NOTE the sentence: These are not instructions on testing it for safety in use, only for fault tracing.

This describes the typical layout of a violet wand in terms of its three 'circuits'. For each circuit I give simple fault-finding tests. I also explain about its mode of operation in simple terms.

These tests can be done by anybody who understands them as they don't use mains power. You need a multimeter or some other low-voltage tester that can tell the difference between an open circuit and a resistance of 100k. With a multimeter you need the high-resistance position to check for resistances that should be high to infinite, and the lowest position for the resistances which should be low to zero. If none of this means anything. you may want to stop reading here. These are not instructions on testing it for safety in use, only for fault tracing. If you can identify a component that needs fixing or replacing, you may be able to make it work but not to be sure it is safe. To do that requires more specialized tests and a greater level of knowledge and I don't think this is something which can be given in a FAQ. Lastly I'm describing 'typical' arrangements and I've only taken apart two VWs. I'd be surprised if the basic circuit varies in any important detail from what I describe, but I put a section called EXCEPTIONS at the end to list common known differences which might affect the tests.

THE SUPPLY CIRCUIT
The primary circuit, the path which the mains power takes, consists of the following components in series, though their order can vary. One power lead goes to the on-off switch (in dual voltage sets, to a two-way switch). The other pole goes to the coil in the box. For a two-way switch, each of the other poles goes to a different tapping on the coil. The core of the coil (usually a bundle of iron wires) has a vibrating contact near one end, an iron reed or spring with a fairly hefty contact at the end. The other, 'fixed' contact is on the end of the adjusting knob, which has a threaded spindle so that as it is turned up this contact is brought nearer to the spring contact. The other end of the coil winding will go to one of these switch contacts, and the other contact will go to the other supply lead, completing the primary circuit.

This circuit acts like a buzzer or an old-style electric bell. At rest the contacts should be closed so that the supply is connected through the coil. When power is supplied, current starts flowing in the coil and the magnetic flux causes the spring contact to be attracted towards the coil's iron core. This pulls it away from the 'fixed' contact and breaks the circuit, and the coil's magnetism decays, allowing the vibrating relay to spring back into contact and the current to flow again. And so the cycle begins again, causing the buzz which an old-style mechanical VW produces. The fact that mains supply is AC could complicate matters, but the contact time is shorter than a half- cycle so some contact periods will have positive polarity and some negative.

HOW TO TEST THIS CIRCUIT

Connect a multimeter on a resistance setting across the power supply leads. This is INSTEAD of the power, ALL OF THE TESTS ARE DONE WITHOUT CONNECTING TO THE POWER SUPPLY. Switch on and turn up the adjusting knob. The meter should register a resistance, it should be a few hundred ohms. If the resistance varies between infinite and normal as you turn the knob, the contacts are burned. If the meter reads infinite resistance at all settings, probably the coil is burned out or some wire has come adrift. This needs to be checked with the compartment which the power lead goes into opened.

If this can be easily done, check for any obvious damage like connections come adrift or perished insulation. Then the next thing to do is to NOTE DOWN what wires connect to what, so that even if you don't know which bit is which, you can reconnect anything you have to disconnect. Drawings, dabs of marker pen on wires, little bits of adhesive label wrapped round, whatever you do you need to make sure each wire and connection is identified, and which others it connects to listed.

Next, you should be able to trace the primary circuit from the power leads through the on-off switch, the coil and the contacts, and should be able to check each individually. The coil measured directly across its leads should read a hundred ohms or so, the contacts zero ohms, at least at some position of the adjusting knob. The last test to do is to check that with the on-off switch to Off, the resistance across the plug is infinite. This checks the switch and that the mains lead is not shorted.

If the coil resistance is infinite and no visible damage or break in connection is visible, it will be burned out internally. This is possible but unlikely. Make sure that you have made proper contact with each end, the coil wire will be lacquered except at the point of contact, and if you need to make a connection further in you must scrape or dissolve off the lacquer.

If the contacts give an intermittent reading over the entire travel of the adjustment knob they are burned and need to be cleaned. I doubt that burning would be so bad as to keep the contacts open at every setting, in such a case I would check that you really are connected to both contacts. If so, something like corrosion or misalignment is keeping the contacts apart.

Cleaning the contacts.
The contact on the adjusting knob moves outwards the more you turn the setting down. There should be end stops of some kind to stop the knob being turned too high or too low. If it's possible to remove the end-stops the knob can be turned anticlockwise until it can be taken off and the contact inspected. The vibrating spring contact will be clamped at the other end so that it is insulated. Remove the clamping screws or bolts, noting the order of the insulating parts. If you find that the spring contact is not properly insulated from the fixed contact or its metal mounting this is probably the cause of the failure. Once you have both contacts free you can clean them. Assuming they are of the same metal throughout you can file them down until clean metal is found. One contact surface is usually flat and one very slightly domed, though both can be slightly domed. Don't use emery paper or 'wet and dry' or other abrasive papers. It's tempting to pull a strip through the contacts while still assembled, but this will probably embed particles in the contact metal which will insulate them. The serious pitting and burning you may find makes a proper job with a file worthwhile. After filing, and even if none was needed, the contacts will need to be adjusted. It may be possible to change the packing on the spring contact mounting to move it up, but if you add extra pieces be sure that it is still insulated from other parts. A better way is to remove both end stops and to find some way to replace them when the new setting is found. When the VW is working and safety tested, you can set the contacts just apart. Switch on, and turn the adjusting knob up until the contacts start buzzing. Turn up a little more until the buzzing is regular. That position is the low end of the scale so find a way to mark the position. I don't know how to test for the high end point so I just say that it's clockwise from the low point by the angle of the scale if there is one or by the angle between the original end stops. Replacing end stops in their new position depends on how it was done originally, if there's a rod coming out of the spindle with two stops drilled into the chassis you can probably drill and fit stops in the new position. Bear in mind that the adjusting knob has a metal spindle which is live, so don't make any metal attachment to it that can be touched. Better in fact if you can find a way to cover the exposed metal.

THE INTERMEDIATE CIRCUIT
There is a second circuit, which flows through the lead between the box and the handle. This starts at one of the contacts, goes through a capacitor (which in the days of old VWs was called a condenser and may be so marked), and then to the lead to the handle, round a coil of a few turns in the handle, and back down the lead either to the other contact or to the primary coil and from that to the other contact. The capacitor is a block with two terminals or wires, possibly paper and wax covered.

HOW TO TEST THIS CIRCUIT.
The leads going to the handle should give a zero resistance. The capacitor should give a high resistance though it will be shorted by the other components and one of its leads must be disconnected from the circuit to test it. A sensitive meter may take a fraction of a second to stabilise at a fairly high value of resistance, some hundred thousand ohms. A modern capacitor's resistance would be unmeasurably high but the old ones with a paper dielectric will be much lower. A much lower value, say 100,000 ohms or less, indicates problems. Equally if your meter is sensitive enough to show tens of millions of ohms and there isn't a flicker as you connect it, a connection has probably come loose inside. If you suspect this you need to connect the meter to the capacitor without holding it at both ends, or your resistance will be measured instead of the capacitor's. This is not a very good test, but can show up some problems.

Next you need to test that the lead from the box to the VW handle is OK, and the coil in the handle which connects to it. To be sure of this you need to open the handle. There may be two screws or bolts towards the back of the handle which will allow the end cap to be removed. If you're lucky, the whole high voltage assembly can then be slid back out of the handle. Whatever the case you should be able to see both ends of the lead from the box to the handle. If either end has perished insulation it should be replaced.

Assuming the lead looks good, disconnect one of the lead connections at the box end. Then see how the other end is connected to the handle assembly. If by terminals you should be able to undo the same colour connection of the lead, and then you can test across the lead wires to check for infinite resistance. If zero and you've really disconnected the same wire at both ends, the lead is internally shorted and must be discarded. Next check across the terminals to see that the coil with a few turns gives a zero resistance.

THE HV (High Voltage) CIRCUIT
If all looks OK up to now, we move on to the last circuit, and the one which unfortunately is most likely to be at fault after the capacitor and wiring. This consists simply of the HV coil, many thousands of turns of thin wire. One end will be connected to one of the terminals of the primary, and the other to the metal socket which the glass electrodes fit into, possibly via a spark gap.

You may be able to open the handle (by unscrewing two screws or bolts at the back of the handle for instance) and slide the whole HV coil assembly out. Other times the whole thing seems to be 'potted' into the handle with melted wax. Whether you can get at it or not, what is in the handle is a step-up transformer. The primary is about thirty turns of moderately thick wire, wound on the outside of the secondary of many thousands of turns of very fine wire. This winding is carefully designed and layered to handle the high voltage without internal shorting, making sure that only turns which are close together in terms of position along the wire of the winding are physically close together. That way turns which are physically close are also close in voltage so that insulation breakdown is less likely. The low voltage side is at the outside of the coil, and the high voltage side is on the inside and towards the electrode end. There may be a spark gap, more likely in later models, consisting of some sort of arrangement of two plates of metal separated by a small air gap of less than a millimeter.

If you can get at the assembly inside the handle, you can check the outer primary winding. If there's a fault with it it can be replaced with wire of a similar diameter and thickness of insulation. It would be safe to use thicker wire or more heavily insulated wire, but it might upset the efficiency of the wand if it changed the resonant frequency.

If there's no spark gap, there should be a measurable resistance (a few thousand ohms) between the electrode socket and one of the supply leads when the VW is switched to ON and the other components are working. If there is a spark gap but you can get into the handle, you can test this from the HV end of the coil, bypassing the spark gap. If resistance is infinite and it's not due to a spark gap, the HV coil is broken or burned out. The break may be small enough for the voltage to jump over it in use, but it is not an ideal situation and could cause more damage.

If there's arcing across a breakdown path on the outside, it may be possible to remove damaged insulation and replace with new. But if it's inside the coil there's nothing that you can do as far as I know. Arcing will greatly reduce or completely stop the voltage output, and could cause overheating and get worse in time. A small break in continuity might not noticeably affect voltage output, the current will find its way across any extra small gaps, though they might do damage depending on where they are. A coil can burn out and this may create large gaps, or a large number of small gaps,and very possibly damage the insulation and cause it to short internally.

One thing which can happen in any of the three windings is SHORTED TURNS, where adjacent turns come into electrical contact due to insulation failure. This causes a high current to be induced in the shorted turn, causing local heating and draining the available power. It is not something which can be tested for with simple equipment but if there are no signs of obvious overheating and you identify other faults, once they are fixed you would be unlucky if this fault was also present.

REPLACING A FAULTY CAPACITOR
Changing the capacitor value changes the resonant frequency. A larger capacitor doesn't necessarily give a larger output, and a smaller one than the original might give the best voltage as it might 'tune' the intermediate circuit to the resonance of other parts of the circuit. It's not at all critical, but if you have a range of suitable capacitors try them for best output. It's probably not a good idea to use a capacitor very much larger than the original one. Equally the longest or brightest spark may not be the only thing to go for, it may be within the original design spec but if it causes the elderly insulation to break down it wouldn't have been worth the trouble. But that's a risk you take with any old equipment. The typical capacitor has a value of 0.1 microfarads and should have a DC working voltage of at least 1500 volts.

EXCEPTIONS
If a VW has a three-core supply lead, ie it has a ground/earth connection, this may be used as the return path for the high voltage. If these exist they are probably uncommon. Modern VWs will most likely have a ground connection but they are electronic and quite different from the ones I'm concerned with here.

Some VWs have all the working parts in the handle, in other words the lead to the handle carries the mains supply, the coil with the vibrating contact is in the lower half of the handle and the step-up transformer in the upper part. I don't know if the basic circuit is any different due to this.

SAFETY FEATURES
Don't use an isolation double-wound transformer with a WV. The low voltage side of the HV coil needs to return to earth, and if the easiest way back is between the windings of the isolation transformer, that's the path it will take.

An RCD may add safety, but VWs often trip them without being faulty.

HOW VIOLET WANDS WORK.
The supply current is connected to a coil via a pair of normally closed contacts. When first switched on the contacts are closed and current starts to flow. A coil has a property called inductance, and the nature of this property is that if the current flowing through it changes, a voltage proportional to the rate of change is induced in the coil, in the direction opposing the change. When you switch on, the initial rise in current is limited by the reverse voltage opposing the supply voltage. Therefore the current takes a certain time to build up. In fact, though the relation of current change to voltage is true, it is slightly more complicated than that, because the current through a coil creates a magnetic flux, with density proportional to the current and the number of turns in the coil, and it is the change in the magnetic flux which generates the voltage. So as the current grows the flux density in the coil grows also, and this attracts a spring on which one of the contacts is mounted, breaking the circuit.

Now the energy used in getting past that opposing voltage is stored in the magnetic flux and it has to go somewhere. Breaking the circuit is going to stop the current flowing abruptly, and the property of inductance is again manifested. This time the current is very quickly changed in a downwards direction, and the voltage induced is correspondingly high. The energy which was stored as the input current slowly built up is now suddenly all returned as a high voltage and high current surge.

If it helps, you can imagine inductance as similar to momentum. It takes a certain amount of force to start a flywheel or a top turning, because changing its speed from rest requires energy. Once turning it will continue without extra force. Now if you grab it to try to stop it instantly you're trying to return it to rest instantly. The energy you put into it to start it is returned to you in a fraction of a second, possibly as a friction burn. In both cases the force that is exerted opposes the change, when you're starting it it 'wants to' stay stopped, and when you're stopping it it 'wants to' keep going.

Incidentally, the way in which the adjusting knob regulates (more or less) the output is I believe by altering the time it takes before the contacts open. The less the spring contact presses on the fixed one, the quicker the magnetic flux can pull them apart and the less current and flux can have built up in the coil.

If we connect a second coil via a capacitor across the contacts, much of this energy will take this path. A capacitor stores energy in a related but different way to an inductor. Its property is that current flows easily across it at first, but the more current flows and the longer it flows, a proportional opposing voltage develops and energy is stored. These two properties, inductance in which a current is hard to start and gets easier, and capacitance in which a current is easy to start but gets harder, are opposites. A capacitor and an inductor together can bounce current to and fro between each other like a ball. Or more like a piano string, because they have a frequency at which they are equally balanced and work best.

What happens in this case is that the burst of energy from the first coil at the opening of the contacts is thrown back and forth at a high frequency.

The last part of the story involves just inductance. The coil which is throwing the.., let's use a grown-up word, resonating, with the capacitor, is wound round a bobbin which it shares with another coil, so that the magnetic flux through them is the same. They both have the property of inductance, in which the rate of change of current induces an opposing voltage. The current flowing through the first coil is changing at high frequency due to resonance, and is causing a proportional magnetic flux. The second coil, sharing that flux, has an opposing voltage induced in it. Just as the flux is proportional to the current multiplied by the number of turns, the voltage which opposes it is also proportional to the number of turns. Each turn of the second coil has the same voltage across it and the more turns there are the higher the voltage is.

The second coil has a very large number of turns, and the induced voltage in it is very high and this high frequency, very high voltage is what we buy VWs to get.

Now it's well known that energy is conserved, you can only transform it, you can't destroy it or create it from nothing. So some people may be wondering how you can get this high voltage from normal supply voltage. Well, the payback comes when you come to use it. If the very high voltage in the second coil has to do any useful work, it has to move, or a current has to flow. And that current creates an opposite magnetic flux counteracting the one which the first coil is creating. What's more, because the second coil has so many turns and magnetic flux is proportional to current times number of turns, the current needed to counteract the original flux is correspondingly smaller. This balance means that current is effectively limited to the current which will completely counteract the flux created by the first coil. This has the incidental advantage that we don't fry ourselves or our victims.

In a perfect transformer, the voltage difference between the primary and the secondary windings is in the same ratio as the number of turns (each single turn, on either winding, has the same voltage across it. And also, for the reasons above, an output voltage of N times the input voltage can only give a current of an Nth (1/N) of the input current. As the energy over a given time is proportional to the voltage multiplied by the current, this will also fulfil the law of the conservation of energy, that the energy you get out can only equal what you put in. In practice a transformer will have losses and these are theoretical best values which real components can approach but not equal.

Another question is how does the AC of the supply voltage affect all this. VWs always worked on AC or DC supplies in the days when some supplies were DC. The things that happen in a VW, the build-up of current in the first coil and the frequency of the resonance in the next coil, all happen much faster than the supply frequency of 50 or 60 Hz. The buildup of current and flux will tend to get synchronized to the mains frequency, as it happens fastest when the supply voltage is at its peak, but the contacts will open in the same half-cycle. Different discharges will occur in different (positive or negative) half-cycles but the whole life-cycle of a discharge occurs during one half-cycle so the supply is effectively DC for it.