## Ampere Ac Arc Welder

The welder, although not an essential item in the hamshack, is a welcome addition to the experimenter's supply of tools. It serves such areas as tower construction and general repair. The high cost of an adequate commercial welder and my native curiosity started me on this project which has seen a year and several different transformer designs pass. The present unit is the culmination of my experimentation. It is a continuously variable unit that will weld to 185 ac amperes maximum. The duty cycle for this unit at maximum current is about 15%. which would not satisfy the industrial user, but for the average shop is quite adequate. A simple duct and fan cooling arrangement could raise this value.

The type of transformer I have designed allows one to control the leakage of the unit, which in turn controls the maximum amount of current that can be drawn. By varying the amount of leakage, one can control the current available at the welding electrodes. Most transformers are designed to minimize leakage. This application is an exception. The formula used to calculate the number of turns for the primary and secondary windings is:

V/N = 4.44 B F S A x 10"8 B = flux density in gauss (flux is the number of lines of magnetic force and is dictated to a certain degree by the metal used in the core).

F = frequency of operation of the transformer (in this case 60 cycles). S = stacking factor (this allows for varnish between the laminations of the iron core).

A = cross-sectional area of a core leg in square centimeters. V = volts applied or desired. N = number of turns required.

After choosing the flux density at which to operate the transformer, considering the iron used (I chose 15,000 gauss), the other terms are fixed by external conditions, except for the number of turns and the cross-sectional area. 1 picked a cross-sectional area that would be large enough to

The completed welder. Note the on-off switch, fuses and shunt mechanism in place.
Fig. I. Welder schematic.

give a reasonable value for the number of turns required. The cross-sectional area that was chosen required approximately 235 turns on the primary (the winding connected to the 220V main line) and 19 turns on the secondary (the winding the welding cables are connected to).

To vary the leakage of welding current, there are two controls. The range switch, which is the coarse current control, actually varies the proximity of the primary winding to the secondary winding by a series of taps which puts a greater or lesser part of the primary directly over the secondary and the rest around the other leg of the core. The

Here, the U-shaped part of the core is being assembled. The large bolts and plyv/ood base are part of the temporary clamping device for the core.

farther apart the primary and secondary, the greater the leakage and the lower the welding current.Vernier current control is accomplished by sliding a laminated iron bar between the two coiis and thus shunting flux away from the coil on the other leg of the transformer (leg 2), see Fig. 1. Flux is the term used to describe the magnetic field which couples the primary to the secondary and thus allows the transformer to function. The less flux, Ihc less coupling and the lower the welding current.

### Construction

The core itself is a laminated structure made of .025 in. thick .25% silicon steel sheet. Plain cold rolled steel will serve if the silicon steel cannot be located. The steel is cut into strips as per Fig. 2. The laminations are then assembled with varnish in alternating layers 1 and 2 (Fig. 2). A metal shear is necessary to cut the laminations. Make sure that the shear is in good condition or you may have to file burrs off each and every lamination since it is necessary to avoid metal-to-meta! contact between the laminations. If liberal amounts of varnish are used and the laminations are reasonably free of

Here, the U-shaped part of the core is being assembled. The large bolts and plyv/ood base are part of the temporary clamping device for the core.

burrs, no problems should be encountered.

After the U-shaped part of the core is completed, it. should be placed between two % in. plywood boards and clamped tight to remove excess varnish. After clamping, the core should be approximately 2% in. high. This will require approximately 109 layers of iron. Thi core laminations may need to be aligned after clamping the first time. Although the varnish will not completely dry for over 9 months, it is very difficult to realign the core after only a few hours, so make sure that il is square. The shunt can also be assembled at this time. The partially finished core and remaining lamination can now be put aside to cure and the coils can be wound. After a week, the clamps can be removed if the core is handled with a little care.

The first job is to build the bobbins that the coils will be wound on. They are constructed out of 3/32 in. plywood, which can be purchased al any hobby shop.deahng in radio controlled airplanes. I have left out the exact dimensions for the bobbins because they should be custom fitted to your

TAPS TAPS

Fig. 3. Bobbin. Note: Adjust the inside dimensions of the bobbins to fit the core that you have constructed. The bobbin shown is the one for the secondary in which three ] 2-gage wires are wound together. For the primary, only one wire slot is required.

although it can be done by hand. Note the tap wire protruding from the coil. Here, paper insulation is being placed between two layers of wire.

Fig. 4. Coil orientation pictorial. Note the direction of the v/indings of the various coils and follow it exactly!

core legs. The bobbins were assembled with gauze and glue (Fig. 3).

About 700 ft of 12-gage copper magnet wire is required to make the coils. It was obtained through a local wire distributor. The secondary was wound first. It is 3 trifilar winding consisting of (hree strands of 12-gage wire side by side and wound together.

Cut a piece of wood that will just fit inside the bobbin and clamp it in a vise; then get a friend to help arid start winding. Seventy-nine turns are required on the secondary (Fig. 4). This will be approximately 6 layers of wire. Between each layer place a couple of layers of paper insulation (wrapping paper is fine) and four 'A x 4 x 3/32 in. plywood spacers on each of the four sides. Varnish the paper after it is wrapped around the coil. The air spaces provided by the plywood spacers allow for air cooling of the coils during operation. The primary coils are wound similarly, except they are mono-

filar windings consisting of one strand of wire. One hundred and sixty turns are wound directly over the secondary. One hundred and thirty turns of primary arc wound on the second bobbin. A layer of gauze is placed over the last layer on each bobbin and varnished.

In wiring the two primaries together, the relationship of the two coils in respect to the direction they are wound and the way they are connected together is critical, so follow the diagram to the letter. If by chance a mistake is made, the welder will hum loudly and heat up excessively when turned on. When winding the primaries, observe the positions of the taps and place them as you wind, using a 100W soldering iron and rosin-core solder. The ends of the coils, both primary and secondary, can be held securely by fastening the ends of the wire with string. Tie the string onto the end of the wire, then wrap the string around the coil several times and tie it.

Remove the U-shaped part of the core from its clamps and clamp the bottom of the core with two I lA x P/i x 1/8 x 10 in. angle-iron pieces with holes drilled at the ends to accept 3 in. machine bolts, insert the bolts and tighten. Place the coils on the legs of the U-shaped piece and complete assembly of the core. Clamp the new assembled portion of the core with plywood and C-clamps. After several days, angle-iron clamps can be placed on the top of the core as on the bottom in place of the C-clamps. The phenolic board for the range switch can be mounted directly on one of these angle-iron pieces.

imjw, i

A complete bobbin ready to he placed on one of the core legs. Note the siring used to tie down the end of the primary winding.

Fig. 5. Shunt clarnp plate. Note; (1) Two plates are required; (2) One of the holes marked for M in. hole in one of (he plates is drilled with a ifS drill and tapped with a Vn x 20 tap. This will ride on the threaded rod which, when turned, moves the shunt in and out of the core.

A complete bobbin ready to he placed on one of the core legs. Note the siring used to tie down the end of the primary winding.

Fig. 5. Shunt clarnp plate. Note; (1) Two plates are required; (2) One of the holes marked for M in. hole in one of (he plates is drilled with a ifS drill and tapped with a Vn x 20 tap. This will ride on the threaded rod which, when turned, moves the shunt in and out of the core.

if the shunt has not as yet been made, it can be done at this time. It is made just like the core. It is a stack of 3 7/8 x I 'A in. iron plates varnished together; however, It is only made 2 in. high instead of 2% in. as the core proper. Two 3/8 in. plates, one aluminum and one phenolic, are drilled appropriately to clamp the shunt (Fig. 5). One of the guide holes in the aluminum plate is threaded (Vi-20). A threaded rod along with a support rod on the opposite side of the shunt both support and move the shunt in and out of the core as the threaded rod is turned. Minimum current for a particular range is when the shunt is all the way inside the core and maximum is when the shunt is all the way out.

For the main switching deck, a piece of 3/8 in. phenolic board is used. Brass bolts are mounted in the board and serve as binding posts. When shorting wires are connected across the appropriate posts, a particular current range is selected.

Observe the wiring diagram. The arrangement of parts is not critical but try to keep the secondary leads as short as possible. The electrode and ground clamps as well as a face shield can be purchased from Sears, Roebuck. The welding cables, two 12 ft lengths of 4-gage stranded wire, can be purchased from any welding supply house.

At this time there are three of these welders in operation and functioning quite satisfactorily.

0 0