All posts by GardenBallistics

Bipolar Marx Generator


A Marx Generator is a type of voltage multiplier circuit constructed using capacitors, resistors and spark gaps. They produce high voltage,  high current DC pulses (non-continuous). They consist of multiple similar stacked stages where each stage effectively doubles its input voltage. The simple design and basic components allow one to that make this electrical circuit quickly and relatively cheaply.

The bipolar Marx generator I built in this project took a 15kV input and produced an output of approximately 220kV. Each discharge occurred with an excitingly load bang every couple of seconds! Unfortunately, before I could take some pretty pictures or a video of the device running I managed to get carried away and blew the device by raising the input voltage above 20kv which was too much for the maximum rating of the ceramic capacitors I had used! Looks like I will have to build a bigger and better version 2…

Background Theory

You might have noticed that I have mentioned the device I built as a “bipolar Marx generator”. The bipolar part indicates that I have built two Marx generators and connected them in an arrangement that supplies opposite polarity to their first stage inputs. This results in a positive and negative high voltage output relative to ground for each Marx generator, however, between the outputs of the two Marx generators there exists a voltage difference twice that of the magnitude of just one of the Marx generators alone.

Looking at just a couple of stages in single Marx generator, a DC input voltage charges the capacitors through the resistors until the voltage across the capacitor exceeds the breakdown voltage of the spark gap and they discharge resulting in a spark. It can be seen that the reason this functions as a voltage multiplier is that the capacitors charge in parallel through the resistors, but they discharge in series once the spark gap voltage break down is exceeded. This can be cascaded in multiple stages to keep doubling the voltage to a desired level. This is an ideal case where you could theoretically keep doubling indefinitely, however, this is both impractical and not possible due to losses and inefficiencies with the components used. In theory my bipolar Marx generator should be able to produce 300kV with a total of 20 stages and an input voltage of 15kV (20x15kv=300kV). However, it only produced sparks up to about 22cm which implies an output of approximately 220kV.


Another important and interesting point to be aware of is that the voltage across the resistors at each stage within the Marx generator is only equal to (or less than) that of the input of the first stage. Also remember that the capacitors discharge in series. This means that the capacitors and the resistors in each stage only need to have a maximum voltage rating equal to the input of the first stage (i.e. the components in each stage do not need to progressively be rated for higher and higher voltages as you add more stages). This is good news for us as it means we don’t have to buy expensive and rare ultra high voltage components! It is however, never sensible to load components at their absolute maximum ratings especially when dealing with high voltages and pulsed discharging of capacitors – a margin of around 50% should be incorporated (i.e. if your input voltage was 10kV, you should try to use components rated for at least 15kv). Otherwise you do what I did and end up killing your lovely new Marx generator!

Making the device

I decided that a nice round number of 10 stages in each leg of the bipolar Marx generator would be sufficient, giving me a maximum of a x20 multiplier. The bipolar Marx generator arrangement that I have used is shown in the circuit diagram below. Note that this diagram shows only five stages per leg to keep the diagram compact for easy viewing.


The next thing was to source 20 high voltage capacitors. My input voltage was supplied by a variable high voltage DC power supply that I built using a flyback transformer set to about 15kV. I found a cheap set of 20kV rated ceramic capacitors on eBay for about £30. If you try to source quality high voltage polypropylene/polystyrene capacitors with similar voltage ratings you will find that it will cost an awful lot more!

For the resistors you will also find that buying quality 20kV rated resistors will cost you a small fortune, that’s if you can even find any at all. There is a nice trick you can do and that is to buy multiple lower rated resistors (standard 1kV rated resistors, which can handle a fair bit more than their trusty data sheets tell you) and solder them in series. Each resistor will see a relative fraction of the total applied voltage across the bulk resistor string, e.g. 10kV across three resistors in series will subject each resistor to 1/3 of the 10kV. If you use power resistors 5W+, the carbon/metal-film tracks will be thicker and wider spaced with the insulation coating keeping a reasonably high voltage at bay. The distance across each resistor (from leg to leg) also needs to be considered – it needs to be long enough that it doesn’t arc over the whole resistor (a 1mm air gap per kV is a sufficient guideline to the breakdown voltage of air). The resistance value needs to be reasonably high, this is to reduce the bleed/leakage current during charging and discharging of the capacitors which will increase each stages voltage doubling efficiency. The resistance value can’t be too high as this will increase charging times between each full discharge. A value between 500kΩ and 2MΩ is a reasonable guide but not absolutely critical – I settled with two 5W 500kΩ resistors soldered in series for a total of ~1MΩ for each resistor in each stage.


I decided to mount the array of components onto strips of polypropylene sheet which I obtained from a couple of cheap cutting boards from the supermarket. For the spark gaps I wanted some nice looking wires/electrodes. For this I decided to get some 3mm dia. steel ball bearings and solder one to the end of a length of thick solid copper wire. Each wire was then bent the same distance from the ball. I then marked and drilled the necessary holes in the polypropylene strips to allow me to mount the components similar to through hole components with all the connections soldered together on the rear of the strip. Keeping the holes a tight fit ensured the spark gap electrodes and components remained rigid. The air gap between the electrodes was estimated to the equivalent breakdown voltage of 15kV which was about 15mm.

DSC00043 DSC00046

Below is the finished bipolar Marx generator mounted on a polypropylene base.

DSC00727 DSC00726 DSC00728




Black Powder


Black powder or Gunpowder is an explosive pyrotechnic composition comprising of charcoal (C), sulphur (S) and potassium nitrate (KNO3, also known as saltpetre). Black powder is classed as a low explosive due to its relatively slow (sub-sonic) combustion or deflagration rate when compared with that of high explosives (super-sonic combustion rate) such as TNT. It can be effectively used as a propellant used in firearms, fireworks and other pyrotechnic displays.

A simplified general chemical equation of the combustion of black powder is:

10 KNO3 + 3 S + 8 C ⇒ 2 K2CO3 + 3 K2SO42 + 5 N2

Making black powder is more complex than simply mixing the three ingredients together. This is in fact called green powder or polverone which burns slowly and inefficiently leaving lots of un-burnt solids behind. Proper black powder is mixed thoroughly to a fine powder using a ball mill or with other non-sparking grinding media. This burns much more rapidly and leaves very little un-burnt residue.

The standard composition of black powder is:

  • 75 parts potassium nitrate
  • 15 parts charcoal
  • 10 parts sulphur

All parts are by weight, not volume – the constituents of most pyrotechnic compounds stated in this convention unless otherwise stated.

Potassium nitrate can be obtained as a pure (99.9%) ground compound from a chemical supplier or as an agricultural fertiliser (<99% purity). Charcoal can be also be bought from a chemical supplier pre-milled to a fine dust or can be made at home from pure wood charcoal (the best wood to use for the charcoal is Pacific Willow). Sulphur can be obtained from a chemical supplier or as an agricultural fertiliser.

The main tool used to make home made black powder is a ball mill or the more labour intensive mortar and pestle. There are two main methods of making black powder each with their own advantages and disadvantages. Professional or commercial black powder will likely always be superior due to its more sophisticated industrial milling, pressing and corning methods.

The two methods discussed here are the ball mill method and the precipitation method. The ball mill method is far simpler than the precipitation method but produces an inferior powder.

Please note: the information provided here is for information purposes only; the methods discussed herein should never be attempted unless carried out by a person who is fully trained and is totally compliant with all relevant local laws and regulations.

Ball Mill Method

This method is potentially very dangerous if not practised with caution. The milling location should be away from any buildings/people/animals in case of accidental ignition of the powder during milling.

  1. Grind the raw charcoal in a mortar and pestle until small grit sized particles are obtained.
  2. Weigh out 15 parts charcoal and add 10 parts sulphur (powdered). Bearing in mind that you should aim to fill the ball mill to a maximum of about half full when all the substances are added.
  3. Ball mill the charcoal and sulphur mix for about 3 hours (without adding any potassium nitrate).
  4. Add 75 parts of potassium nitrate (powdered) to the charcoal and sulphur mix in the ball mill. Also add 6 parts water to the mix before ball milling, this will dramatically reduce the probability of accidental ignition of the mixture during milling.
  5. Ball mill the mixture for about 5 hours checking every hour or so that the mixture hasn’t dried out; if it has add a little more water and continue ball milling.
  6. After ball milling, empty the mixture out onto a large sheet of paper, spreading it out and allow it to dry naturally in a warm, non-humid place.
  7. When dry, very carefully grind the mixture in a mortar and pestle to break it up into a course powder. This should then be sieved using different sized sieves to obtain different grades of powder for different purposes.

Precipitation Method

The precipitation method is much more difficult. You first ball mill the charcoal and sulphur together (just like you would with the ball mill method), but this is followed by dissolving the potassium nitrate in hot boiling water which is then mixed with the milled charcoal-sulfur mixture. The potassium nitrate is then precipitated from the solution by mixing with ice cold isopropyl alcohol. This is followed by filtering (messy) and drying which takes a long time, and a good place with no ignition sources is required, since there is a flammable liquid involved.

  1. Grind the raw charcoal in a mortar and pestle until small grit sized particles are obtained. This should be done outside as it is very messy.
  2. Weigh out 15 parts charcoal and add 10 parts sulphur (powdered). Bearing in mind that you should aim to fill the ball mill to a maximum of about half full when all the substances are added.
  3. Ball mill the charcoal and sulphur mix for about 8 hours (without adding any potassium nitrate).
  4. While the mill is running; per 100g of charcoal-sulphur mix, place 600ml of isopropylalcohol into a large container and place that in a fridge.
  5. In an old pan, take 75 parts of potassium nitrate and add 40ml of water per 100g of potassium nitrate. Place the pan on the stove and bring it to the boil while continuously stirring. When the solution starts boiling add small amounts of water until all the potassium nitrate has dissolved. Then add another 10ml of water.
  6. Add the charcoal-sulphur mix to the boiling solution of potassium nitrate. Stir until the liquid is consistent and free from lumps. Allow to cool to room temperature.
  7. By now the isopropylalcohol should have cooled to about 0°C. Take the isopropylalcohol outside and pour the charcoal-sulphur-potassium nitrate mixture into the cold isopropylalcohol.
  8. Make sure there are no sources of ignition nearby as there will be a large amount of flammable vapour given off from the isopropylalcohol. Stir for a few seconds.
  9. Place this mixture back into the fridge/freezer and allow to cool back to 0°C as fast as possible.
  10. Now filter the mixture through an old cloth and squeeze the liquid out. Discard the black liquid.
  11. Spread the black mixture out on a large sheet of paper and allow it to dry naturally in a warm, non-humid place, e.g. outside in the sun light. Keep away from any sources of ignition.
  12. When dry, very carefully grind the mixture in a mortar and pestle to break it up into a course powder. This should then be sieved using different sized sieves to obtain different grades of powder for different purposes.

Ball Mill


Originally published: 2006

This project will show you how I built a ball mill for the process of making finely ground powders. Predominantly for the production of black powder and other pyrotechnic compounds. I have shown how I build the ball mill and how I made my own lead ball grinding media.

Ball mills are machines which are used to reduce the granular size of powdered chemicals and to efficiently mix multiple chemicals together. This is an important requirement for good pyrotechnic compounds, the finer the grain of a pyrotechnic compound and how well it has been mixed with other constituents determines how efficiently the compound will combust.

Ball milling pyrotechnic compositions is a very hazardous procedure. Care and the correct materials and procedures must be followed when running the mill. You should never mill any shock sensitive compounds such as flash powder. The grinding media within a ball mill must be non-sparking, common grinding media used are lead balls or ceramic media. Chrome steel or glass grinding media should never be used. Milling should always be performed outside away from buildings and the public in case of accidental ignition of the compound being ground.

Materials used

Two 32cm x 19cm x 1.2cm ply wood boards.
110mm Ø, 250mm long PVC pipe.
110mm Ø, PVC pipe end caps.
Four 10mm Ø, 130mm long threaded bars.
Nuts, bolts and small wood screws.
High torque electric motor.
Four caster wheels.
Two PVC pipe brackets (diameter similar to motor).
25mm Ø, 30mm long aluminium bar (to be lathed for motor belt wheel).
Computer mouse ball or similar diameter ball.
Lead (flashing, tubing or scrap).
Butane gas torch.
Stainless steel ladle.

Construction of the ball mill

The first thing I did was to construct the tumbler barrel. This allowed me to deign the rest of the mill around the size of this barrel. I used a 250mm long length of 110mm Ø PVC pipe as the main part of the barrel. the end caps I made myself with a vacuum forming machine at my school because of the expensive price of commercial end caps.

ballmill2 ballmill3

The next part was to build a platform on which the barrel could roll. For this I used four castor wheels bolted upside down on a sheet of plywood. The gap between the wheels must be about the right distance apart so that the barrel can roll freely without the possibility of rolling off the casters. A large slot was also cut out of the centre of the platform (see picture) which would allow a motor belt to pass through it and around the barrel.

ballmill4 ballmill5 ballmill6

The next part was to mount the motor to the underside of the same board of plywood. The motor shaft has to be placed in the centre of the plywood platform to allow a rubber belt to pass through it. To mount the motor to the plywood platform, I used PVC pipe brackets which fitted the motor almost perfectly. I used some rubber inner tube wrapped around the motor where the brackets would be clamped to add grip and to stop the motor spinning in the brackets when in use. I screwed the brackets down using small brass wood screws and a couple of washers. The mounting turned out to be surprisingly sturdy and ridged.

ballmill7 ballmill8

The motor shaft had to have a wheel/pulley attached to hold the rubber motor belt. I used a short length of aluminium bar for this. I drilled a hold straight through the centre of the bar with the same diameter of the motor shaft. The pulley needed a grub screw to grip and hold the pulley in place on the motor shaft. This was done by drilling a hole perpendicular to the shaft hole on the pulley and then threading and counter sinking the hole. A short section of thin inner tube was stretched over the pulley to add grip.

Now that the platform and motor mount have been finished, a base is needed. This was simply another piece of plywood of the same dimensions of the platform used before. Four holes in the corners of both plywood boards were drilled. Then four threaded bars of appropriate length were bolted to the base board (as shown in photo) the top platform could then be bolted to the threaded bars as well.


The next thing to do before the ball mill was completed was to make a motor belt. I used a length of bike inner tube. The belt needs to be under a fair amount of tension when in operation to take this into consideration when performing the next step. Determine the length needed to pass around the barrel and motor pulley by simply wrapping it round by hand and marking it, making sure there is tension in the inner tube band. Now using rubber solution found in nearly all puncture repair kits, adhere one end of the tube into the other following the instructions for using the rubber solution carefully.


Due to the high costs of pre-made lead balls for grinding media, I decided to make my own lead balls by casting them. A ball mill should be filled approximately a third to half full with grinding media, and that is equivalent to about 50 to 80 lead balls for my mill. The first thing I did was to mould some wet clay into two rectangular blocks. I then needed a ball pattern which I used a computer mouse ball – one of those old track ball mice. I pressed it into the centre of the clay blocks and formed the clay tightly around the ball. I inserted it slightly deeper than half way and then removed the ball. The reason for making the depressions deeper than half way is that to make clean flat surfaces I would later sand the faces flat which would require some extra depth to ensure the final void was spherical. I used a thin wooden dowel to make depressions for an inlet channel and air an escape channel, making sure each block was a mirror image of the other.

I then baked the clay in an oven at a low starting temperature of about 80°C for about an hour to slowly evaporate the water out of the clay. This avoids cracking the clay due to heating it too quickly. After being dried I turned up the oven to full power at about 220°C for about 2 hours (this isn’t actually hot enough to properly cure the clay into pottery but it does set it hard enough for use as a casting mould).

After the moulds had cooled, I sanded the faces of each block flat and also sanding back until the depression became as close to hemispherical as I could. I glued a sheet of sandpaper to a flat sheet of MDF board to provide a flat sanding surface. The ball used as the pattern was placed in the depressions to align the two mould halves and the outside edges of the moulds were scribed to enable me to realign them without the ball pattern.

To cast a lead ball I set up the two mould halves and gently clamp them together using a clamp and some packing board. I placed a few scraps of lead in a cooking ladle and using a butane torch melted the lead in the ladle until it was completely molten (this is clear when the molten lead rolls around freely with high surface tension and does not wet the ladle surface). The molten lead was carefully poured into the mould through the inlet channel until the mould was filled. The air outlet channel should allow the air/gases in the mould to escape easily to avoid bubbles/void defects in the ball casting.After waiting a few minuets for the lead to solidify the mould was parted gently to reveal the casted lead ball. The excess lead from the inlet and vent ports can be easily cut off and filed smooth.

ballmill11 ballmill12 ballmill13 ballmill14

30kV Flyback PSU


This high voltage PSU utilises a high voltage flyback transformer (also known as a Line Output Transformer – LOPT) salvaged from an old black and white TV. It is capable of producing a variable DC output of up to approximately 30kV. This circuit uses a relatively simple transistor based oscillator circuit to drive the flyback transformer – there are far more complex driver circuits out there but this circuit can be built simply with minimal knowledge and parts, most of which can be salvaged from an old CRT TV. This PSU can be used to power many other small but interesting high voltage projects such as plasma globes (high voltage AC would work better), Jacobs ladders, Marx’s generators, and Ionocrafts.

Background Theory


Very high voltages of up to 40kV can be generated using the standard flyback transformers discussed here. 40kV is more than enough to ionise a relatively large air gap! Air has a breakdown voltage of approximately 30kV per centimetre. The nominal output current is relatively low <10mA, but can still be dangerous.

The output voltage of a flyback transformer has a high frequency, usually around 20kHz. Our nerves are not very sensitive to high frequency currents and therefore little or no sensation of shock is felt at high frequency, even if a significant current at high voltage flows through your body, contrary to expectations. You will however feel flesh burns due to high current ohmic heating and/or the extremely high temperature of plasma arcs, the former being less important in this example due to relatively low output currents for flyback transformers.

Plasma arcs have temperatures in the thousands of degrees and can easily set flammable materials on fire and cause flesh burns.

Flyback Transformers

Flyback transformers (FBT) or Line Output Transformers (LOPT) are special step-up transformers designed to generate high voltages at high frequencies. They were invented for use in cathode ray tubes (CRT), commonly found in CRT TV’s and PC monitors.

FBT’s come in all shapes and sizes and can usually generate voltages between 20kV and 50kV and operating at frequencies between 15kHz and 50+kHz depending on the FBT.

Standard operation of a FBT uses a low voltage saw-tooth input at high frequency applied to the primary windings of the FBT. The fast switching of the input current and large step-up turn ratio between the primary windings and the secondary windings induce a very large voltage in the secondary winding.

There are three main FBTThere are three main FBT variants which can be obtained by salvaging them from various electrical appliances.

1. Old Black & White TV FBT:

This is regarded as the most ‘useful’ FBT. This is mainly due to the transformer consisting solely of primary windings and a secondary winding with no high voltage rectification or high voltage smoothing capacitor being incorporated into it as a unit. They have an exposed ferrite core which allows easy access and space to wind you own sets of primary and feedback windings. Not having the high voltage rectification allows for a full-wave AC output and not having the high voltage smoothing capacitor allows one to pull longer continuous arcs rather than high frequency arc pulses. For a high voltage AC current the correct driver circuit must be used to produce an AC current on the FBT secondary; the simple driver circuit I have used described on this page will not allow for a high voltage AC current output. The above advantages allows one more options regarding experimentation with high voltage experiments such as plasma globes, Jacobs ladders etc, which all require or perform better with high voltage AC. If a DC supply is required for lifters or ion motors for example, the AC output can be rectified and smoothed using external components. The old B&W FBT’s generally output a lower voltage (15kV to 25kV) but with higher current capabilities than the modern colour TV or PC monitor FBT’s (20kV to 40kV).

2. Colour TV FBT:

These are the next best thing if an old B&W FBT is not available to you. These usually incorporate high voltage diode rectification across multiple secondary output stages producing a full wave rectified output but with no smoothing capacitor which allows one to pull longer continuous arcs. These are known as diode-split FBT’s. These FBT’s can be ‘pushed’ further than the old B&W TV FBT’s and can output up to 60kV with the correct driver circuit. These FBT’s are great if you want to experiment with lifters (ionocrafts) and ion motors that require a DC output.

There are old colour TV FBT’s which are known as cascade (or voltage multiplier) FBT’s. They output around 8kV AC at the secondary winding and use a voltage multiplier (tripler) to step the voltage up to ~25kV DC.

These FBT’s are generally slimmer than the monitor FBT’s as they don’t have the built in smoothing capacitor. The units consist of a horseshoe shaped set of pins encased in a black/grey plastic case with a thick red high voltage wire coming out of the top.

3. PC monitor FBT:

These are the least ‘useful’ FBT’s. They are very similar in design to the colour TV FBT’s except that they include a built-in high voltage smoothing capacitor making drawing arcs difficult and therefore not useful for experiments such as Jacob’s ladders.

These FBT’s are generally fatter than the colour TV FBT’s as they include a built-in smoothing capacitor which can be identified from the outer plastic casing as a long cylindrical extrusion on the side of the unit near to the thick red high voltage out put wire. The units consist of a horseshoe shaped set of pins encased in a black plastic case with a thick red high voltage wire coming out of the top very similar to the colour TV FBT’s.


Driver Circuit

There are many different driver circuits out there that can be used with a FBT. There are some very simple and some very complex designs which allow you to control driving frequency, current and voltage supply etc – see here for some good driver circuit variations. I decided to use one of the simplest designs based on a resonating transistor based circuit. The circuit schematic is shown below.


This circuit operates as follows:

When a DC voltage is applied at “+” and “–”, the transistor will conduct current through the FBT primary via R1 and R2. This induces a current in the secondary windings and the feedback winding. The feedback coil will generate a current, which will trigger the transistor to stop conducting. The magnetic field in the ferrite core collapses and a HV spike will appear on the secondary windings by induction. There will no longer be any feedback current on the feedback windings and therefore the transistor is allowed to conduct again. This cycle repeats several thousand times a second, and the cycle will repeat at the natural frequency of the FBT.

By having a circuit using such a feedback system, adjustments are minimal, and the circuit becomes dynamic i.e. the operating frequency automatically settles at the natural frequency of the FBT. This can be observed when arcs are drawn; the frequency of the arc increases as the arc is drawn longer.

Depending on the ratings of the transistor used, this circuit can supply a relatively large current; but the transistor will need a large heat sink.

Flyback Transformer windings

To use a FBT one needs to either find or wind their own primary and feedback (and sometimes even secondary) windings.

Custom primary and feedback windings:

Winding your own primary and feedback windings allows you to supply a much larger input current than can be supplied to the commercial FBT’s that use very fine gage wire which can be damaged by Ohmic heating. Secondly you will have full control over the turn ratio between the primary and secondary windings and between the primary and feedback windings. Usually winding a new primary and feedback winding on the ferrite core and utilising the high turn secondary of a commercial FBT is adequate for good output results. The usual number of turns for a custom primary and feedback is between four and six turns for the primary and between two and four. However depending on the FBT and transistor used, these numbers need to be experimented with to get optimal performance. When winding and testing the primary and feedback windings you should be aware of the direction of each winding – if you can’t get an output try switching the connections across either the feedback or primary windings.

I could never quite get as good results by winding my own primary and feedback windings (probably needed to experiment more), and so, I stuck with the primary windings already present on the FBT.

Finding the primary and feedback windings on a commercial FBT:

Finding the appropriate pairs of pins to serve as your primary and feedback windings can be a somewhat tricky task. I have played with various modern commercial FBT’s and they usually have a very similar pin layout and pin designation. This site has a good guide that can be very useful for finding appropriate pairs of pins for the windings using a standard DVM.

The first thing to know is that in general the ‘useful’ pins are located on the horse shoe shaped array of pins (primary, feedback pairs and HV GND pin). The pins aligned in a straight line at the edge are not usually useful and can be ignored (they are usually connected to focus and screen adjustments).

When finding/selecting pin-out pairs your first need to work out which pins have continuity with other pins using a DVM on continuity mode. After this taking resistance measurements between pairs of pins showing continuity will give an indication of the relative number of windings (not necessarily correct as the internals of FBT’s are more complex than simply tapped windings). The feedback winding should have less winding than the primary and the primary should have as fewer windings as possible to allow a higher output voltage. Generally the lowest resistance between pins is between adjacent pairs.


I found the primary and feedback pin-out pairs by experimenting on a trial and error basis, whilst using the knowledge from the sites mentioned above and from various other site that I have found that document FBT pin-outs and diagnostics (other flybacks will likely have a slightly different pinout).

I have noticed that the pin-out pairs which work best on one particular FBT, commonly occur in similar positions on various other FBT’s (be it a type 2 or type 3 FBT variant, mentioned in the Flyback Transformers section).

The standard layout and pin-out pairs that I have found to work most effectively and are generally commmon between FBT variants is shown in the above diagram. This diagram could be wrong for some FBT’s but it should provide as a first basis starting point to test a new FBT. A bit of experimentation with various pairs of pins and swapping the polarity of the connections to each pin-out pair will ultimately provide one with the most appropriate pin-out pairs for a particular FBT.

Here are a couple of internal schematic diagrams of some commercial FBT’s to help with understanding the pin-out design.

flybackpsu_6 flybackpsu_5

Here are some more helpful links worth reading:

FBT pin-out:

CRT diagnostics:

Making of my variable 30kV PSU

Circuit Design

I decided that it would be convenient to power this device by the mains as this would eliminate the need for an external DC power supply. For this I required a mains transformer and a bridge rectifier with smoothing capacitor. To allow me to vary the DC voltage into the flyback driver circuit I decided to use a voltage regulator to produce a variable DC voltage from 1.25 to 27.95 volts (see this very useful LM317 calculator). The next stage of the circuit incorporates some protection from the driver circuit as I have found high voltage currents tend leak back towards the power supply end of the circuit. For this I have incorporated a transient voltage suppression diode constructed using two standard zener diodes connected as shown in the circuit diagram. I have also included inductors used as chokes to suppress any voltage spikes from the driver circuit. Finally, I have put two in-line diodes to stop any AC currents flowing back up from the driver circuit. The last part of the circuit is the flyback driver circuit as seen previously and is separated form the DC supply via a switch.


Component list:

V1 Mains power, 240VAC RMS, 50HZ.
S1 Rocker switch, 250VAC, 6A.
Tr1 Mains Transformer, 240VAC primary, 12-0-12V secondary centre tapped, 250mA.
D1 Bridge rectifier (KBPC104), 280VAC RMS, 1A.
C1 Polarised electrolytic capacitor, 330uF, 50V.
Reg1 Voltage regulator (LM317T).
R1 Carbon film resistor, 220Ω 5%, 0.25W.
POT1 Linear carbon track potentiometer, 4.7kΩ.
C2 Polarised electrolytic capacitor, 1uF, 50V.
D2, D3 Zener diode (BZV85C 36V), 36V 5%, 1.3W.
L1, L2 Ferrite core inductor, 20mH.
D4, D5 Rectifier diode (1N4005S), 600V surge and peek reverse voltage, 1A max forward current.
S2 Rocker switch, 250VAC, 6A.
R2 Ceramic power resistor, 235Ω, 14W (parallel pair of 470Ω, 7W).
R3 Ceramic power resistor, 27Ω, 3W.
Tr1 NPN Power transistor (C5387).
TR2 Flyback transformer.

My Flyback Transformer

I managed to salvage an old styled FBT from an old CRT TV. IT was relatively easy to find primary and feedback pin-out pairs. The two wires/pins on the plastic base seen in the front of the photo below are the primary input pins and on the other side were six other pins relating to tapped output of the secondary winding. The HV ground pin can be seen on the far right of the plastic base in the photo. I did have a go at winding my own primary and feed back windings however I couldn’t get the FBT to produce a reasonable output. I therefore resorted to using the existing windings and pin-outs.


Designing and Soldering the Circuit

I like to use MS excel to layout my circuit components and connections when I’m using strip board. I set up a matrix of small square cells representing individual holes on a sheet of strip board. I mark soldered pins as pink cells and red for track breaks. I also colour in yellow blocks of cells approximating to each of the component sizes and shapes to help with compacting and arranging the layout of all the components.

I then started to solder all the components on a sheet of strip board cut to the required size. The voltage regulator required a heat sink and the power transistor required a very large heat sink. These were attached to the components using self tapping mounting screws and thermal paste. The heat sinks were salvaged from an old TV and had mounting pins which could be soldered to the strip board. The voltage regulator potentiometer and switch S2 were mounted in the lid of the project box with wires connecting directly to the circuit board. The mains power was connected to a fused and switched IEC euro connector module that was mounted in the side of the lid of the project box. Wire to the the mains transformer was soldered directly to the module’s live and neutral terminals.

flybackpsu_10 flybackpsu_9

The FBT was mounted on a sheet of rigid polystyrene (heat moulded with a 90° bend in it) using some rather convenient threaded mounting holes on the FBT base. Connections to the FBT primary and feedback windings were soldered directly to the FBT pins and connected to screw clamped terminal posts on the circuit board (the terminal posts allow the connections to be removed easily if ever needed). The HV (hot) output wire was extended (as it had been cut short) with some spare HV wire from other FBT’s that I had salvaged. Some more spare HV wire was soldered to the HV GND pin of the FBT. These HV wires were connected to two terminal posts mounted on the lid of the project box.

Some pictures of the finished PSU:

flybackpsu_12 flybackpsu_11

Coilgun I


Originally published: 2007

This page documents the build of my first coil gun. It is a 704 joule coilgun, capable of firing a small projectiles at high speeds. A coil gun is a type of a linear electromagnetic accelerator. It fires ferromagnetic projectiles using electromagnetic coils or solenoids.

I have tried to include some supplementary information regarding the basic theory behind a simple single stage coil gun and how to go about building one. It involves basic knowledge of electromagnetic physics and high current/voltage electronics.
Coil gun basics

A coil gun is a circuit comprising of three main components – a powerful capacitor bank, a high current switch and a coil of wire (or solenoid) from which the projectile is lunched. The capacitor is charged and then discharged through the coil by the switch.

A ferromagnetic projectile is usually used but any ferrous material could be used. A ferromagnetic material is one that interacts with a magnetic field but cannot be magnetised. This projectile is placed within the coil at specific distance from the centre. The capacitor bank is usually comprised of a many smaller capacitors combined in a parallel and series combination to obtain the voltage and capacitance values desired for the bank. The capacitor bank is charged with a DC current via a charging circuit. The capacitor bank is then discharged through the coil via a high current switch usually a semiconductor switch in the form of an SCR (Silicon Controlled Rectifier) or thyristor.

A pulse generator circuit is often needed to activate the SCR by sending a low voltage signal to the SCR with a particular pulse width which will then allow a pulse of current to flow from the capacitor bank through the SCR and coil. This is necessary because if the pulse of current through the coil is too long the projectile will oscillate at the centre of the coil. This is because the projectile is being attracted to the centre of the coil in the presence of the magnetic field produced by the high rate of change of current flowing through the coil. The magnetic field ideally needs to be turned off after a particular time to stop the projectile being attracted back to the centre of the coil.

This however causes another problem known as inductive kick-back. Once the current is switched off, the coil has a strong magnetic field present. Coils/solenoids are inductors which like to oppose a change in current and in this case wants to keep current flowing but cannot as the circuit is broken via the switch. The solenoid’s magnetic field collapses into a high voltage pulse which can easily damage other components of the circuit. The solution is to install a power diode and resistor bleeder circuit in parallel with the solenoid to bleed this inductive pulse away.

The circuit diagram below is a simplified circuit diagram and shows the layout of components described above with some example component values.


At some point during the design of a coil gun you will have to play with some equations. When looking at capacitors these are the three main equations you will need.

eq1 eq2 eq3

The capacitor bank energy is proportional to the square of the voltage so an effective way to increase the bank energy is to have a relatively large bank voltage. However increasing the capacitance will ultimately dictate the maximum surge current and its duration when discharged.

The solenoid design relies on a compromise between different variables. It needs to have a low inductance to reduce Ohmic inefficiencies but also needs many turns to increase the magnetic flux and hence magnetic force induced onto the projectile. Looking at the force and inductance equations below it is possible to see how the different variables affect the total force produced. In general increasing the force will also result in increasing the inductance – a compromise of the two has to be made this can be done experimentally or with the aid of some simulations.

eq5 eq4

Force is in Newtons, Inductance is in Henries, A is the cross-sectional area of the coil in square meters, μ0 is the permeability of free space, μ is the permeability of the coil core (in this case it is air: 1), N is the number of turns of wire, I is the current in Amperes, l is the length of the coil in meters.

A really useful tool is Barry’s RLC circuit simulation (Java required). This tool enables you to enter your coil inductance, coil resistance, capacitor bank capacitance, and capacitor bank voltage to simulate the resonance of the circuit and its damping. It is useful when finding out what pulse is needed when operating the thyristor – this time period should be the time it takes for the current to hit it’s first maximum peak.

Building the coil gun

The first thing I did was look at what things restricted my design. Capacitor values are somewhat fixed in terms of voltage and capacitance. Everything else in the circuit could be built around the capacitor bank such as SCR ratings and power supply. I found a listing for a box of 16 new old stock capacitors with a rating of 200VDC and 2200µF which seemed like a good set for a first capacitor bank. I decided to arranged two sets of eight capacitors in parallel, with each set connected in series giving me a combined bank rating of 400VDC at 8800µF (704J).

I laid out the capacitors into two sets of eight capacitors in parallel, with each set connected in series. I stripped some lengths of single core copper mains wire and straightened them out. I then soldered each capacitor terminal to a length of this wire until they were all connected correctly.


I made a switching terminal where the charging switch and firing switch would be located. I designed a net and cut it out on a sheet of ridged polystyrene sheet. I then used a strip heater to bend the net at the marked fold lines to form the terminal box. I then cut the holes where the switches were to be inserted. I also screwed a set of chop block connectors to the top of the terminal where the capacitor bank was connected to the charging circuitry.

I decided that I would separate the charging of each side of the capacitor bank so that I could first charge one set of eight capacitors at 200V and then the other side. This was because I preferred the idea of using a lower rated power supply (200VDC which is what I had available) to charge the capacitor bank.

coilgun3 coilgun4

For this I needed a four gang three pole switch. I wired the switch up so that a DC line input could be switched off, charge the left bank or charge the right bank. Basically, in position-one the switch would not charge the bank at all, position-two it would charge the left bank of capacitors and in the third-position it would charge the right bank.

This switch comprises of four inner terminals and twelve outer terminals. Each inner terminal is associated with three consecutive outer terminals. The switch has three positions, each switch position connects all inner terminals to one of their three associated outer terminals. The inner terminals are labelled A, B, C and D. The outer terminals are labelled 1 to 12. So terminal A is associated with outer terminals 1,2 and 3 and the continuity between A and one of the other three terminals depends on the position of the switch.

coilgun5 coilgun6

There are two chop block terminals screwed to the top of the switching terminal. The DC input block has two inputs for positive and negative from the charging PSU. The other block terminal has three inputs which is wired from the four gang three pole switch (described above) such that the DC input charges one set of capacitors and then either the two sets of capacitors. The other connections on the larger block terminal are connected to the capacitor bank with a mid-point tap. The bank also has wires connected here which lead to the thyristor and coil.


I needed to start mounting the components onto a solid base, so I cut a base board from plywood (which was later coated with polyurethane varnish). The capacitor bank and switching terminal were connected together and placed on the board and screwed into place. I used four small wood screws to screw the switching terminal to the base. For the capacitor bank I cut large washers from a sheet of ridged polystyrene and screwed long screws directly to the base board clamping the capacitors down with the large washers.

The next thing was to mount my thyristor which was another great find that only cost £35. It is defiantly over rated for my coil gun but I decided I could always use it for bigger experiments later on. It is a Westcode P300KH08EJ0 rated at 300A (average current), 550A (RMS current), 10,450A (surge current), 800V (average voltage), 25µS (turn-off time), 300mA (gate current). There were holes in the corners of the base of the thyristor so I simply screwed it down to the base board with wood screws and washers. The base of the thyristor is also the anode and the positive wire from the capacitor bank is connected to the base here as well. The cathode of the thyristor (the very thick red wire coming out of the top of the component) was bolted to an L bracket which was also screwed to the base board.


The all important coil was now needed. The coil needs to be wound on a tube called a coil form. I have found many different coil forms on the internet ranging from glass, brass and plastic. Metallic coil forms are not advised as eddie currents are created when in operation that oppose the magnetic field that created them, reducing the coil’s efficiency. Glass has a tendency to shatter under coil compression when fired so I decided against this too. I believe that plastic coil forms are the easiest to work with as they can be easily worked and obtained.

My coil form has an internal diameter of 10mm and is made from clear polystyrene (acrylic could be used also). The tube had grooves at each end that allowed me to include end guides that were large external cir clips that coincidently fitted perfectly in the grooves on the tube. I have read that metal end guides also improve the magnetic field strength of the coil in operation but haven’t verified this. I fixed them in place using epoxy resin.


I have used single core copper mains wire with a PVC coating for the windings. I think using some enamelled copper wire with extra thick coating would also be viable and also improve the winding turn density but I couldn’t source it easily. It is important to work out the length of wire needed before cutting a certain length and before winding. My coil is made up of four layers, each layer consisting of about 30 turns. The windings were secured by tightly wrapping the outer windings with insulation tape.


I cut two ‘V’ blocks from some wood and screwed them to the base board. This allowed me to mount the solenoid but to also allow other different sized coils to be mounted on the same ‘V’ block mounting that may be used in the future. The picture only shows the temporary fixing of the solenoid using insulation tape, but I have since secured the coil more permanently with clamping straps and stops. It is important to fix the coil down securely before firing as there is a surprisingly strong recoil when the coil gun is fired which can throw the coil in the opposite direction to the projectile.


A nearby chop block was screwed to the base board and the solenoid wires were connected to it along with the wires from the thyristor cathode and negative wire from the capacitor bank.

The power diode and resistor bleeder circuit protection was to be built next. I was unable to work out the exact maximum current that this circuit would need to handle so I tried to over compensate – I soldered three P600J rectifier diodes rated at a total of 1200A surge current together in parallel. I also bought a power resistor rated at 2.2 ohms at 26W. This was soldered in series with the diode network. I then screwed the diode to a ridged polystyrene sheet which was also screwed down to the base board. This simple circuit was then wired in parallel with the solenoid via the chop block terminal that was connected to the coil.


The last part to add to the coil gun was the pulse generator circuit which operates the thyristor. This circuit activates the thyristor for a certain amount of time set by this circuit. I have used a variable timer circuit kit based on a 555 timer chip purchased from Maplin Electronics. The circuit had to be customised do what I wanted. I wanted the pulsed signal to go to the thyristor and not operate the relay supplied, so I didn’t use the relay at all, I just added two connections from Co2 and Co1 on the circuit board which go to the thyristor. Co2 goes to the thyristor gate and Co1 goes to the thyristor cathode. In those connections I included a couple of sacrificial diodes to attempt to protect the circuit from surges or revered polarity. The second modification was to do with the minimum output pulse of the circuit. It was originally 0.5 seconds, I needed about a 1.5ms pulse. So to change this I needed to change a capacitor value of the pulse generator circuit. I replaced C3 in the circuit diagram above with a 0.1µF capacitor. This changed the RC characteristics of the circuit and with a bit of testing and RC modeling I could calibrate the variable potentiometers to give me an output pulse of about 1.5ms. The circuit is operated by a 9V battery (PP3) which has a series push button switch added, that is located on the switch terminal (this is the firing button). This circuit board was then screwed to the case board as well. The original circuit diagram of the timing circuit is shown below with red notations notifying the changes that I made.

pulsecircuit coilgun13

For the ferromagnetic projectiles I used lengths of ferrite rod which were obtained from some old radios. Different lengths and profiles will be experimented with later.

Everything was complete so some low powered tests were carried out to test the device before a full power test was successfully undertaken. I will be looking into tuning the circuit and making some tweaks here and there to try and improve the system better in due course.

Tennis Ball Mortar


Originally published: 2004

This is a brilliant launching device – simply due to the sound of the boom when fired! It doesn’t have much of a range when compared to the potato cannons but it does manage to launch a tennis ball about a hundred feet into the air. It also doesn’t have much of a service life as the recoil of firing crumples the bottom-most cans which then have to be replaced in the field, but, it is great fun! The following content is from the original project page.

Materials used

5 baked bean tins with stacking ridges on base of tins and standard top
1 to 3 baked bean tins with a standard top and base (no stacking ridges)
Tennis ball
5mm Ø drill bit and drill
Duct tape
Stanley knife
Tin opener
Extended BBQ Lighter
A few bricks and/or something to hold the mortar in firing position

Construction of the mortar

The first problem was to find some stackable baked bean tins and some that did not stack. The photo below on the left is the base of a ridged based tin (one that stacks) and the photo below on the right is of a standard tin with the base cut out (one that does not stack).

tennisballmortar (2) tennisballmortar (3)

I made baffles/holes in four out of the five ridge based tins by cutting out a circle in the bases using a Stanley knife. The hole in each of the bases is about 25mm in diameter. The baffles help to keep the mortar rigid and can improve the overall combustion of the propellant.

tennisballmortar (4)

I then cut the top and bottom off the standard baked bean tins using a tin opener. These tins are then used for extending the ‘barrel’ part of the mortar. With the remaining ridge based tin I drilled a 5mm Ø hole on its side about 25mm away from the base of the tin.

tennisballmortar (5)

The diagram below shows how the tins should be stacked (the ‘barrel’ part of the mortar can have any number of extra cans added which is a bit of trial and error but does change the performance of the mortar significantly). I joined each tin together simply using a couple of layers of duct tape.

tennisballmortar (6) tennisballmortar (7)

Last of all I modified the tennis ball using duck tape. To keep the ball spherical, I wrapped the duck tape round in an “X” shape first, and then in a “+” shape. I kept adding layers until the ball fitted smoothly yet snugly down the top section of the mortar.

tennisballmortar (8)

To use the mortar stack up some sturdy bricks that can support the mortar well. Place a modified tennis ball down the mortar and spray a very small amount of propellant through the firing hole (or about a teaspoon of lighter fluid). Place the mortar securely and position yourself next to the firing hole and using an extended lighter ignite a flame at the firing hole to ignite the propellant inside – BOOM!

tennisballmortar (1)

Mini Cannon


Originally published: 2003

I had some parts lying around my bench and I had an idea for a mini combustion cannon. I had a plastic tube that fit plastic BB’s perfectly and a nicely sized pot for the combustion chamber that fits the hand nicely. This cannon works brilliantly with standard BB’s or even potatoes or cotton buds. I did lean very quickly to wear goggles as the projectiles do have a tendency to bounce around the room like crazy. The following content is from the original project page.

Materials used

Tablet pot/container (e.g. calcium carbonate or vitamin tablets pot)
Piezoelectric sparker from a lighter
7mm Ø (5mm int. Ø), 220mm long pipe
Two pins/thin needles
Two rings of 22mm Ø pipe
A few lengths of insulated wire
Solvent weld cement or multi-purpose adhesive

Construction of the cannon

To attach a wire to the metal base of the sparker, I made a ‘cup’ to fit to its end (see photo). I made the ‘cup’ out of thin metal and soldered a wire to this as you must not solder to the base/end of the sparker as the heat from the soldering iron causes it to fail and stop sparking!! The wires leading from the sparker should be approximately the same length of the tablet pot plus about 50-100mm.

potatocannon (6)

I then drilled a 7mm Ø hole in the base of the tablet pot to allow the barrel to fit through. I also passed the two wires leading from the igniter through this hole as shown in the photo.

minicannon (2) minicannon (3)

To the end of the barrel that is going to be inside of the tablet pot add a very small amount of multi-purpose adhesive to the inside of the barrel – to narrow the internal diameter to stop the projectile from falling all the way through.

The barrel should intrude about 80% of the tablet pot to keep the barrel length long but not have all of it protruding outside to keep the cannon’s overall size down. I secured the barrel to the inside of the tablet pot with some multi-purpose adhesive.

To make the ignition mounting I used a couple of spilt rings of piping cut from a length of 22mm Ø pipe, and fitted one around the other until it fitted snugly in the tablet pot.

I then pushed two pins through the side of this mounting and cut the protruding excess parts of the pins (to make it flush when inserted into tablet pot) to create a ‘sparking bridge’. I then soldered the two ends of the wires leading from the sparker to each pin (see photos).

minicannon (4) minicannon (5)

I then taped the sparker to the side of the tablet pot near where the barrel protrudes from the combustion chamber and also taped a ‘stop’ made from layers of cardboard behind the sparker to stop the sparker from moving when it was pressed. After the sparker was fairly secure I wrapped a final layer of duck tape around it, just to hold it permanently in place.

minicannon (6)

And that’s it – a pocket mini combustion cannon that you can take anywhere!

Pneumatic Potato Cannon


Originally published: 2002

After building my first combustion cannon I wanted to try a cannon that was slightly different. The pneumatic cannon works on the principal of using compressed air in a sealed chamber to be suddenly released to propel a potato rather than using the combustion of a propellant. This cannon was again build using basic heavy duty plumbing materials and operated up to a pressure of 100psi. The following content is from the original project page.

Materials used

All piping/plumbing components are made from pressure rated ABS – 120psi.

Two 50mm Ø, 90°, bend
50mm Ø, 1200mm long pipe
40mm Ø, 1200mm long pipe
Two 50mm Ø, double socket
50mm Ø, access plug
50mm Ø, 60mm long pipe
Two 50mm Ø to 22mm Ø tundish
Two 22mm Ø, 80mm long pipe
50mm Ø to 40mm Ø reducer
50mm Ø, 190mm long pipe
50mm Ø, 220mm long pipe
22mm Ø, levered ball valve
A car tire valve
50mm Ø, 10mm long pipe (or ring)
50mm x 50mm sheet of 5mm squared wire mesh
Solvent weld cement

Construction of the cannon

I solvent welded the two 50mm Ø, 90°, bends; the 50mm Ø, 1200mm long pipe (chamber pipe); and a tundish together as shown. I then solvent welded a double socket and access plug to the other end of the chamber pipe, also shown in photo. To add the car tire valve to the access plug I simply drilled a tight fitting hole in the centre of the access plug ‘cap’ and inserted the valve.

pneumaticpc (2) pneumaticpc (3)

I then made the valve fixings. Using a 50mm Ø, 60mm long pipe and the ball valve fittings, as shown in the first picture, I solvent welded the pipe into the tundish and screwed the parts together attaching the ball valve (making sure that the ball valve was in the correct position for firing) using pluming wrench to make sure they were tight. I did this for other tundish to complete the valve section.

pneumaticpc (4) pneumaticpc (5) pneumaticpc (6)

To stop potatoes from getting stuck in the tundish/reducer section, I used some wire mesh and a ring cut from an ABS pipe. I solvent welded the mesh to the inner face of the reducer and then welded the ring inside to hold the mesh securely in place. See photos. The barrel can now be solvent welded into the tundish.

pneumaticpc (7) pneumaticpc (8)

To make the cannon structurally strong, I made two separator supports to fit between the barrel and the chamber. The support nearest the valve is positioned between the chamber and the tundish attached to the barrel. I did this by cutting the pattern shown in the photo to a short length of pipe, which fit flush with the walls of the tundish/chamber. This was then solvent welded into position. The second support fits flush with the end of the chamber and slots over the barrel (the barrel passes through a hole in the support at 90°). This is then also welded into place.

pneumaticpc (9) pneumaticpc (10) pneumaticpc (11) pneumaticpc (12)

Here is the cannon completed – I do like the double-backed design of this cannon!

pneumaticpc (13)


Potato Cannon II


Originally published: 2004

This was my second combustion potato cannon. It was a significant improvement in efficiency and size over the first cannon. The following content is from the original project page.

This is my latest combustion cannon, and I have tried to make it as efficient as possible with a 0.8:1 barrel to chamber ratio. It measures 3.65 meters long and weighs quite a lot due to its 6mm thick walls. The power behind this new cannon is a brilliant feeling! It quite happily fires potatoes well out of sight, I just need to remember to remove the length of dowel used to load the potato before firing…it’s not supposed to be a javelin cannon!

The one disadvantage is its size and weight which make it particularly difficult to transport places – I usually get a friend to help and we usually get to deal with the public’s bemused looks on our way to the firing range.

Materials used

All piping/plumbing components are made from PVC.

110mm Ø, 80mm long pipe, 3.2mm thick
110mm Ø, 65mm long pipe, 3.2mm thick
Two 50mm Ø, 3000mm long pipe, 2mm thick
110mm Ø, access plug
Two 110mm Ø, double socket
110mm Ø, to 50mm Ø, reducer
110mm Ø, strap boss
50mm Ø, to 32mm Ø, reducer
32mm Ø, access plug
32mm Ø, double socket
32mm Ø, long tail bend
32mm Ø, 40mm long pipe
50mm Ø, 20mm long pipe
Two 300mm lengths of 2mm thick insulated copper wire
Piezoelectric sparker from a lighter
Sheet of PVC 80mm x 80mm x 2mm
Two: 25mm lengths of 2mm Ø thick insulated copper wire (mains wire)
Solvent weld cement

Construction of the cannon

To make the combustion chamber section, I took the 110mm Ø, 65mm long pipe and marked out a straight line on which I would cut along. This was to produce a sleeve to fit on the outside of the 110mm Ø, 80mm long pipe as the chamber needed to be thicker to withstand the large pressure build-up from the combustions when firing.

Getting the sleeve onto the other pipe was quite tricky – I used pieces of wood to pry the sleeve open. I then applied a small amount of solvent weld to the outside of the other tube and passed the inner pipe through it. I then removed the wood to close the sleeve over the inner pipe, making sure it was in the centre.

potatocannonII (2) potatocannonII (3)

I then cut rings from a length of 110mm Ø pipe; and cut a plate to fit in the 110mm Ø reducer (see photo). This is simply to add strength to the reducer as it needs to be thicker and stronger around the ‘corners’ of the join in the reducer.

potatocannonII (4) potatocannonII (5)

I solvent welded two 110mm Ø double sockets to each end of the combustion chamber section. I then solvent welded the 110mm Ø access plug and 110mm Ø reducer into these double sockets.

potatocannonII (6)

Using a hole saw, I cut a 50mm hole in the front, top of the combustion chamber to allow the igniter to enter the chamber and allow the strap boss to fit snugly when solvent welded to the side (see photo). The position of this whole was to position the handle as close to the centre pivot of the cannon as to compensate for the weight of the 3 meter barrel! This will then make holding the cannon easier.

potatocannonII (7)

This next bit was another tricky bit. The strap boss now doesn’t fit perfectly around the chamber as I have added the sleeve and increased its diameter. I applied solvent weld all round the boss and whole on the chamber, and QUICKLY held it in position with a couple of g-cramps to hold the strap boss flush with the side of the chamber – it was very important that this was a good, strong fixing! I also used a bent screw to fasten the strap boss shut shown in the photo.

potatocannonII (8)

To make the handle I solvent welded the 50mm Ø to 32mm Ø, reducer; the 32mm Ø, 40mm long pipe (female to female connector); the 32mm Ø, long tail bend and the 32mm Ø, double socket (male to male connector).

potatocannon (3)

I cut a segment about 20mm wide in the 50mm Ø, 20mm long pipe as shown in the top right of the photo. This needs to fit inside the reducer and leave enough room to let the two copper wires to fit snugly between the gap. With the sheet of ABS I made a blanking plate by cutting out a 50mm Ø circle. I cut a notch out the side to allow the copper wires to pass through snugly as well (see photo on right). The blanking plate is used to stop the combusting fuel pressure from blowing up in to the handle section when the cannon is fired.

potatocannon (4)

The copper wires were passed through the handle and the blanking plate and ring were then solvent welded into place. The gap between the end of the wires is important for successful ignition of the propellant; set the gap between the two wires to about 5mm and change accordingly when cannon is complete.

potatocannon (5)

This next bit was probably the trickiest. To attach a wire to the metal base of the sparker, I made a ‘cup’ to fit around the end of it, out of thin metal and soldered a wire to this. You must not solder to the base/end of the sparker as this causes it to fail and stop sparking!!

Using the rest of the ABS sheet I cut and made a casing around the sparker. I used a short section of oval plastic tubing and cut square sections of ABS for sides. It was a bit of a ‘make it up as you go along’ method, but it seemed to work just fine!

potatocannon (6) potatocannon (7) potatocannon (8)

I then cut a square hole in the access plug to allow the sparker to fit through it; as shown. I solvent welded the sparker holder to the access plug and held it in place for about 5 minutes to allow the weld to set.

potatocannon (9) potatocannon (10)

I soldered the wires from the sparker to the two copper wires in the handle, and then separately wrapped the exposed parts of the wire with insulting tape.

potatocannon (11)

This handle was solvent welded onto the chamber of the cannon, making sure that the ignition wires where positioned correctly within the chamber.

potatocannonII (9)

This is yet another tricky bit, which involves a lot of brute force and a wall! I cut a straight line down one of the 3 meter 50mmØ pipes using a steady bench, clamp and a jig saw to make a sleeve for the barrel. Then I inserted the end of the uncut 3 meter pipe into the sleeve. Using the wall I ‘carefully’ banged the pipe into the sleeve (making sure I didn’t fracture the piping). the sleeve should be short of being flush with the inner pipe to allow the barrel to fit into the reducer (see photo). I also cut the other of the barrel so it was flush. The barrel was then solvent welded into the reducer, with the help of a friend.

potatocannonII (10) potatocannonII (16)

After leaving the cannon’s joints to fully set it was ready to go!

potatocannonII (1)

Potato Cannon I


Originally published: 2001

This was my first potato cannon. It used basic solvent welded plastic pipe fittings from the local hardware store and I didn’t really consider the combustion chamber to barrel ratio in its design. It only took about an hour or so to put the thing together and I was well on my way to getting hooked on building cannons. The following content is from the original project page.

Materials used

All piping/plumbing components are made of ABS plastic.

50mm Ø, 90°, sweeped tee connector
50mm Ø, access plug
32mm Ø, access plug
32mm Ø, double socket
32mm Ø, long tail bend
32mm Ø, 40mm long pipe
50mm Ø, 20mm long pipe
50mm Ø, to 32mm Ø, reducer
50mm Ø, 750mm long pipe
Piezoelectric sparker from a lighter
Sheet of ABS 80mm x 80mm x 2mm
Two: 25mm lengths of 2mm Ø thick insulated copper wire (mains wire)
Solvent weld cement

Construction of the cannon

I solvent welded the 50mm Ø, 750mm long pipe; the 50mm Ø, 90°, sweeped tee connector and the 50mm Ø, access plug together to construct the barrel and combustion chamber.

potatocannon (2)

To make the handle I solvent welded the 50mm Ø to 32mm Ø, reducer; the 32mm Ø, 40mm long pipe (female to female connector); the 32mm Ø, long tail bend and the 32mm Ø, double socket (male to male connector).

potatocannon (3)

I cut a segment about 20mm wide in the 50mm Ø, 20mm long pipe as shown in the top right of the photo. This needs to fit inside the reducer and leave enough room to let the two copper wires to fit snugly between the gap. With the sheet of ABS I made a blanking plate by cutting out a 50mm Ø circle. I cut a notch out the side to allow the copper wires to pass through snugly as well (see photo on right). The blanking plate is used to stop the combusting fuel pressure from blowing up in to the handle section when the cannon is fired.

potatocannon (4)

The copper wires were passed through the handle and the blanking plate and ring were then solvent welded into place. The gap between the end of the wires is important for successful ignition of the propellant; set the gap between the two wires to about 5mm and change accordingly when cannon is complete.

potatocannon (5)

This next bit was probably the trickiest. To attach a wire to the metal base of the sparker, I made a ‘cup’ to fit around the end of it, out of thin metal and soldered a wire to this. You must not solder to the base/end of the sparker as this causes it to fail and stop sparking!!

Using the rest of the ABS sheet I cut and made a casing around the sparker. I used a short section of oval plastic tubing and cut square sections of ABS for sides. It was a bit of a ‘make it up as you go along’ method, but it seemed to work just fine!

potatocannon (6) potatocannon (7) potatocannon (8)

I then cut a square hole in the access plug to allow the sparker to fit through it; as shown. I solvent welded the sparker holder to the access plug and held it in place for about 5 minutes to allow the weld to set.

 potatocannon (9) potatocannon (10)

I soldered the wires from the sparker to the two copper wires in the handle, and then separately wrapped the exposed parts of the wire with insulting tape.

potatocannon (11)

And that’s it! Here are some more pictures.

potatocannon (12) potatocannon (13)