Making the Quantum Tunnelling Telescope

The idea for the telescope came from Stephen Bournes, the pier owner. He couldn’t understand why people bothered to use the ordinary telescope as there is usually so little to look at out to sea. He asked if I could make it more interesting. It seemed obvious to try and add ‘incidents’ to the view, but I couldn’t think practically how to do it. Any image or model would need to move with the telescope to make it appear to be part of the world outside. This meant it had to be some sort of virtual reality, but I didn’t want to simply end up with a video image as it just seemed too predictable.

I puzzled for ages not getting anywhere, and had just about given up on the idea, when I was showing someone round the pier in February. Describing my idea while looking at the existing telescope, I had a rare ‘eureka’ moment. The existing telescope had a platform round the base for kids to stand on. I suddenly realised I could fill the space under the platform with models, with a periscope to view them. Moving the telescope would also move the periscope so the models would look fixed in space.

		I enjoy playing with optics, so the project seemed like fun. Initial experiments showed that I could use an ordinary pair of binoculars focused on infinity (to give the standard view out to sea) combined with a periscope with weak lenses to adjust the focus down to 1.5 metres, the distance to the models. I then bought two expensive beam splitting mirrors, thinking I could superimpose my models on the real view. This was a complete failure. The view out to sea is usually so bright that any model looked pale and transparent, however bright I lit them. Anyway the superimposed images just didn’t look interesting – our eyes are so used to them. Looking into windows we very often see a mix of what’s inside and what’s reflected from outside.

I then had the idea of the model appearing in the centre of the binocular image, with the real view round the edge. This was surprisingly effective, though in the workshop, I couldn’t get an accurate idea of what it would look like surrounded by the bright seascape. To get the effect, I simply placed two black disks in front of the binoculars, blocking most of the direct view. Where the light was blocked by the disks, the beam splitting mirrors reflected it all down the periscope. Rotating the disks with a geared motor, the model appeared to come and go as the disks rotated. Flat on, they were invisible though the binoculars so the view looked completely normal. A subtlety of the optics that hadn’t occurred to me was getting the periscope image to work in stereo, through both eyes. In my initial experiments I saw two images which I couldn’t get my eyes to converge into one. I had to make a system for very accurately adjusting the mirrors so I could tilt them inwards a fraction. It took ages.

		With the prototype sort of working, I could start on the finished exterior. I pulled out all the useful shaped bits from my own pile of scrap, and went to the scrapyard, where I found some military jet engine exhaust pipes – beautifully made things with stainless steel bellows. The tricky thing was designing the exterior all to be completely waterproof – its so exposed on the end of the pier, the rain is often horizontal. At the same time it had to be easy to open up for maintenance. I had originally thought of building it all on a pallet truck, so the whole telescope could be swiftly moved indoors for maintenance, but the prototype showed that I needed all the height I could get for the models, and the pallet truck would have reduced it too much. Having roughly made the decisions about waterproofing and access, it only took a couple of days to weld together. Making the ring bearing for rotating the telescope took much longer. It had to be a ring so the periscope could pass down the middle. After a couple of failures, I just managed to turn a 10mm thick, 400mm diameter steel ring. Almost too big for the lathe, the ring jumped out at one point and wrecked my live centre. I’ve since found I could have bought an off the peg ring bearing for less than Ł100.

The next problem was to decide what models to make. In May I took my prototype down to Dorset to show Gavin Morris, who runs the Digital Funfair. His initial reaction was that my optical effects made him feel sick which knocked my confidence, but we then started to have some good ideas. I had got sidetracked by pretending the telescope was a cold war relic, through which you would be able to see the Kremlin etc. Talking it through we decided the scope was limited, and sticking to stuff that happens out to sea was more straightforward. Gavin’s greatest idea though was to add sound to every scene. I was nervous about the extra complexity, but could see it would make the experience more engrossing, so I decided to go for it, and commissioned Gavin to do the sound tracks.
Adding sound to each scene meant I now had to fit a ring of sensors (to trigger the right sound in the right position) in an already cramped and inaccessible area next to my ring bearing. I eventually found a place to mount a ring of reed switches. The good thing about reed switches is that the tolerances are quite forgiving (they trigger at up to 18mm from the magnet). The bad thing is that the switching positions are never very precise. It all just fitted in, though I hope they don’t cause problems as the whole machine now has to be stripped down to get at them, certainly not ideal.

At the end of May I panicked that I wasn’t going to be able to get it finished in time for the summer holidays. Most of June was already booked on other projects. I rang Patrick Bond to see if he could come for a couple of weeks in July to help me catch up. Patrick has worked for Cabaret Mechanical Theatre for years and has a lot of experience of making automata and also has lots of good ideas! By chance a student (Mimi Ling Yan Leung) had written, asking if she could do some work experience, sending me some blurry pics of model people and dogs she had been making. 10 year old Ben Bournes also joined in and made excellent fish. I had rough ideas for some of the models but most of the ideas came while we were actually making them. It was a great team, and in 2 weeks we had all 12 models finished and motorised. I sent videos of them off to Gavin to do the sounds while Graham and I spent the next two weeks wiring everything up.

Technically, the difficult thing was to get an MP3 player that would reliably play the right track in the right position. Ordinary MP3 players are hopeless. They automatically switch on at very low volume (to avoid damage to the ears), have tiny buttons which are tricky to solder to reliably, and don’t have individual buttons to play each track. ‘Kiosk’ or ‘industrial’ MP3 players are much more suitable. I rashly Pay-Paled the cash to a small US supplier, who then admitted he was out of stock. Meanwhile my friend Andy had been trying a Medeawiz DV68 video player for playing MP3 tracks and was pleased with its stability. Will Jackson very kindly made me a pic chip circuit to plug into the DV68s RS232 port so I could control it by sending a binary encoded track number from my Programmable Logic controller. To my amazement it worked pretty reliably. It can get in a muddle if asked to change tracks too frequently, but I’ve programmed the PLC to only send it play signals at least 1.2 seconds apart. Track zero is a default track which it reverts to when no other track is selected. Making this track several minutes long also improved its stability, as the player also gets in a muddle if it gets a play signal just as the default track is looping. Thorough earthing also seems to be very important.
As with all my simulators and arcade machines, I have to wait until they’re almost finished to get any feel of the total effect. Programming the PLC is a bit like editing a film, finding the best way to put all the ingredients together. Once the technical problems of the sound were roughly sorted, I could start to enjoy the process. I realised I could use the position sensors to switch off the halogen light (which illuminates the periscope models), when its not pointing at the centre of a model. This improves the effect enormously as you just never see the joins between the models. I played with different effects with the rotating shutters and worked out an instruction commentary to get people started. Gavin’s sounds added enormously to it all – well worth the effort.

Graham and I finally got the telescope on the pier on Monday 7^th august. Although we had no serious problems, it took most of the day, simply because its so difficult doing anything in such an exposed position. I still feel nervous about its weatherproofing. Seeing the effect of the models surrounded by the real sea for the first time was wonderful – much better than I had dared hope. Everyone seems to read the cold war sign, some obviously believing it, which is very satisfying. And people who do put Ł1 in do seem to enjoy it.

Simulated view through the telescope,
with model surrounded by the real view

THE QUANTUM TUNNELLING TELESCOPE

England Heritage is delighted to present the Quantum Tunnelling Telescope, first tested here on Southwold pier in 1968. It was developed by the Royal Radar Establishment in their secret East Anglian laboratory at Orfordness. An astonishing achievement for the time, it exploited the latest advances in particle physics and quantum dynamics. The project was led by Dr Tim Hunkin, ably assisted by Graham Norgate, Gavin Morris, Patrick Bond, professor Will Jackson, Ben Bourne and Mimi Ling Yan Leung, Without such an extraordinary team the telescope would never have been so successful.
In use at the government’s GCHQ, it completely changed the course of the cold war, monitoring activity in communist territories. Recently declassified, the telescope has been restored to full working order by Quinetiq with a grant from the European Union’s crossing boundaries fund.

	Of course there were the usual teething problems. Some bits came loose and fell off. People often didn’t get the idea of tipping the telescope – and often didn’t even rotate it – obviously thinking the next scene would appear automatically, so I had to revise the instructions a few times to make them clearer. They aren’t so stylish as they were, but its much more important that they are clear. Some people were deterred because they weren’t sure if it was working as it has no flashing lights. Fortunately, they usually peer through the telescope anyway, so I added a bit to the program that switches on the sound effects for 10 seconds whenever the telescope is moved. Trying to work on the end of the pier is still really difficult – I can hardly see the screen to my laptop in the bright light – I sometimes have to spend ages just looking for the mouse pointer on the dim screen. I really hope it turns out to be reliable. The day after I wrote this it poured with rain all night. The next morning when I opened the 'waterproof' control box, water poured out of it. Inside, the little box containing Will's RS232 circuit was completely full of water. I managed to dry everything out without any damage but I will obviously have to reseal all the cable entry holes - I've still no idea where the water got in. The program is now on its sixth version. I managed to shorten the instructions. I changed the size of the rotating disks a couple of times - too big and there is no sea visible to frame the models, too small and there is so much sea that people keep using it as an ordinary telescope without putting money in! I completely abandoned my nice effect of switching off the light between scenes, because seeing the joins intuitively gives people the idea of turning the telescope. As it became easier to use, people spent longer looking through it, so I had to increase the time. Then people didn't seem to know when it had ended, so I added another sound effect to round it off. It makes a big difference having a full stop, people often now come out laughing.