The Atomium Nixie Clock

If you have ever been to Belgium and visited Brussels one of the attractions there is the Atomium.

This was built for the Expo 1958 Brussels World Trade Fair and was designed by Andre Waterkeyn. Originally constructed in Aluminium it has since been renovated in 2005 and is now Stainless Steel. It stands 102M (335ft) high and consists of nine 18M (59ft) diameter spheres connected to denote the shape of a single Iron crystal magnified 165 billion times!

It was this design that gave me the inspiration for the Atomium Nixie Clock, albeit a much simplified design. My design consists of the six outer spheres connected to a middle sphere with the clock figures in the outer spheres and a Dekatron spinner in the central one.

My 'spheres' are actually tubes made from 50mm white oak dowel with brass inlays and are interconnected with 8mm brass tubing.

OK, on with the show!

My apologies for the long winded Instructable but it is to enable others to be able replicate this clock in detail.

Step 1: The Atomium Clock Materials Listing.

Where possible I have added sources where the materials can be found.

1M of 50mm White Oak dowel (local timber supplier)

1.5M of thick walled 8mm OD Brass tubing (Ebay)

8 x 3/4" Brass back nuts (local plumbing parts supplier)

2 x 1" Brass back nuts

2 x 3/4" to 1/2" Brass reducers

9/16" Brass round bar

Offcut of Spalted Chestnut or any hardwood of choice (local timber supplier)

6 x GN-4 Nixie Tubes ( Ebay)

1 x OG-4 Dekatron tube (Ebay)

1 x DINK Remote Tube Nixie Clock kit (PV Electronics UK)

1 x Dekatron Spinner kit (threeneurons @ Ebay)

1 x 12VDC 1A Wall Wart (PSU) (PV Electronics)

4 x 11mm PCB Stand Offs with screws (PV Electronics)

22 AWG Wiring (Ebay) (0.7mm can be used as it is more compact)

6 x B13-B Nixie Tube bases (Ebay)

3mm Heat Shrink (Maplins, RS Components, Radio Shack, Auto Electrics store etc)

Wiring Identification kit (RS Components)

1 x Custom Engraved Plaque (Engraving Studios UK)

Assorted small brass fittings for shrouds on the Neons and switches (made by me)

Assorted screws, brass and plated

Satin finish Yacht Varnish (DIY store)

Step 2: Where It All Happens!

A bloke just has to have his special place and this is mine!

I built up my gear over many years with the lathe and mill being the last added a few years ago. I have tools dating back over 45 years and I still accumulate more! Hopefully I will convert the mill to CNC and also build a flat bed CNC router but where I will put it I have no idea!

My wife gives my shed a wide berth apart from getting the watering can out!

Back to the clock now!

Step 3: Tooling for the Atomium Clock.

As you would have seen, I am fortunate enough to have a reasonbly well equipped workshop. I know that not everyone will have access to some of the tools but things can be outsourced to friends with similar tools.

Hobby metal lathe - used for the woodwork too!

Hobby Milling machine, like wise abused for wood

Small pedastal drill

Bench top grinder / belt sander

Dremel

Soldering Iron

Electronic side cutters

Angled nose pliers

Tweezers

Phillips No 0 screwdriver

4mm Flat bladed screwdriver

Mini ratchet screwdriver

Multimeter

Bootlace ferrule crimping tool

Forstner bits, Drills, Slot drills, End mills, Countersink

Measuring tools, Ruler, Micrometer, Vernier, Tape measure

Metal polishing kit, Sandpaper, Emery paper, Diamond files

Squeeze clamps, Spanish Windlass, Crocodile clips

Step 4: The Atomium Clock Base.

The Spalted Chestnut offcut that I got from the local timber merchant was the ideal size from which to make the base. You can find good deals at timber merchants for offcuts of various hardwoods, have a look or search on the internet for them and save a few Euros, Dollars, Pounds or whatever. After cutting what I needed from the piece I then drew a circle making the most of the size but allowing for machining off the rough edges.

I cut most of the waste off by hand and then drilled a hole in the center to take a mounting bolt to secure it to the milling machine. Having it secured I hand fed it against the rotation of the cutter which stops it getting dragged out of your hands and gives a better finish. As it was thicker than the milling cutter I had to reposition it when turning it over to ensure that the cut direction was the same as the first one.

Once this was done I need to take out a slight bow in the wood and face off both sides to get my desired thickness for the base.

Making the pocket for the main circuit board was done on the milling machine and details are on the photographs. Another pocket was created for the Dekatron board which measures a tiny 2.3 x 2.3 inches and one for the clock control switches.

On the "Hands of Time" clock I fitted bun feet that were made from 1" plain dowel with a brass ring in the middle. To make these I rounded the free end of the dowel in the lathe with sand paper and then using a parting tool I cut the 'stalk' of the mushroom. The diameter of the stalk is the same as the inner diameter of the 3/4" brass pipe repair tube that I sliced to make the rings. The length has to be able to hold the brass ring and have enough protruding to get glued into holes in the base section. The bun feet for this clock will be made from the dowels I created from the stair spindles.

The hole in the center of the base will be covered by the name plate with another hole further back that will be used later to pass the wiring from the circuit boards to the Nixies and Dekatron tubes.

Step 5: Woodwork and Brass Work Part 1.

If you have seen the Nixie Clock No.2 that I made then you will see the basics of the preparation of the oak tubes. This time they are a bit more complex internally with a shoulder and reliefs inside to enable the mounting of the GN-4 Nixie tubes I used.

The same methods were used to cut the dowel and make the holes for the nixie tubes, 25mm Forstner bit as a starter and then a boring tool in the lathe for the two different diameters at each end of the tube

The mounting holes for the connecting tubes were made using a rotary table on my hobby milling machine with pairs of holes at 120 degrees and an additional pair at 60 degrees between the outer pairs. Using the rotary table made easy work for the holes and ensured accuracy that would be needed for the assembly.

The 3/4" brass back nuts were sized on the lathe to remove the taper on the edges, then the 'back face' was machined flat and then the thread was bored out to give clearance over the GN-4s. After that the hex was removed to reduce the size of the mounting hole for them. The oak tubes have a recess at the front to enable the brass fittings to sit flush when mounted. The brass surrounds are epoxied into place after fitting the GN-4 nixies. Before fitting they are polished yet again and a coat of clear spray on lacquer. How to get a good finish with lacquer

Satin finish yacht varnish is then applied to the oak tubes. This is thinned down with white spirit at a 70% varnish to 30% white spirit ratio, this allows the wood to pull the varnish in deeper. I prefer yacht varnish to polyurethane varnish as it has a far better quality to it. After a few coats of this with sanding in between coats, the last coat being 100% varnish, it allowed to dry thoroughly for 48 hours.

The center section of the display is longer than the tube carriers as the Dekatron is a much bigger unit. I was fortunate to get exact details of the size from the Ebay vendor and machine the tube before it was delivered from the Ukraine. (this was after a botched attempt at guessing the size!) The wood used was different as I was not going to lash out for a 1 Metre length of white oak as I only needed 80mm. I might stain this to take care of any colour difference between this and the white oak of the Nixie holders, then again, maybe not!

The brass tubing that links all of the oak tubes together was cut to size on my lathe using the tail stock as a fixed measure and the parting tool locked to the length required. When deciding on the fitting of the brass tubes to the wooden housings I decided to add some brass detail pieces instead of just the brass direct into the wood. I got some 14mm solid brass rod which I bored out to 8mm for the tubing and using a bull nose lathe tool created the 'sweep' from the outer diameter to the tube diameter. With a hexagon design, all of the connecting tubes are the same length as they form an equal sided triangle. When all the tubes were cut to length and the burrs removed I made a simple jig to hold them while they were being soldered. This failed as the solder would not flow into the joints and I resorted to changing the jig and used superglue which did flow. Once this was done it only needed them to be polished clean and The tubes are roughened on the ends that are concealed by the wood are then epoxied into place. Set everything level and apply tension with a Spanish Windlass', which is just a loop of rope tensioned by twisting it tight with a piece of wood, to compress the parts together and leave for 24 hours to get a good cure of the epoxy.

I made the end caps for the rear of the oak tubes from the square section of stair spindles I had from making the Nixie Clock No.2. Turning them from square to dowel and keeping them to as large a diameter as provided from the square section. These will be turned down to size for the oak tubes and pre varnished before final assembly of the clock. What is left over will provide the bun feet for the base section.

Step 6: Woodwork and Brass Work Part 2.

Originally I was going to have the support column as a 10mm ID brass pipe connected directly to the housing for the Dekatron but after making a mock up to see how it would look I revised this as it was looking a bit wimpy and went for plumbing fittings instead.

I used 2 off 1/2" x 4" brass pipe repair tubes along with some reducers, back nuts and a 1/2" bend to make the support column and it looks far better than the 10mm ID pipe original.

As the fittings are castings that are then threaded it was necessary to remove the rough surfaces and size IDs from them, this was done on the lathe and all the hexes were removed as well to give a more aesthetic look to the assembly. At 4" the repair tubes were not high enough so I removed about an inch from each of them including the thread. The cut off section of one forms the attachment from the elbow to the back of the Dekatron and is soldered into the 3/4" to 1/2" reducer after enlarging the 1/2" hole in the reducer.

The plain ends of the pipes were turned down to form a socket connection which is soldered together to get the height required for the display section. Both 1" back nuts were drilled for screws to fix to the Dekatron housing and the base unit. All interior edges were chamfered to remove any burrs that could damage the wiring.

Polishing off the rest of the casting surface on the 90 degree bend was done with a Dremel and a small sanding drum. The pieces were buffed up on a felt wheel mounted in the milling machine with fine abrasive then metal polish applied to all parts before coating with clear lacquer to protect them. There is a photgraph showing the kit I used for polishing, this is available under many brand names but is the same kit and it pays to shop around. I got mine in LIDL for less than 5 Euros, I have seen other sites charging as much as $30 for the exact same kit in a different package!

Once the wiring is completed in the display section it is just a case of feeding it through the support and fixing this to the rear of the Dekatron with 6 brass wood screws. I was then able to fit the support and display to the base, securing it with brass machine screws which were tapped into the base section and then tailor the wiring to the circuit board.

I needed to make brass push buttons to activate the switches for the clock settings which are remotely mounted from the circuit board. As I had plenty of brass left from the detail pieces of the brass tubes I did a quick sketch and then made the parts. Finding 3 identical springs was a piece of luck because I had ordered a bag of mixed ones for another hobby. They proved to have the ideal compression as well as length for what was needed and, when the button was assembled, worked perfectly.

There are two LEDS that are capable of indicating several functions, AM, PM and the Seconds pulse. I made miniature 'Bulls Eye lanterns' to house them and secured the LEDs in position. Once they were central I used a clear twin pack epoxy to make the 'lens', the air bubbles from mixing help to diffuse the glow of the LEDs and give a nice effect. These were then fitted to the base section and glued into place.

Step 7: The Atomium Clock Electronics

PV Electronics provide first class kits for Nixie clocks and the DINK Remote kit is no exception. All the instructions for assembly are straight forward if you have some basic electronic knowledge and can solder well.

I left off the switches and LEDs as these will be mounted elsewhere on the base and the power inlet socket will have to be removed from the board as it needs to be on the base plate.

I prefer end view neon lamps to the vertical ones but this clock will use 3mm LEDs and these will be shrouded in brass fittings that have been machined to mount them. The wiring comes down from the top of the base and then goes through the hole to the circuit board.

The circuit board for the Dekatron Spinner is easy to build and this is available from Three Neurons via Ebay. This will be remotely wired to the Dekatron in the same way as the GN-4s. It will share the 12VDC supply of the Nixie board as it is capable of powering both and more.

From the photograph you can see how compact the fitting of the electronics are in the base section As I did not have any more stand offs for the Dekatron board I just made some delrin spacers and used 2.5mm screws to mount it directly to the base into 2mm holes Screwing directly without tapping the holes gives a strong enough mounting but to make the 'threads' more durable I let a little thin cyanoacrylate soak into the holes and this hardens the thread.

I decided to mount the clock controls on the top surface of the base unit and made a recess opposite the Dekatron recess. 3 holes were cut from the recess to the surface of the base with a slot drill to get the position correct in relation to the rest of the electronics.

I was intending to use the original control switches but they were a different pitch from the strip board I had and they were substituted for panel mounted types. These were fitted onto supports made from thin stainless steel strip and screwed into position. The switching point was set by adding collars to the switches and fine tuning done by bending the mounts slightly by pushing down with a screw driver.

I was going to mount the power inlet socket to the base cover plate but fitted it to the base section, unfortunately due to a component being central on the Dekatron board it had to be mounted off center. It is at the back so not too much of an issue!

Step 8: The Dreaded PIA Wiring Up Section!!

As you might have guessed, this is my least favourite part of these clocks after the experience I had with the Nixie Clock No.2, but it has to be done!

With the display section finished the next step is to fit the wiring for the Nixies and Dekatron tubes to it. Fortunately, PV Electronics use a 'daisy chain' method for supplying the voltage to the cathodes and this cuts down the number of wires required immensely. To make life easier with the wiring I use a double feed method which sends 8 wires (5 cathodes and 3 anodes) to 2 opposing oak tubes in the display and then it is just a case of cross wiring them.for the Cathodes and connecting the individual Anode wires to each nixie tube (see drawing).

10 cathode wires and 6 anode wires from the nixie circuit board power the Nixies while the Dekatron uses only 8 wires which gives a total of 24 wires going to the base unit. 50 wires form the 'daisy chain. between the nixies tubes with the 6 anodes directly wired back to the circuit board.

Wiring up the Nixies is a total PIA! I found that out with the Nixie Clock No.2 and also doubly so with another clock I built called the "Hands of Time". Why I don't just do board mounted tubes I don't know, it would be a hell of a lot simpler than these nightmares!
Using B13-B bases however, makes it a bit easier and because they are not in a blind hole as was the Hands of Time, I allowed 'storage' space for the wiring behind the support for the bases in the oak tubes so it was easier as I could do the soldering and heat shrink fitting from the front of the oak tubes.

When connecting to the B13-B bases, the wires are paired to each terminal and heat shrink is slid on before soldering to the tags on the bases. I fit wiring ID Nos. to keep track as I use a single colour instead of buying 10 different coloured coils of wire.

The feed wires to the circuit board are left long until final fitting when they get tailored to length and bootlace ferrules crimped to them and secured to the terminals of the circuit board of the clock and in the case of the Dekatron spinner, soldered to the base pads of the Dekatron

The GN-4s are fitted to the bases before sliding into the oak tubes where they are fixed with hot glue then quickly orientated for alignment and finally the prepared brass collar is fitted with epoxy. Alignment of the Dekatron is not crucial as it is a spinner and is connected using individual sockets on the back of it wired down to the circuit board. Once the wiring is completed and the Dekatron and Nixies are tested for function.

When the LEDs were powered up they were a bit on the bright side and detracting from the warm glow of the Nixies. I changed the value of the resistor for the current limiters for the LEDs and then they gave off an intensity in keeping with the Nixies.

Step 9: The Atomium Clock Wood Finishing and Final Assembly.

The display section is supported by brass tube assembly and this screws on to the back of the Dekatron and onto the top of the base section.

Once everything was tested and proved to be working it is taken apart and final finishing applied to the base with the yacht varnish

. The brass work gets polished and coated with clear laquer then is fitted followed by the neons, circuit boards, power inlet and the Atomium section of the clock. The wiring is fed through the support tubing and into the base section. Connecting the wiring is simple using the terminal blocks and having bootlace ferrules on the nixie / dekatron wiring. The circuit board is secured to the stand offs with the wiring tucked underneath it

Connect the power, select the operating parameters and the Atomium Clock is up and running. Final thing is to fit the covers to the rear of the oak tubes and the cover to the opening in the base and then have a step back and a look at the finished clock.

Step 10: Post Script.

While working on the Atomium Clock I needed to make a partial circle for another project and rather than hand cut and then mill I decided to mill it from the start.

Always feeding into the cutter stops the piece getting grabbed and with several passes adjusting the depth each time it made short work of the process. This was done using a slot drill and then a finishing pass with a same size end mill The top and bottom surfaces received a quick clean up cut with the end mill and this also took out any warp in the wood at the same time.

Mounting a flap wheel sander to the chuck I was able to get a very good edge finish and then milled off the waste from the diameter to get the final piece as shown in the last photo. Minimal finishing is needed this way.

You might think that you need to have extensive skills to do projects like this but they are all gained as you work on them. I could hardly make a straight cut in wood at one time but keeping at it helps you develop those skills.

The old adage applies - "You never know what you can do until you try!"

I hope that this project inspires you to be creative and think out of the box.

Thanks for taking the time to read this.

Roddy.