A Power-pack for "250" Valves

Radio Record, Volume III, Issue 25, 3 January 1930, Page 29

A Power-pack for "250" Valves

"For D.C. or A.C. Operation

(Continued from last week.) Tapped Secondary Windings. ONSTRUCTORS wishing to commence operations of 300 volts with a Raytheon will wind the full secondary coils, tapping each one at a suitable point so that the reduced voltage may be used as long as desired. Care must be taken to follow the proper "procedure in placing the taps and in connecting the coils, otherwise the two halves of the secondary may be acting in opposition, and new voltage will be, th} result, he number of turns to be utilised will be 1850 on each half of the secondary winding. This will give an initial Woltage of 360, which will drop to less n 300 by it reaches the plate of the i t valye. The Raytheon is quite equal to handling the voltage. The first coil is wound with 1850 turns, and a tap is taken out; then the remaining turns are put on. The second coil is to be wound, turning in the same direction and putting on a number of turns equal to those on the first coil in excess of 1850. A tap taken out, and then 1850 turns to camplete the coil. The beginning, or larger number of turns of coil No. 1 is then connected to the end, or larger number of turns of No. 2, and from this connection the centre tap is taken. Thé two taps connect to the respective "filament" terminals of the Raytheon socket, and the end of coil 1 and the beginning of coil 2 are coiled up and insulated with adhesive tape. This same method may be utilised to obtain several tappings-the taps will be near the finish of the first coil, and then repeated in reverse order at the beginning of the second coil, The 800 volts will suit the £45 valve.

Filament Windings. FTEHR insulating the* secondary windings, any filament windings that may be required are put on. These windings are usually of heavy wire, 18’s., as heavy current is to be drawn for a.c. valve heating. The actual voltage of each winding and the amperes to be drawn, will depend upon the valves to be used in the receiver. + ‘The power stage filament supply may be put on first, and for one or a paar of 74 volt valves, 47 turns of 18’s d,@ec. will be required, allowing 24 ampies to be drawn. If a single valve is used, only 14 amperes will be drawn, which will tend to increase the voltage, so that a tap might be provided at the second turn from one end, or provision made for the introduction of a few ohms resistance into both legs of the filament circuit. The power-valve winding is not centre-tapped, as the electrical centre will be accurately determined by means of a potentiometer of 30, 60, or more ohms across the filaments, giving a variable tap by means of the arm, which is directly connected to earth unless a grid-bias resistance is included in the circuit: This method of centre-tapping power filaments was dealt with in the "Record" of May 24 last. The Radiotron 227 valve requires 24 volts, 13 amps, The De Forest and Ceco equivalents require the same, and the Osram 34 volts 2 amps. The Philips

#424 and several others in the heater series. of the \same make require 4 volts and .9 amp.,. or. just under one ampere. Windings to give any voltag ge required by the above or other valves are as follow :-

Some "ee The above table will give voltages as near as possible to those required, but it ig necessary for constructors to exercise care with their a.c, heater current in order not to give any excess, A small amount of resistance can ‘always be introduced to reduce voltage, half in each filament or heater leg. The above table is correct for 18’s wire, which is easier to manipulate than heavier gauges when winding. A good plan.is to provide .a .-separate winding for each 2 or 24-amps., as good regulation is theréby secured. Heavier wire gives rather less drop when large current is drawn, *°

Two 4-volt windings aré easily accommodated in one layer, and three 2}volt. Centre-taps are not required on these windings, as the centre is usually determined by the tapped resistance method, though it is not critical with some valves. Shellac all filament windings, and allow it. to dry before covering with insulation. If any filament windings are put on for future use, one end should be connected to earth (metal floor) and the other coiled up small and insulated by wrapping in adhesive tape. The transformer dimensions given allow for one layer of 18’s for rectifiers between primary and secondaries, and outside the latter one layer 18’s for power valves, and three other layers for heater supply. The outer windings may project very slightly outside the spool ends en the straight sides, but not at the corners. The Smoothing Chokes. WO smoothing chokes of identical proportions are to be provided, and the specifications here given will produce a pair of chokes that will be equal to the task imposed upon them. The inductance -will be about 20 henrys each, when carrying up to 130 mills, with two gaps of 1-16in. in each. The number of turns of 30’s s.wW.g. on each will be 5750, but constructors not having a _ revolution counter may save the trouble of counting by simply filling each spool, which will take half of the 33lb. of wire allowed for the pair. The cores are constructed of 1}in. stalloy, built to the same thickness. The long piece-that which is packed inside the centre of the spool-is 4#in. The other sizes are 3}, 2, and #-inch, all 13 wide. The spool ends are 2%in. square, and the outside length 2tin. bare. Remember. that there will be a -high potential difference between the winding and the earthed core, so let the preliminary insulation be goodempire cloth or two brown papers in addition to a layer of tape. The only great difference between the construction of a choke and a transformer is in the arrangement of the core. In order to provide: the "gap" in the choke, it is necessary to cut four sizes of stalloy, one heap of each to the thickness to which the core

is to be built. The centre of the spool $s packed tightly with the longest pieces, to which the thicknesses of card~ ‘board which. determines. the gap may be fastened with seccotine, when the remainder of the core may be ase sembled with the three smaller sizes, The direct-current resistance of. each choke will be 260 ohms, or 520 ohins the pair, which is fairly low for un eliminator choke. If a current of 50 mills is passing through the pair the drop will be 26 volts, 70 mills will drup 38 volts, and 100 mills will drop 52 volts. ; This drop in voltage is well provided for in the specified turns, especially. when it is noted that the drop in filas men rectifiers of the 281 type is piactically negligible up to 60 mills, after which the internal resistance gradually rises, Wooden clamps 1} wide by 3-8 thick and 4 5-8in. long are drilled with bolt holes, centres 8 7-8in. apart. The chokes aye secured to the baseboard by means of holes drilled through it, and through which screws are passed from underneath to secure the lower edge of the sottom pair of clamps. The lead-outs should be at the lower end of the spool, is they pass through the baseboard in every inastance in order to connect to the smoothing condensers. The wire is run in "without" insula-.. tion, but it is just as well to put in a layer of tissue-paper occasionally as @ precaution, keeping it close at the ends. Two dozen 8ft. lengths of stalloy will be required for each choke, about 80 of each size of cut piece for each. The chokes and transformer cores are earthed by placing under them a few thicknesses of tinfoil folded up, thus making good contact with the metal floor. The tinfoil in which adhesive tape is packed answers well. For a greater output than 130 mills the chokes would have to be construct~ ed of 14in. stalloy built to the thickness, and the window enlarged in both directions to take the same number of turns of 28’s wire. The Smoothing Condensers. DIAGRAM shows a neat arrangement of the smoothing condensers in the base, but other makes may necessitate a different arrangement in order to suit the space. It.is a wise plan to make the container after the conden- sers have been procured. The four 4 mfds must be of 800 to 1600 volts test -half the test voltage is the working voltage, and a margin is always good.

ay, Hitérest ciastruntors to know Seo hing: of the actual function of the condensers marked Ci, ai O3.in the circuit diagram. ae The function of C1 is that of reduc- > ing "ripple? or hum. Its capacity ‘ should not be too low, because a reasoneb high capacity improves | regulation. -The condensér betweerl" the two = C2, is merely to reduce ripple, aoe Same has no effect upon; régulation.. If ‘this: condenser is: not of suitable value it can cause trouble by resonating with the previous choke at 100 cycles on a 50-cycle supply... The extra demand is all made upon C3. Some. circuits employ only 2 mfds. in: this position, but it pays to be liberal with-the smoothing capacity. The last capacity, C38, controls audioquality in:a perhaps somewhat indirect way. There are large audio-fre-quency: variations in the plate current of the last -valve-or valves, and it is the duty of C3 to supply the extra current required when a heavy signal arrives on the grid of the power-valve. This condenser is, then, acting as a storage capacity, and when an extra demand is made its voltage drops, but is-rapidly raised again. The output through the last. choke is steady, direct current, and as fluctuating demand cannot be met. If the capacity of C3 is too low, the extra demand will not be met, and so distortion results. Laboratory experiments have shown that increasing the capacity of C3 up to 6 mfds,. gives constant improvement with each microfarad added. From. 6 to 8 mfdé. there is still improvement, but less marked, and after 10 mfds. no difference is noticeable. This shows 3 mfds. to be adequate for ordinary purposes, and the writer’s own experience shows this to be the case. An important point to note is that the common or negative sides of all the 4 mfds. are connected together,

---. but this must be done with insulated wire, which connects to centre-tap of the B supply and the high grid-bias output. If this lead is earthed the bias resistors are cut out of the circuit. As ©1' has to withstand the full a.e: voltage of half the combined secondaries, it may well be of 1000-volt test rating, Any condenser following a choke is ‘liable to receive heavy surges from that ‘choke at the moment the mains current is switched off, and as that surge voltage is added to that already held by the condenser, the total voltage may be very much higher than the working voltage. This is one reason for having a high test rating for the condensers, The Diagrams. N°: 6 shows the transformer complete, as viewed from the back of the eliminator. Note the convenient position of necessary filament and heater leads, which are cut short and soldered to the flex, each joint to be insulated with tape. The rectifier fitament leads.will also go down through holes in the base. The fuse panel is clearly shown. Keep fuses: near edge of panel so that solder tags bolted on back will project. Use 1-8in. brass bolts and washers to suit. Bach pair of bolt holes drilled with centres lin. apart. ‘ No, 7 shows’ the strip of tin already mentioned, and No. 8 the method of tapping and connecting up two separate high-tension secondaries, The plate leads are both connected to either the taps for the lower voltage, or to out 1 and in 2 for the high voltage. No. 9 gives the approximate position of leads through spool ends. Mark these out with the 1 38-8in. running ix the direction shown. The Voltage Divider. HE type of voltage divider for tapped resistunce to be employed consists of a continuous high resistance of wire, connected across*the positive and negative, and provided with taps at certain points to give a selec-

tion of plate voltages. Connected to the negative end of this resistance is a further resistance made up of two 400-ohm wire-wound potentiometers in series. ‘These connect through to the high-voltage secondary centre-tap, and will provide two variable . grid-bias voltages. There are a number of voltage dividers on the market, most of them arranged to reduce 200 volts to the necessary voltages for the average receiver. In this case we have 400 volts available for the plate of the power valve, and this voltage must be reduced to suit other stages of the receiver. We can utilise a 300-volt potential or voltage-divider if we precede it by extra resistance of the value that. will effect the necessary reduction. ~The "pilot" voltage divider is one that would be suitable, and by placing an additional one in series for extra resistance, the 400 volts will be suitably cut down. In case of utilising a make of divider capable of carrying less than 60 mills, it is a better plan to use de Jur or Ward-Leonard fixed resistances totalling about 12,000 ohms as the extra resistance. These are connected in series and placed between the powervalve lead-off from the second choke and the high-voltage end of the voitage dividex. The "pilot" divider is made up of resistances of 4000, 3650, 2250, 2800 ohms, totalling 12,700 ohms. The 4000-ohm end is B — and the tappings available will be 180, 135, 90, and 45 volts. For a low detector voltage ft will be necessary to include on the panel a variable high resistance of 500,000 ohms and include it in_ series with the 45-volt tap connection to the panel socket. The voltage divider may be made up in any way, provided that the total resistance of about 22,000 to 24,000 ohms is included, also adding in the grid-bias value. Many constructors will have to use the type of voltage divider that they are able to procure. Few dividers will dissipate sufficient heat to allow of their being used as extra resistance, $0 it will be safer to use de Jur or Ward-Leonards if the divider available is of very low dissipation, as considerable heat is generated in the extra resistance. « Make the total resistance work out as follows :--Divider 12,000 ohms, extra resistance 12,000 ohms, grid-bias r2sistors, 800 to 2000 ohms, giving a total of about 25,000 ohms. A thousand ohms difference either way is not very important. .