Neil Hopcroft

In this article the Author describes some simple methods of making at home Mains Transformers of any type – by Frank Preston

Every practical man, whether he be interested in model making, electrical experimentation, accumulator charging, chemistry or wireless, at some time or other requires a source of eletrical energy. This can frequently – and expensively – be derived from batteries or generators, but the electricity mains provide a far more convenient and reliable form of supply. AC mains, in particular, are extremely useful to the experimenter, since the voltage from them can be changed to any require figure with the greatest of ease. All that is needed is a steup-up or step-down transformer.

Suitable transformers can very easily be constructed, whilst the work of making them is both interesting and instructive. It is fairly well known that a transformer consists essentially of an iron core upon which are placed primary and secondary windings, although most people find it very difficult to decide upon the size of core and numbers of turns required for a particular purpose. The necessary data can be obtained by making various somewhat complicated calculations, but if a few simple facts are known the arithmetic involved is not outside the scope of an intelligent schoolboy.

The relevant facts will here be given as briefly as possible, and also some tables which will remove any trace of tedium from the process of design.

Core Stampings.

The type of core most frequently employed for small transformers is that consisting of a number of pairs of “U” and “T” shapped stalloy stampings of the kind shown in Fig 1. When these are assembled they form a semi-solid core with two “windows” and a winding limb (see Fig 1). Assuming that the stampings are of correct proportions (as all those on the market are) the numbers of “turns per volt” for both primary and secondary windings depend upon the cross-sectional area of the winding arm and the frequency of the mains supply. For example if the area is 1 sq. in. and the frequency is 50 cycles, eight turns should be allowed for every volt. If the area were halved the numbers of turns must be doubleds, and vice versa; on the other hand, if the frequency were doubled, the turns hould be halved, and vice versa. This rule, although so utterly simple, is invariable and forms the basis of all transformer design.

The stalloy stampings mentioned above are made in a variety of sizes, some of which are listed in Table I, where “A”, “B” and “C” dimensions are those defined in Fig 1. In order to determine the most suitable size of stampings it is necessary to known the power, in watts, which the transformer has to handle. This is easily calculated by multiplying together the voltage and current (in amperes) of the secondary winding. For example, suppose the transformer had to supply 20 volts at 2 amperes, the wattage would be 20×2, or 40 watts. This assumes an efficiency of 100 per cent., but as the actual efficiency is generally about 80 per cent. the result must be increased by 25 per cent., which gives the power to be handled as 50 watts. Reference to Table I then shows that a core consisting of six dozen No 4 stampings will be suitable.

Choosing the Correct Wire Gauge

Once the core size has been determined the winding data can be conmpiled. Starting with the primary, which has to handle the total amount of power (50 watts), it will be seen from the Table that eight turns per volt will be required, so it only remains to decide upon the gauge of wire necesasary to carry the current involved. The current is found by dividing the wattage by the voltage of the supply; for instance, supposing the voltage to be 200, the current would be 50 / 200, or 0.25 ampere. The correct gauge of wire could then be determined by looking up a book of wire tables, but to save this trouble the necessary information in regard to the gauges in most common use is given in Table II, where the smallest possible gauge suitable is seen to be Number 30. As this table is based on a current density of 2000 amperes per square inch, however, it is slightly better, where space permits, to emply a gauge of wire one size higher than the minimum shown.

The secondary winding will consist of 8×20, or 160 turns, and since it has to carry 2 amperes the wire should be not less than 20 gauge.

In regard to the covering of the wire, this may conveniently be enamel in all gauges less than about 24, but for the stouter gauges it is better to use double-cotton-covered, since enamel is liable to crack and so allow turns to short-circuit.

Winding area

The size of core was provisionally decided on in the first place, but as the winding data is now known, a check should be made by finding the actual “winding area” required (see Fig 2). This area can easily be determined by making use of the “Winding turns per square inch” given in Table II. Taking the same example as before, we see that 28 gauge enamelled wire can be wound 3760 turns per square inch, and therefore our 1600 turns will occupy rather less than 1/2 sq in. The secondary consists of 160 turns of 20 guage dcc wire, which can be wound 472 turns per square inch, and will therefore take up apporximately 1/3 sq in. In other words, the total winding area required is 5/6 sq in, and as the No 4 stampings provide 1 1/2 sq in winding area they will be amply large.

The Winding Spool

We can now turn to the practical side of the question. We know what core stampings are going to be used and so we can make a spool to fit them. The simplest method of making the spool is illustrated in Fig 3. A square-section cardboard tube is first required and can be made by scoring and bending a strip of stout card of the dimensions shown. Next a pair of end cheeks must be made to fit tightly over the ends of the tube, and these can be cut out of stiff card or thin plywood and secured by means of strong glue. To make the bobbin more rigid it should finally be given one or two applications of thin shellac varnish and dried quickly. To cover the sharp edges of the bobbin, which might cut the wire whilst winding, a few turns of empire cloth or insultaing tape should be wound on.

Winding the Primary

Now solder a short length of flex to the end of the 28-gauge wire, anchor this by passing it through a pair of holes in an end cheek and wind on the correct number of turns for the primary. The winding can be done most expeditiously by fitting the spool to a mandrel which can be turned in the lathe or a hand-drill gripped in a vice, but it can be done by hand if desired by cutting a handle of wood which is a tight fit in the spool. In winding, attempt to arrange the wire in reasonably even layers and maintain steady tension on it to avoid slackness. After every four layers, or approximately 500 turns, it is advisable to cover the winding with a layer of empire tape, oiled silk or waxed paper to avoid the possibility of any two turns at widely differing potential getting close together. Take care that no later turns are allowed to slip past the layer of insulation.

After winding the requisite number of turns a second length of flex should be soldered to the end of the wire, taken once around the spool and anchored as before. Thoroughly insulate the primary by covering it with two or three layers of empire tape, etc., and then continue to wind the secondary, following the same procedure as with the primary. Finally cover the outer layer with insulating material to ensure that the windings cannot be damaged in any way.

Assembling the Core

The core stampings must next be fitted, and the method of fitting is clearly shown in Fig 6. First a “T” and then a “U” are inserted from one end of the spool, after which a similar pair of stampings are insertered from the other end, this process being repeated until the spool is quite full. In order to make the core a tight fit (as it must be to prevent vibration), it might be necessary lightly to tap the last few stampings into position, but undue force must not be used or else there might be a danger of “bursting” the spool. It will be noticed that one side of each stamping is covered with a white insulating film and, to ensure that every one shall be insultated from the next, the white sides must face in the same direction.

Core Clamps

The last step is to fit suitable clamps to the core to hold the stampings tightly together and provide a simple means of mounting the complete transformer. These clamps can be made from 1/8 in thick strip brass or steel, shaped and bent as shown in Fig 7. They are attached by means of 1 1/2 in bolts and can be fitted with a terminal strip if desired, or connections can be made directly by means of flexible leads from the winding. Both methods of finishing are shown in Figs 4 and 5.

All details given above, although they have been applied to a particular component, are equally applicable to any pattern of mains transformer that the reader may require. In some cases it is more convenient to design the transformer, so that it can be used on any mains having a voltage of between, say, 200 and 250 volts. In that case the primary winding would require an additional 400 (eight 50) turns and tappings would have to be taken after winding 80, 240 and 400 turns for 240, 220 and 200 volts respectively. The tappings would be made by soldering suitable lengths of flex and passing these out through holes made in the end cheeks. To safeguard against short circuit between the tapping points the soldering joints should be covered with a strip of insulating tape, or even with a piece of stamp edging.

When more than one secondary winding is required, such as for HT and LT supply for a wireless receiver, it is geenrally most convenient to divide the winding spool into three or more sections by fitting extra cheeks. The positions of these will be determined by the area required for winding in the different sections. In order to prevent mains hum it is best to place the LT secondary in the centre section, where it will serve as an effective screen between the primary and HT secondary windings. With all other kinds of “dual-secondary” transformers the primary winding should be arranged between the other two.

Figures to follow once scanner is available