Fred's Fusor

Engineering Works

Contents of this page

    • C

construction helmet


The purpose of this page of our Fusor website is to give a detailed description of all workshop construction work regarding components of the Fusor. This includes constructions required for peripheral equipment (monitoring equipment) and for stands and supports.

Individual pages of the website, relating to specific subjects (vacuum, HV-power supply, etc.), may give ideas for components to be constructed, though in theory only and accompanied with suggestions where to obtain parts or where to find the original sources for these ideas.
However, here on this page, the actual construction work will be dealt with.

General Techniques

Finishing Stainless Steel


The following section describes chemical processes, which can be very dangerous to your health!
The author of this website is a qualified professional (radio)chemist, who has the training and the experience to handle dangerous chemical substances. Unless you have the same qualifications DO NOT attempt to perform the electrochemical processes below.
When you think that you should apply the described processes, you do that entirely at your own risk. At least apply minimally the following recommendations:
  • When handling strong acids or alkalines always wear safety goggles that fully protect your eyes;
  • Also wear plastic or rubber gloves and a plastic or rubber apron;
  • Always add strong acids or alkalines to water and never water to strong acids or alkalines;
  • Work in the open or under a powerful fume extracting hood;
  • Never inhale fumes from liquids or from chemical processes taking place;
  • Keep children (including adolescents) away from chemicals and out of your chemical workshop. Store chemicals in locked closets;
  • Be prepared that while mixing or diluting liquids strong exothermic reactions may occur, spraying the mixture unexpectedly and violently out of the container;
  • After spilling liquid chemicals onto your skin wash away with plenty of water. Consult a physician.
  • Never leave chemical solutions in the basin for more than a few hours but empty the basin after finishing a job, Put the liquid in a tightly closed plastic container and store in a safe place.
The author of this website, nor the hosting company, nor any related legal entity bear any responsibility for incidents and damages occurring and possibly relating to any information on this website or due to lack of information.


Stainless steel does not rust onto air because it has a passive surface mainly consisting of chromium trioxide. Welding stainless steel (TIG) produces locally an intense heat that melts the steel and causes locally chromium depletion, leaving an iron rich zone. Despite the fact that an inert gas protects the weld seam, the high temperature causes oxidation that can be observed as blueish to brownish discoloration of the workpiece. The oxides formed are iron oxides, free iron, chromium oxides and nickel oxides. From a structural point of view the workpiece is sensible now for further (iron) oxidation. Therefore it is required to remove iron oxide, free iron and nickel oxide and to passivate the steel to a mainly chromium trioxide surface with a thickness of 20 to 50 Angstrom prior to further using it. Several methods exist such as mechanical techniques (grinding and polishing with passivation by exposure to air) and chemical methods such as pickling or etching with pastes or liquids.

Interestingly, also an electrochemical method exists, electropolishing, which can be done by submersion the total of the workpiece into an electrochemical bath, or locally by means of a carbon brush connected to a power source.
The principle of electropolishing is quite simple: by means of electrolysis a polarisation film is formed onto the surface of the metal. The "peaks" on the surface of the steel receive the highest current density and are eaten away faster than the "valleys" leaving a matt satin to a shiny surface. At the same time the surface becomes passivated. The combination of duration of treatment, temperature and current density determine the shining effect obtained.
By raising the voltage (V) from a low level upwards, the current (A) also rises evenly up to a certain critical voltage. From the critical voltage upwards the current remains stable until the critical volttage plateau has been passed and the current will start rising again.
Agitation of the bath is always required during the electropolishing process, e.g. by swirling gently with the workpiece.

After electropolishing it is essential to clean the workpiece by rinsing and sponging it with lots of water, followed by thorough cleaning with a steam cleaner. Failing to do this will leave chemical residues deep in the structure of the metal, which will continue to react or which might interfere with future use of the metal object.


Electrochemical processes are usually performed in stainless steel basins, such as used for professional kitchens. For a Fusor with a diameter of 20 cm basins with dimensions of 325 mm x 265 mm and  a height of 150 mm are suitable. The wall thickness is 0.7 mm and they cost about 12 Euros each. When filled up to a height of 120 mm they contain a bath volume of liquid of 9 liters and the size of the basin is sufficiently large to contain a half sphere of the vacuum chamber, about the largest in-house manufactured single part of our Fusor.
By clamping one of the current wires to the edge of the basin, the basin will be the negative pole, or cathode, of the electrolysis system.

However, this requires that we put a plastic thin mazed basket inside the basin to protect the workpiece (the plus pole or anode) hitting the wall and causing a short cut, which may burn a hole through our workpiece due to the high amperage applied. It may be difficult to find a thin mazed basket with the required dimensions and if we find one it may be necessary to find a suitably dimensioned stainless steel basin to fit with the basket, as the choice in dimensions of ss basins is far more larger. Therefore, another solution is to buy cheap plastic paddock plates and cut them to dimensions to suit the basin: one for the bottom and four for the sides. Tie them together with plastic strings. Or, similarly take Perspex plates, cut to size and perforate them with a 2 mm size drill (lot of work!) and glue them with chloroform into a box fitting inside the basin. The purpose of the inner basket is simple: insulate the workpiece from the basin wall and provide sufficient direct contact with the electrolyte to be in contact with the basin wall.

Next, we need a power supply and that should ideally be a supply able to provide AC as well as DC at 12 to 18 Volt but also capable to deliver hundreds up to thousands of Amps. For the DC supply we can make use of one or two lead acid car batteries, which are able to deliver over a thousand of Amps, though during a short while. This is not problematic as the process requires rather short times for the electrolysis. Never use Lithium or NiMH batteries as they may explode!


Electropolishing is the electrochemical process to obtain a shiny surface on stainless steel parts as an optical effect. It adds an improved passivation of the surface and can be used as an alternative to chemical pickling or etching.

The composition of the electrolytic bath is (reference 1):
  • one part by volume of 80% phosphoric acid
  • one part by volume of propylene glycol
  • one part by volume water
  • part by volume of sulphuric acid (d = 1.84)
  • part by volume ethanol or isopropanol
  • 0.5 gram of liquid dish soap per liter of water
The formula can also be used without sulphuric acid but with a less shiny result, producing more a matt satin finish. Leaving out sulphuric acid may also reduce somewhat the current flowing as it is a strong acid yielding a higher conductivity of the bath liquid.
The process temperature is 65 to 100C but heating is not required. Flowing the current will heat up the bath and instead cooling may be required.
The amperage to be applied ranges from 2.5 to 16 A/cm2. This seems not a lot, but when we want to electropolish a half sphere with a diameter of 20 cm we find an outer surface for a total sphere of 4 times pi times radius squared is 1257 cm2. For our convenience we assume that the inner surface is equal to the outer surface and therefore the total surface remains that of a full sphere or 1257 cm2. The amperage for elelctropolishing the total sphere therefore ranges (theoretically) between 3390 A to 16,970 A! The CCA rating of a car battery tells us that a certain type of car battery (for light trucks) can deliver during 30 seconds 700 to 1000 A at a temperature of 0C! This means that two car batteries in parallel just can be able to perform our electropolishing job. The point is that we do not have a method to control the amount of Amps that go into the workpiece.

When one has the opinion that this is a rather brute force method than you are correct. The process more or less short cuts the batteries during the very first (parts of a) second(s) until the "iron peaks" are eaten away and the resistance of the workpiece goes up. Short cutting a battery may lead to exploding of the battery or making it useless due to warping of the plates inside the battery causing an internal short cut. When this is too much of a risk than apply the brush method.

The brush method consists of a carbon brush connected to a (positive) variable current output 12V DC 100A power supply. The brush is dipped into the electrolyte solution and the welding seams (workpiece connected to the minus)  are brushed until passivated. The surface of brush, hitting the metal, is usually less than 2 cm2 and a current up to 40 A will do the job. Using larger surface brushes may enable polishing of larger objects.

What are the costs of the chemicals? For a bath volume of 9 liters we will need
  • 2.4 L 80% phosphoric acid
  • 2.4 L propylene glycol
  • 2.4 L demineralised water
  • 1.2 L sulphuric acid (d1.84)
  • 600 mL ethanol or isopropanol
  • 1.4 gram liquid dish soap
A local chemical supplier asks the following prices for the chemicals:
  • Phosphoric acid 85% pure: can of 5 L costs € 64.85
  • Propylene glycol: can of 5 L costs € 55.80
  • Demineralised water: can of 5 L costs € 19.00
  • Sulphuric acid (d1.84): bottle of 1 L costs € 8.30
  • Ethanol 99%: bottle of 1L costs € 4.50
  • Liquid dish soap: from our kitchen sink
A bit of calculus tells us that our electropolishing bath of 9 liters than costs just under € 67.00.
This can be lowered by buying our demineralised water in a supermarket where the same 5 L can costs as low as € 2.95 and propylene glycol is just car anti-freeze and can be obtained pure in a 5 L can for € 18.30.
Resourcing these components lowers now our bath prize to € 53.35 per 9 liters.
In reality the price for the bath is even slightly lower because the Phosphoric acid that we bought has a strength of 85% whereas 80% is needed. This is obtained by using a volume of 2.26 L of Phosphoric acid 85% and adding this to a volume of 140 mL of water to obtain 2.40 L of Phosphoric acid 80%.

A far more simple formula is the following:
  • Citric acid 55% w/w
  • Sulphuric acid (d 1.84) 15% w/w
  • Demineralised water 30% w/w
The working temperature is 90C and the current density is 0.1A/cm2 at 6 to 12 V.

Electropolishing of stainless steel can also be done with 3 to 18 V AC and this requires the following bath formula:
  • Nitric Acid (d 1.42): 70% w/w
  • Glacial Acetic Acid: 30% w/w
The bath temperature is 12 -15C and the current density is 0.04 to 2 A/cm2.


Construction of Vacuum Components

Diffusion Pump Flange and Baffle

As described on the Vacuum Page the diffusion pump outlet has a collar for inserting a metal tube as connection to the high vacuum  system. This is rather unique because normally an ISO DN sized flange or a Conflat flange is present to connect to other system components. Because I had now to construct myself an adapter that provided a junction to a common vacuum flange size, it seemed a good idea to include a baffle and a cooler in this contraption to be made.

For this junction piece it had to be taken into account that the next component upstream in the system is the high vacuum isolating gate valve. In accordance with the general rule that components upstream a vacuum system should not create a restriction in the passage of the gas flow and knowing that the diameter of the outlet of the diffusion pump measures 80 mm diameter, a vacuum flange size of minimally 100 mm iso DN or 4 inch Conflat is required.

Materials Specifications


An ISO DN 100 mm vacuum flange could be acquired cheap for EUR 28,36 (image 1 and 2):

flange DN 100  flange DN 100 b 
Image 1 and 2: ISO DIN 100 mm vacuum flange ( mcr0710)

The flange from images 1 and 2 was made by TCI Precision Metals as type ISO-100-000 N, Heat No. S2859-W1, Blank Vac NR 6.50 D ISO 100SS and has as dimensions an outer diameter of 165 mm, a thickness 12 mm, a number of 8 holes with a diametral pitch distance between holes of 145 mm and a hole diameter of 9.1 mm. The groove in the blank flange has an outer diameter of 101.9 mm, an inner diameter of 99.3 mm and is 4.8 mm deep.

The flange needs to be drilled or laser cut for a concentric hole of 80.0 (+ 0.1, - 0.0) mm in order to fit a 80 mm stainless steel tube, which will be brazed into the center hole, flat to the flange side with the groove and pointing out from the blank flange side.

Bolts and Nuts

The flange is fixed to a DN 100 gate valve with 8 bolts M8x40 mm, 16 washers for M8, and 4 nuts M8, all in stainless steel.

Center Ring
For connection to a gate valve with ISO DIN 100 flange, a DN 100 center ring is required (image 3):

DIN 100 center ring
Image 3: ISO DN 100 Center Ring


A stainless steel 304 tube with an outer diameter of 80 mm and 2 mm wall thickness was acquired in a length of 500 mm. From this length of tube two sections will be cut for brazing to the under and upper faces of the cooler housing. The ss tube has been polished on the outside.


A 200 mm length stainless steel 304 cylinder was acquired with an outer diameter of 129 mm and a wall thickness of 2 mm. The cylinder will be used in full length or cut to a smaller length. The openings at both side will be closed by brazing onto it (two) flat flanges of 140 mm diameter and 6 mm thickness. Each flange has a laser cut center hole with 80 mm diameter intended for brazing lengths of 80 mm diameter tube into it. The ss cylinder has been polished on the outside.

Copper Tube

As a liquid cooling element inside the baffle housing a length of 1 meter soft copper tubing with outer diameter 8 mm and inner diameter of 6 mm was acquired together with two couplings for connecting flexible tubing. A jig will have to be made for bending the copper tubing into a predefined shape. The tubing will pierce through the cylinder wall at two places as an inlet and an outlet and the piercing will be hermetically closed by brazing the tubing to the stainless steel wall.


Two stainless steel gitters (image 4) were acquired for bolting onto the copper tube inside the baffle housing. The gitters act as baffles and they have a diameter each of 126 mm and a thickness of 0.75 mm. This means that, in order to fit into the baffle housing, minimally 0.5 mm will have to be cut from the rim. It is the intention to remove this excess material by grinding.

Image 4: Stainless steel gitter as baffle

The gitter has slits with a total length of 520 mm and a height of 3 mm, yielding an open surface of 1560 mm2. At the rim the grid has four holes for bolts, with a diameter of 3.35 mm, placed in a 90 degree rectangle with sides of 80 mm length.


For completion of the baffle/cooler some more components are required, but these do not require alterations made in the workshop:
  • Thermoelectric cooling unit with integrated liquid cooler
  • Electric pump with combined water tank
  • Flexible tubing
  • Temperature control system
  • Coolant for low temperatures, e.g. undiluted propyleneglycol (-59C) or undiluted monoethyleneglycol (-13C)

Construction of the Bending Jig

This pump flange and baffle construction project has been set on hold for the time being, due to procurement of a turbo vacuum pump. This causes to leave out the oil diffusion pump from the vacuum system, at least until the turbo pump appears to be unsatisfactory.


Construction of a High Voltage Feedthrough

High voltage vacuum feedthroughs can be bought ready made, but the price is rather high. For our Fusor it was decided to construct a HV feedthrough from an industrial spark plug.
A description of the method can be found in reference 2, though we used a modified design.

The most important requirements for the park plug, however, is that it does not have an integrated resistor. Most spark plugs used in motor vehicles nowadays have a high ohmic resistor "built in". Therefore we will need to check that the spark plug measures a resistance of almost zero Ohm.

The spark plug is mounted into a KF25 vacuum connector and vacuum integrity is secured by placing an O-ring between the spark plug and the connector.

In order to insulate the feedthrough for the high voltage applied we need to extend the distance between the HV cable connection and the mass of the connector by using an extending brass rod, insulated with a ceramic alumina tubing. Connecting a cable just to the connector on top of the spark plug will be good only to about 10 kV, which is too low a voltage for our purposes.

Similarly we will have to insulate the electrode in the vacuum chamber by fixing a ceramic alumina sleeve over the electrode, followed by a ceramic alumina tubing. for fixing these components we will use low outgassing epoxy glue.

Materials Specifications

Spark Plug

For the spark plug we make use of a industrial type, the Champion (599) J99, which has no resistor inside. The screw thread for fitting the spark plug is M14x1.25, the screw thread for connecting the HV wire is M4x0.7 and the diameter of the stainless steel electrode is 2.05 mm (image 5).

spark plug
Image 5: Spark plug Champion (599) J99 ( Supplier)

The ceramic alumina isolator at the electrode has a diameter of 4.76 mm at the electrode and a diameter of 6.23 mm near the bottom of the screw thread.

When measuring the resistance of the spark plug we found a value of 7.1 m
Ω, indicating that indeed no resistor is present inside the spark plug. A Milli-Ohm meter with the 4-wire Kelvin method was used for the measurement.

Vacuum Connector

The vacuum connector is a stainless steel KF25/NW25 size connector with a female thread of M14x1.25, suitable for fitting a sparkplug (image 6).

KF25 connector
Image 6: KF25 threaded connector (Source: Supplier Hong Kong)


A Viton O-ring, type 12x2.5 mm with an ID of 12.0 mm diameter, an OD of 17.0 mm and 2.5 mm thickness, 75 Shore hardness, is fitted between the spark plug and the vacuum connector for ensuring vacuum integrity.

Ceramic Alumina Gas Collar

For insulating the electrode from the connector we will use a ceramic alumina gas collar. A suitable one was found from our plasma cutter machine, which uses PT-31 LG-40 shield cups with a top largest opening of 18.15 mm, a bottom smaller opening of 9.45 mm and a height of 30.5 mm (image 7).

gas collar
Image 7: Ceramic Gas Shield [upside down] ( FRS 2017)

The largest opening with an outside diameter of 18.15 mm fits nicely into a KF25 center ring, which is required for a vacuum tight connection of the KF25 feedthrough connector on the reactor KF25 connector.
The smallest opening of 9.45 mm diameter will be widened by means of a diamond rotary burr dril to 9.54+ mm in order to provide for a passage for the ceramic tubing.

The ceramic alumina gas collar
is epoxy glued over the ceramic collar of the electrode, which is extending from the bottom of the spark plug. The smaller opening of the gas collar fits over a ceramic alumina tubing that covers the electrode.
When glueing, care should be taken that the epoxy glue does not contain air bubbles
after mixing the two components, as this may cause problems when applying vacuum. Also, a low outgassing type of epoxy glue should be chosen.

Ceramic tubing

Commercially available ceramic alumina tubing, with dimensions of OD 9.54 mm, ID 6.36 mm and a length of 304.8 mm, will be used to cover the (extended) brass rod conductors for both the HV connection and the electrode (image 8).

ceramic tubing
Image 8: Ceramic Alumina Tubing (source: Supplier)

The dimensions of the ceramic alumina tubing permit a tight fit around the ceramic collar at the spark plug electrode. It leaves a very small gap between the tubing and the metal base of the spark plug with the advantage that only a limited surface of the epoxy glue will be exposed to vacuum. This will greatly reduce eventual outgassing problems of the epoxy.

The tubing will be cut to required lengths by means of a Dremel with a diamond cutting disc.

Brass Rods

For extending the HV connector, a brass rod will be used with a diameter of 6.0 mm and a length of 250 mm, consisting of a metal alloy complying with MS60 quality norms (image 9).

brass rods
Image 9: Brass Rods (Source: Supplier)

For extending
the spark plug electrode, a part of a stainless steel TIG welding rod will be used with a diameter of 3 mm.

Epoxy Glue

As mentioned before, the requirements for filling and glueing parts subjected to vacuum require that a low outgassing product will be used that can join different materials, e.g. stainless steel to ceramics.
For this purpose special products exist, such as Torr Seal from Varian and Hysil 1C from Loctite, which are basically identical. These products are good up to 10-9 Torr but they are quite expensive, not easy to locate and subjected to transport regulations forbidding shipment by air.
We have made use of J-B Kwik, a generally available two component epoxy (image 10).

JB Weld
Image 10: J-B Kwik Epoxy (Source: Supplier)

J-B Kwik has a set time of 6 minutes and after 4 - 6 hours it has become a hard, solid product that can withstand a temperature of over 200C and has a tensile strength of 16.7 N/mm2 (156.6 kg/cm2). The relatively high application temperature makes this product very suitable for outgassing with heat after the product has fully hardened. The heat applied for outgassing the product is usually just over 100C.
After outgassing, a vacuum better than 10-6 Torr can easily be achieved. For our Fusor we will never exceed this level of vacuum.

Construction of the HV Feedthrough

Construction of the feedthrough is started by removing the metal gasket ring from the spark plug. It cannot be pulled off, but it needs to be unscrewed.  Next is placing the Viton O-ring over the screw thread at the base of the spark plug. By using a 12x2.5 mm O-ring it will fit in the groove between the thread and the bottom flange of the spark plug (image 11).

spark plug and o-ring
Image 11: Spark Plug and O-ring ( FRS 2017)

Next step is to fill the male thread on the spark plug, as wel as the female thread on the KF25 connector with high viscosity vacuum grease. Because the threads might be in direct contact with vacuum, vacuum grease is applied in order to prevent repeated outgassing of air from the threads, which may take a long time to cure. Assemble the spark plug and the connector by screwing the two parts slowly together to a tight fit; observe that the O-ring needs to be slightly compressed for a vacuum tight fit (image 12).


Image 12: Spark plug and Vacuum Connector Assembled ( FRS 2017)

At the bottom of the connector, with the electrode protruding, scrape away any excess vacuum grease. Then, use a cotton swab with acetone to degrease thoroughly the metal and the ceramic alumina isolator. Degreasing is required for preparing these surfaces for permitting glueing with two component epoxy.


control systems
Ion Source


Ref. 1: Electropolishing: http://bbs.homeshopmachinist.net/archive/index.php/t-38026.html

Ref. 2: Sparkplug HV Feedthrough: http://home.earthlink.net/~jimlux/hv/sparkplug.htm

Last Updated on: Tue Jun 6 23:29:29 2017

Flag Counter

Document made with KompoZer