a sharp right-angle they probably would have done.
The second problem, also expected, was that when the coins fell they tended to pile up into a non-moving heap - the coins in a pile would all overlap, and when another coin fell onto the back of the heap, they would all be pushed so that the coins closed up a little, only for them the 'unclose' as the top bed moved backwards.
Both these problems had a simple solution: a gate was added to the leading edge of the moving bed. Coins then fell straight down, usually flat, thus pushing other coins towards the pay-out tray as intended. In subsequent trials, coins sometimes got stuck behind this gate, but they always cleared by themselves as other coins fell on them.
Real Machines Make Money!
Although the machine was intended for amusement only, one of the purposes of it was to demonstrate to children that money which is inserted into a slot machine generally ends up in its cash-box.
At the left and right edges of the lower bed there are two slots, each the length of two yellow connectors. This is enough to enable the machine to gobble up between 15% and 20% of the inserted coins. This is not apparent when playing the game, even though every now and then a cascade of coins can be heard which don't end up in the pay-out tray!
The slots are covered by a row of black rods to make them less obvious.
The cash-box is hidden behind the panel at the front of the machine under the static bed.
Weight of Machine
The machine had to be heavy to resist jogging.
The weight of the laden beds helped here, as did the construction of the base and sides which required many rods and connectors.
The base was constructed mainly from blue-rod-sided cubes with the yellow diagonal rods on opposite faces at right angles to each other - see Making Strong K'Nex Structures for the construction method.
Playing the Machine
Fortunately, the game is great fun to play!
Even though the cash-box is continually being visited by coins, there are enough wins to make the player believe that they are doing well.
The best time to insert a coin is when the moving bed is furthest back - there is a higher chance then of a coin falling on an empty section at the back of the bed rather than falling on top of other coins.
...are available - just ask!
Typos and Errors
Please let me know if you find any.