> INTRODUCTION a fascination with imagination   > THEME PARK MAGIC inspirational rides and attractions   > A PHANTOM PLOT UNFOLDS a 2003 test from Disneyland ideas   > THE HAUNTING BEGINS five years of Halloween, 2004-2008 > BACK FROM THE GRAVE 2010-2011 with new technology   > A 2013 RESURRECTION a mix of new and age-old effects   > DAWN OF THE UNDEAD 2014 show, part one   > THE ZOMBIES EMERGE 2014 show, part two   > A CHANGING CONCEPT 2015 show, part one   > BUILDING PNEUMATIC FIGURES 2015 show, part two   > ROLL UP, ROLL UP! 2015 show, part three   > INTO THE TUNNEL... 2016 show, part one   > MAKING MONSTERS MOVE 2016 show, part two   > TO HELL AND BACK 2016 show, part three   > SHARPENING THE SENSES 2017 show, part one   > A MAGNETIC ATTRACTION 2017 show, part two   > THE BIG EXECUTION 2017 show, part three   Animation and decapitation       This year, there was a lot to animate! All these props and figures that I’d come up with… I wanted them all to move! There was lots of variation in these required movements too; linear motion, rotary motion, dropping and winching… but how best to do each one? While Disney's latest series of Audio-Animatronics have been designated A-100, mine might be classified as A-0.1! But what my figures might have lacked in realism and life-like human movement, I hoped they would make up for in old-fashioned ghost train clunkiness and charm! I think even figures animated in a very basic way have the power to startle and shock when combined with the right sound and lighting. Both are useful tools that can emphasise and highlight, to help create a more impressive overall effect, as well as mask and disguise, in order to misdirect the observer's attention. The list below describes the scenes that I had come up with for the dungeon. All of them were to be sensor-activated, and incorporate movement.       SCENE A (IR triggered) - axeman, transformation into headless figure (one-way mirror effect with two pneumatic figures, sound and lighting cues) SCENE B (IR triggered) - hanging figure falling downwards (pneumatic figure with sound and lighting cues) SCENE C/D (IR triggered) – figure holding rope, guillotine with victim (pneumatic figure, motorised prop with moving victim figure, sound and lighting cues) SCENE E (IR triggered) - figure on torture rack, standing figure operating it (motorised figure with moving handle, static figure, sound cues) SCENE F (IR triggered) – figure on gallows, falling through trapdoors  (motorised figure on rope, static figure, pneumatic doors, smoke, sound and lighting cues) SCENE S (IR triggered) (in porch) – skeleton hanging on wall, jumps forward at guests  (pneumatic figure with sound cue) SCENE X (IR coin detection trigger) (at exit) – ratcatcher figure, moves when coin dropped into donation slot (motorised figure, sound cues)             Most of the standing figures were to be moved by pneumatic cylinders, much like in the previous year’s show; tilting towards visitors as they approached, then backwards again to a resting position. I liked using a pneumatic method for these as this could be used to create movement that was quite fast and fierce! For this reason I chose also to use a cylinder on each of the trapdoors in the gallows scene at the end of the walkthrough (scene F). This would allow them to quickly fall open, allowing the man in the noose above to fall through, and then close again once this figure was winched clear of them.           I planned to use the same reliable but relatively small airbrush compressor as in the previous two shows to drive these cylinders. There were to be seven in total, and this was probably a sensible limit, given that more than one effect was likely to be triggered at a time. However, there was a trick that I used to increase the number of animated objects without needing extra cylinders – the idea of ‘linked’ movement. In other words; don’t separately animate a figure, attach it to an already moving part! This trick was useful for moving the lever held by the 'executioner' figure on the gallows as the trapdoors opened (linking the lever to the nearest door with wire), moving the hands of the torturer (fixing them to the rotating handle of the torture rack) and also with the falling blade of the guillotine - it could push the head of its victim downwards too as it fell.                                         This worked by means of mounting the head on a weighted 'see-saw' type arm. Ordinarily the head was slightly less heavy than the weight, so it was held upright in position. When the blade came down and hit the arm, the combined weight was now heavier, so it tilted downwards. As soon as the blade was lifted clear again, the head was restored upwards. No motor or air cylinder required! It would have been great to have had a much larger compressor offering all the air in the world, but I needed something that wasn't going to rattle the neighbours' windows, so I had to stick with my trusty little airbrush model with its modest 3-litre air tank! As a result, it was all about finding a balance between making sure that enough things moved for as many visitors as possible (rather than them walking through and finding every scene inactive or resetting) and programming enough 'rest' time for each effect (only a few seconds) so that the compressor would not be heavily drained in the likely event of multiple people walking the circuit at the same time.         I used a mix of air valves this year. The previous air-leaking problem I had with those pesky 5-port, 4-way valves was solved this year by generous winds of PTFE tape around each fitting on their inlets! However, my emergency fix of using two single way valves per figure had actually produced perfectly acceptable movement, and this method also only required ‘half’ the air demand, since elastic was used to restore each piston to its resting position, and not another cylinder-sized quantity of air to push it back.       For this reason, I decided that all three standing figures that were to be animated, and the two trapdoors in the final scene, would use the same single-valve arrangement as the previous year. Three 4-way valves were also used, notably on the hanging figure in scene B, which I wanted to drop down suddenly into view and startle visitors!           For this movement, it was important that there was sufficient air pressure to lift the figure, mounted on a hinged arm, upwards and out of sight. But also I wanted it to fall downwards and crash into the mesh on the window in quite a violent way when triggered. With a 1-way valve simply ‘releasing’ the air pressure, the figure fell slowly and gently under its own weight – not very scary... So instead, I fitted a 4-way valve, to provide air pressure that moved the cylinder quickly in both directions; the figure no longer fell, it was forced downwards with some speed!                                           Rotary motion     There was another type of motion to be created too… I wanted the axeman at the entrance to repeatedly tap the blade of the big axe he was holding against his other hand, in eager anticipation of his victims! This movement would use another air cylinder, but this time, it would need to produce rotary motion; pivoting the axe up and down by about 45 degrees.       I hadn’t tried creating this movement with a figure before, so for a few hints, I got in contact with my number one pneumatics advisor…David Buckley! His suggestion was to use an arrangement called a bellcrank, to convert the linear ‘push-pull’ motion from the cylinder into the rotation needed to move the axe. The shaft of the axe was threaded onto a long metal rod. Fixed at the other end of the rod was a metal plate, about 5cm long, extending perpendicularly from it.         A 50mm air cylinder was then attached via a clevis fitting to the opposite end of this metal plate. As the cylinder piston pushed and pulled, it caused the plate to be moved around the fulcrum created by the rod; the rod rotated. The cylinder for this was driven by another 4-way valve, since I wanted the axe to be lifted and lowered in a controlled way. Again, this provided movement at the same speed in both directions.                                           Motors and magnets     Aside from pneumatics, I had another favourite way of animating things – electric motors! Whenever I’ve got a figure or prop that needs large amounts of rotation, or requires long periods of movement, the answer is usually to stick a motor in it instead! The man being ‘stung by wasps’ in the first scene of the 2016 show was moved using a 12V windscreen wiper motor, as he was particularly heavy. Much like the air cylinders, this motor was very strong and capable of moving heavy loads. I wanted to use motors like this to animate some of the other trickier and more complex pieces for this year’s show.         The thing that was proving the most complex was devising a way of animating the guillotine. It would have been easy to leave it as a static prop once I'd built it, but this simply wasn’t good enough in my book – I wanted it to move for real! The resting state was for the guillotine blade to be held in a raised position at the top of the device. Then, on cue, the blade would fall downwards and ‘chop’ a figure’s head off. Crucially, I wanted the blade to actually fall under gravity, rather than be winched down in a leisurely manner by a motor – an actual drop would provide downwards acceleration, for maximum impact! So the challenge was to develop a mechanism capable of lifting an 0.5kg approximate mass through a height of about 1.4 metres, hold it in position, then allow it to 'free-fall' on cue, before winching it back upwards to the reset position. This called for some sort of motor-powered assembly!     Eventually, I came up with a way to do it. It was based on the function of a clutch mechanism in a car, and worked on the idea of being able to engage and disengage power from a motor with a cable drum, onto which the rope holding the guillotine blade could be wound. How to make possible this connection and disconnection? The plan was to use my secret weapon for 2017...the electromagnet!                     < This is the assembly that I came up with. A 12V motor was connected to a long threaded rod. On this rod was a steel disc, fixed such that it was rotated by the motor. Next was a cable drum with a length of rope attached. This drum was able to spin freely on the rod. Mounted on the side of the drum was an electromagnet. This is the small silver cylinder in the photo to the left. When this was powered, the drum would be magnetised to the steel disc, and would be rotated when the motor was switched on. This would wind the rope onto the drum, and thus the weight attached to the rope would be winched upwards.             When the weight reached the required height, the motor would be switched off, but the electromagnet would remain powered. This would hold the weight in the raised position. When it was time to drop the weight, the magnet would be switched off. The drum would lose its grip on the steel disc, and the weight would fall downwards, unwinding the rope from the drum.                     Looking at the other ideas I had planned, I realised that two other scenes could be animated in the same way. < The figure on the gallows, held in a noose above trapdoors, could be dropped and winched using the same method.             The figure on the torture rack could also be ‘stretched’ using a similar mechanism, this time operating on a smaller scale – instead of displacing the moving part by about a metre, this one would move the figure by only 30cm or so, before releasing. This all sounded good, but it needed testing. So I got the bits and stuck it all together. And... it worked! The motor and magnet seemed quite capable of lifting loads over three times heavier than what was required! But the guillotine presented another problem; how would the electronic system that would eventually be controlling this prop know when the blade had been winched up high enough, and stop the motor?                     < I attached strong magnets at both edges of the guillotine blade, in the spaces covered by the wooden frame. In this frame, at the height that I wanted the blade to stop, and directly in line with these magnets, two reed switches were fitted. These are components that can open or close an electrical circuit when placed in a magnetic field. When the blade was winched upwards, its magnets would eventually pass over these reed switches. The controller for the guillotine would be waiting for this trigger, and would stop the motor once it was sensed. In the event of a fault and this not happening, the controller was programmed to switch the motor off automatically after six seconds of waiting, to prevent damage to the prop or motor assembly!                                           Magnetic nastiness, and how to avoid it     This was a system that relied on a motor, electromagnetism and gravity to operate. I hope Messrs Faraday and Newton would have approved…! Having incorporated some of the scientific principles of these two great physicists into my own contraption (not to mention the design principles of a certain Heath Robinson...) it seems only right to mention an important point about using electromagnets in an electronic circuit... For my early tests, I was simply plugging in and unplugging two 12V power supplies; one for the motor, one for the magnet. This was eventually going to be connected to a relay system and controller, so what about protecting them from the collapse of the big magnetic field created by the electromagnet?       I connected a metal oxide varistor (MOV) in parallel with the electromagnet. This is the circular, blue component in the photo to the right. This had a rated voltage of about 14V, chosen as it was slightly higher than the standard 12V operating voltage of the system. Below 12V, the MOV had a very high resistance. Thus while the power supply to the magnet was connected at the standard operating voltage (12V), it had no effect. It only worked its magic when the power supply was disconnected. At this point, the back EMF (electromotive force) created by the magnetic field collapsing was clamped by the MOV at its rated voltage; around 14V. This way, it dissipated the unwanted energy safely, without harm to the control circuits or relays. This happened in a fraction of a second; the magnet was able to completely lose its grip on the steel disc very quickly, allowing the unwinding of the drum to begin instantly.         Initially, I had struggled to come up with a way of getting power to the magnet, since it would be mounted on a freely rotating drum, and I didn’t want its wires to get tangled…a slip ring assembly, perhaps? I later realised during testing that the drum would only complete about five rotations before unwinding the opposite way again…the answer was to simply fix the wires along the nearest metre of rope such that that they would be wound and unwound from the drum with it! Getting these motor assemblies working reliably was a big step forward in the development of the moving elements of this year's show. Props in three separate animated scenes were to use these to provide their key 'drop' movements. Seeing that this idea had worked spurred me on to build the pneumatic figures and other set pieces to go with them.       The brains of the show       The system in charge of running the show was made up of seven Arduino Uno controllers, which ran the animation and effects for the five scenes inside the walkthrough, and the two moving figures outside. Each of these controllers recalled a CD-quality audio file, at random (for scenes with multiple tracks) from one of seven Adafruit FX audio boards. There were also about 25 relays, controlled by the Arduino boards to switch 12V circuits for air valves and motors, as well as 230V circuits for scene lighting. The triggering for each scene controller was done in the same way as in 2016; infra-red sensors fixed into the wall at the necessary points. When a person walked past, they interrupted the IR beam and the sequence was triggered.     There were five IR sensors inside the walkthrough, and a sixth in the porch to trigger the skeleton on the wall. The seventh figure was to be part of a charity collection feature, and relied on detecting coins being dropped through a donation slot (again, via infra-red sensing) to begin animating. Inside the house were three computers running VenueMagic DMX timelines; one for running looped flickering effects for the green PAR 16 lighting inside the tunnel, and another two for controlling DMX lighting elements in the first and third scenes. The sound system inside the walkthrough used four amplifiers and nine speakers, with another two speakers and two single-channel amplifiers for the two moving figures along the path outside.       The dungeon takes shape     Building the structure again was good fun - it brought back lots of memories of the 2016 show. But the job was still as huge as before, if not bigger! Props such as the guillotine and the torture rack were so big that they wouldn't have fitted through the finished walkthrough to be positioned, so they had to be transported and stored at the back of the garden before any of the structure was built!     The wall and floor pieces were reassembled to produce the same walkthrough structure as the previous year. Some of the scenes required the windows to be a different shape (mainly to obscure 'workings' that needed to be hidden!) so some of the wall pieces were re-cut, and some were replaced by new pieces. I wanted the inside to feel a bit different to the 2016 show, so I replaced all the red brick vinyl with a grey, stone effect, to match the dungeon theme. The walls themselves were also repainted with black emulsion to freshen them up. Nylon material was stapled across the tops of the walls to create a roof and to keep out unwanted light.     Once the structure was completed, I began building the scenes. The challenge this year was making sure that there was enough space behind each window to fit the props and figures! < There was a lot going on in the centre of the tunnel! Behind this figure with the rope was a curtained-off space containing the lighting dimmers and most of the air valves. On the other side of all that was the workings of scene A, itself covered over by a black drape on the side nearest the path so it couldn't be seen when exiting. > The gallows scene with the victim in the noose was another tight fit! Behind the black drapes, there was about a 1cm gap between the back of the trapdoor floor and the window of the house behind!     < Attached to the joists above the roof material, hidden out of sight, were the two motor boxes for moving the guillotine and the gallows victim. The whole walkthrough took about a week to complete, starting with laying the base cables on October 23rd. From there followed several long days of assembling and testing everything. The weather was very kind to me once again, with lots of sun and dry spells, which made it much easier to install. There were some great effects to be unleashed on visitors this year, and it was exciting for me to see the scenes fully operating for the first time. By the end of October 30th, everything inside the walkthrough worked correctly - phew!     On the morning of the big day, I added in the signs and the external lighting for the front of the walkthrough. When the darkness fell in the evening, it was, at last...time for the dungeon to open!