Building a Wimshurst Machine – Part 2.

October 27th, 2008

I finally got around to finishing off the Wimshurst machine the other week and it’s about time I updated this page about it.

Part one of this project is available here:

There were a few changes made to the machine since the last post. The main one was I discovered glass Leyden jars don’t work very well! Well, mine didn’t at least. I re-made the jars using plastic containers and these seems to work much better.

newjar Plastic Leyden jar.

They are made the same way as the glass jars and the inside and outside are covered in aluminium tape. I machined up some plastic discs on the lathe to use as lids. Using the original plastic seal the lids fit tightly to the top of the jars. I used some copper foil tape to ensure there is a good contact between the foil on the sides and the base of the jar.

I also made sure the main discs spin true on the hubs.  Between the hub and the discs I use three small rubber O-rings. These allow me to make sure the disc is spinning true by tweaking the tension on the adjusting screws. I made a temporary jig and using a DTI (dial test indicator) was able to measure the run-out on the edge of the disc and adjust it until this was minimised. You can never get it perfect but this step helps a lot.

runout Adjusting the main disc run-out.

Once I had the discs and main supports made I started on the base. I made mine from MDF with wooden curtain rod ends (I got for 50 cents each!) as the feet. The base is two rectangles of MDF glued and nailed together. I make one slightly smaller than the other and use a quarter round molding in the corner between the two pieces to give the base a nice appearance. In the middle of the base is a rectangular cutout which is where the main drive pulleys sit. The pulleys are attached to a shaft that runs in bearings in the main supports but as they sit lower than the base of the supports you need space underneath them so they can turn freely. I did this to keep the supports as low as possible. On each end of the base is a wooden box for holding the Leyden jars. This has a top plate with a hole cut in it to allow the jar to slide in.

basedetail Wooden base and Leyden jar holder.

Attached to the back of each jar holder is a support for the collectors. This was made in the same way as the main supports using plastic cutting board and a wooden template on the router. In the front of each box is a brass terminal that pokes through the box. On the inside of the box there is a piece of springy brass which contacts the outside of the jar when it is inserted.

jar1 Leyden jar and collector detail.

The terminals on the front of each Leyden jar holder can be connected together wiring the jars into the circuit of the machine. This allows the charge to be collected giving you much bigger and stronger (and more dangerous) sparks. I used wires with banana plugs on each end to connect each jar to a central terminal plate that sits under the handle of the machine. This central terminal is simply a brass plate with two brass banana terminals attached to it. The wires are coiled for appearance.

base Terminal plate for connecting the Leyden jars.

The  collectors are what pick up the charge from the spinning discs. I used brass doorknobs as my collectors and these form the central point where the combs, the main terminals and the connection to the Leyden jars are made. The brass doorknobs are arranged to sit directly above the center of the Leyden jars and also in line with the center of the main discs. There is a hole through them vertically through which the main terminals rods slide. The main terminal rods are 6mm aluminium rod that are bent in a curve over the top of the main discs. On the ends are brass terminals which I will describe later.

I used a piece of aluminium tube between the base of the collector and the base of the Leyden jar. A small O-ring between the collector and the aluminium tube ensure that there is a little tension between the collector and the base of the jar ensuring good electrical contact. Just beneath the collector the aluminium tube has a threaded hole into which a plastic adjusting handle screws. The handle, when screwed in tight, attaches the tube to the main terminal rod. By rotating the handles you adjust the spacing of the main terminals. The O-ring between the tube and the collector also ensures the rods don’t rotate too easily and move out of position. The handles are simply plastic tube with a brass screw epoxied to one end and a plastic ball (from a roll on deodorant bottle) hot glued to the other to finish them off.

Coming off the side of the collector terminal are the main combs. I made these from welding wire bent into a U shape. The ends and intersection of the wires are terminated in lead balls (fishing sinkers) soldered in place. This is to hide any sharp points that will leak off charge. The actual collectors I made from aluminium tape which has a series of points cut into the edges that face the main discs. I used pinking shears to cut the points. Both the wires and the aluminium tape can be easily bent to allow the combs to be positioned close to, but not touching, the main discs. If the run-out on the discs was properly adjusted the gap between the combs and the discs should remain fairly constant as the machine turns.

collectors1 Collectors.

The main terminals are covered in plastic tubing to insulate them and to make them look better on the machine. The terminals on the ends are made from more brass doorknobs. I cut off part of the base to make a sphere with a hole in it.

terminal_1 Brass terminals.

To cover the hole in the sphere I soldered a nut to a small, circular piece of brass which was in turned soldered to the sphere. This gives me a brass terminal with a screw thread in the base. The ends of the aluminium terminal rods were threaded so the brass terminals can be screwed on.

terminal_3 Threaded fitting on terminals.

The image above shows the terminal being soldered up. After soldering the terminal was sanded smooth. The nut seen here was removed as it is just used to hold everything in place during soldering. The nut that the rod screws into is on the inside of the terminal.

The next component of the machine is the main driving shaft, hubs and handle. I made the hubs from more breadboard cut into circular discs using the router on a small jig again. The discs were then trued up and a driving belt groove cut into the edge on the lathe.

disc_cutting Cutting the driving hubs on the router.

Two hubs are made, one to drive each main disc, and they both mount to the same shaft. Drive belts link these hibs to the main disc hubs. To get the discs rotating in opposite directions you simply cross over one of the drive belts. The belts themselves I made from plastic polycord belts. This is a stretchy, circular profile plastic that can be cut to length then the ends welded together using a small flame. I got mine from a company that sells it as replacement belts for wool spinning machines.

To attach the discs to the main shaft I used small brass rods bent into an L shape that press through holes in the main shaft and into holes on the driving hubs.

crank_2 Main driving hubs and shaft.

The crank handle was made from a piece of steel cut into an appropriate shape and drilled. This was welded to the main shaft. A handle was made from a replacement wooden file handle from the hardware shop drilled through it’s length. A coach bolts goes through the handle and attaches it to the crank. The wooden handle is free to rotate as you turn the machine. The whole assemble runs in skateboard bearings that sit in the base of the main uprights. The bearings are held in place by large fender washers that hold them in position inside the uprights. A brass spacer tube sets the position of the shaft at the front of the machine and an acorn nut holds it in position at the back.

crank_3 The driving handle and bearings.

The final part of the machine is the neutraliser bars. These bars run across the face of each main disc, one on the front of the machine and one behind. Their purpose is to neutralise the charge across the discs and their positioning in relation to the direction of rotation and the collector combs is crucial (as in which sector they must go) to the function of the machine although their exact position (as in the angle in the sector) needs to be experimented with.

I made mine from plastic hubs turned up on the lathe. These are attached to the same shaft the main discs rotate around. The arms are made from brass rod bent into an L shape. These push into holes in the edges of the plastic hubs. The short, bent end of the L is arranged to hold a small copper brush that wipes against the sectors on the main discs.

hubs1 Neutraliser hubs.

To connect the two sides of the arms I used a piece of wire and two alligator clips. I simply coiled the wire for appearance sake and actually the wire isn’t needed as the machine seems to work well enough without it. I think there is enough leakage between the ends of the arms pressed into the hubs and the steel axle shaft for the machine to work.

At the brush ends of the arms I soldered short lengths of brass tube. Into these I insert the brushes. This took a little bit of experimenting as the brushes need to touch each sector as it passes but you don’t want them putting too much pressure on the discs making them harder to turn.

collectorbrush_2 Brush and brush holder.

The actual brushes I make from single core shielded audio cable. I cut an appropriate length and strip both ends. I then cut out the central core. At one end of the brush I bend the wires back 180 degrees. This end is inserted into the tube on the end of the arm. There is enough friction here to hold the brush in place and also to ensure a good electrical contact. The other end of the brush I simply fan out the wires and it is these that touch the sectors as the pass.

collectorbrush_3 Brush.

The arms of the neutraliser can be rotated slightly to allow the brush to be positioned so that it just wipes the sectors. A slight angle on the arms also prevents them from vibrating as the sectors pass beneath them. You do need to replace the brushes from time to time.I think the small step each brush must make over the edges of the foil tape sectors causes a slight bending of the copper wires in the brush. Eventually the wires will work harden (copper is prone to do this) and then they break off. I often find small pieces of broken brush on the base of the machine. You only need one strand to be making contact though for the machine to work. The brushes will mark both the copper foils and the acrylic discs over time. Some people have attached small dome terminals to their sectors so the brushes only touch them and not scrape the acrylic. I don’t mind the scraped look as it makes the machine look old and used to me!

One final but very important addition that isn’t part of the machine itself is a discharge rod. The Leyden jars are simple capacitors that store charge. They can hold a charge for a very long time and are actually quite dangerous. Even though the capicatence is low (around 1nF) the voltage in them is high. To make sure the machine is safe to touch you should ALWAYS discharge the jars. To do this I made a discharge rod. This is a piece of brass rod bent into a U shape with two small steel bearings (rescued from empty spray paint cans) soldered onto the ends. This rod is attached to a length of insulating PVC pipe. I always discharge the jars before touching the machine by shorting our each jar (usually with a large spark and crack) and shorting between the main terminals.

discharge_rod Discharge rod.

All of the pictures above were taken as I built the machine and when I first had it working. I finally got around to painting it and also making a few small changes. I increased the diameter of the main driving hubs to be as large as possible. The bigger these are in relation to the small driving hubs the faster the discs spin. The speed of the discs affects how quickly the charge can be built up. I made new neutraliser hubs using black plastic to better match the rest of the machine. Also, in a fit of maker vanity, I had the small terminal plate engraved with my name and date for posterity. One change I would like to make is getting fabric covered wires for the machine.

Now finished the machine works very well. The largest size sparks I can get from it are about 150mm long. This is with the jars attached to the circuit. This being a static electricity generating machine it’s performance depends a lot on the weather. If the humidity is too high you get a lot of leakage off the machine and it can’t built up as high a charge.

When operating well you can hear the charge building up in the machine as soon as the handle is turned. The room fills with the small of ozone and the hairs stand up on the backs of your hand. Close to the machine you can feel a prickly wind coming off it and if you run it in a totally darkened room you get a nice corona glow from various parts of it. With the Leyden jars attached and the humidity low the resulting arcs are very strong and loud. You need to wear hearing protection to avoid damaging your hearing from the crack of the discharge! I imagine there is a lot of UV emitted from the arc too.

There are some YouTube videos of the machine running available here. The first was the unfinished machine generating 150mm long sparks on a particularly dry day.

Other videos are available if you look at my other videos on YouTube.

Below are pictures of the finished machine which  hopefully make some of the details in the text clearer.








Posted in Projects | | Top Of Page

6 Responses to “Building a Wimshurst Machine – Part 2.”

  1. David Says:

    The paint didn’t reduce the resistance then? Sparks are still just as large?

  2. Simon Says:

    Paint didn’t make any difference actually. It’s more affected by humidity!

  3. chai Says:

    thank you very much

  4. luiz alberto feijo jr Says:

    Nice job! I suppose yuo learned we can never trust in glass as an isolator – in fact most of them do have lead mixed on this composition – and this makes it a little conductive. This problem is bigger than the hygroscopic property of the glass in most cases! And some plastics (specially the black ones ) have also conductive material added. Receive a hug from another electrostatic guy in Brazil!

  5. Simon Says:

    No, glass is definitely a bad idea as I found out! I also know about the black plastic or rather black paint when years ago I painted the water jacket of a CO2 laser with black paint then worried about conduction across it.

  6. Asciimation » Blog Archives » More on the steampunk-o-phone. Says:

    […] To cover the servo I made a small copper cover made from an old soaker plate I had lying about. It is also worth mentioning here that the drive belt I am using is a poly cord belt used on wool spinning machines (again off TradeMe). It comes in a long length that you cut to size then you melt the ends in a candle flame and press them together to form a continuous  belt. I used these belts with great success in my Wimshurst machine. […]