Radio Broadcast (May 1927-Apr 1928)

214 RADIO BROADCAST JANUARY, 1928 FIG. 2 about 20,000 ohms was the most pleasing the writer has ever listened to. Ry the turn of a single knob it was possible to bring out the bass or treble to any relative degree to suit the particular ear of the listener. D. C. Redgrave Norfolk, Virginia. Another Cone and Horn Combination I I KE many other radio fans I recently discarded my horn loud speaker for a cone. When listening attentively, I decided that the cone slighted the higher frequencies somewhat so I dug out the old horn, which rather favors the high notes, with the idea of using the two loud speakers in combination. But the impedance of the loud speakers were so poorly matched that some balancing arrangement had to be devised. 1 evolved the scheme illustrated in Fig. 3. employing a halfmegohm potentiometer, such as is often used for a volume control. As the slider is moved, one loud speaker is cut in and one eliminated. I soon found, however, that I needed a better criterion than my ears to place the slider where it belonged, so I plugged a pair of good phones in the detector plate circuit and adjusted the potentiometer until the loud speakers duplicated the pitch of the telephone receivers. H. D. Hatch Boston, Massachusetts STAFF COMMENT THE choice of the systems outlined above by our two readers rests pretty much on the convenience of what you have in the way of a variable resistor. Fig. 3 provides a greater variation in loud speaker selection than Fig. 2 but sufficient variation can be secured with the latter for the purpose of correct balance. In the arrangement shown in Fig. 2 the variable resistor should always be connected across the loud speaker having the higher impedance, i. e., the loud speaker that seems to produce the most sound when the resistor is disconnected. Mr. Hatch's system for adjusting the relative intensities of the two loud speakers is worthy of special note. While a preference may exist for an emphasis of high or low notes (adjusting to suit individual tastes), this is not necessarily natural reproduction. The phone test method probably provides as correct a balance as can be achieved without elaborate equipment. C Bias from B Socket Power Units IN THIS department for November, 1927, an ' interesting item was published showing how a B socket power device could be changed to supply grid bias voltages as well as plate potentials. In effecting this change, it was necessary to "get inside" the device, break a connection, and make several additions. This method is quite practicable, but there is always a general disinclination to " monkey " with a commercial set-up. It is possible to obtain the C bias potentials by application of external resistors — no change whatever being necessary in the device itself. In so doing, several possible faults in the device can be corrected, such as break down of resistors, as may be evident in noisy reception and poor voltage regulation. It is also possible with the external arrangement to secure voltages other than those supplied by the original device for special sets or purposes. The additional device takes the form of an entirely separate set of resistors, which is connected between the high-voltage and negative posts on the old power unit. The resistors are tapped to supply the desired B and C potentials and are mounted on a base-board. The number of resistors required is determined by the number of voltage outlets. We shall need one Amsco "Duostat" to supply two variable "C" biases, while one resistor will be required between the negative tap and the lowest positive plate tap, and another resistor for each additional plate tap. The Amsco "Duostat" is equipped with two variable arms, making it possible to secure two variable C bias potentials with the one resistance unit. The average arrangement is shown in Fig. 4, which requires three fixed resistors in addition to the " Duostat. " The fixed resistors may be any satisfactory power resistor, such as those of Electrad, Amsco, Metaloid, Ward Leonard, Durham, Carter, etc. The method of calculating the values of the resistors takes into consideration the probable plate current drain through each resistor, and is best illustrated by the example of Fig. 4. An arbitrary value of ten milliamperes is chosen as the loss current — the current through the resistors over and above that drawn by the receiver. The presence of this loss current reduces the variations in voltage with slight differences in load. The higher the loss current the better the voltage regulation. The output from the average power device receiver combination under load is about 200 volts. By an application of Ohm's Law, which tells us that the resistance is always equal to the voltage drop divided by the current in amperes, or one thousand times the voltage drop divided by the current in milliamperes, it is a simple matter to determine the total resistance necessary. The equation in this case is: So the total resistance of the "bridge" will be 20,000 ohms, 2000 of which is already apportioned to the "Duostat," R4. It remains to calculate the values of Ri, R2, and R3. To do this, we must consider the probable plate currents through these resistors. A 201 -a type tube, properly biased, draws the following currents: r. f. amplifier, 90 volts, 4 milliamperes; detector, 45 volts, 1.5 milliamperes; a. f. amplifier, 90 volts, 2 milliamperes. 20,000 Q To B + Max. on B Device and + B Max. on Set ;r2 + 90V. ;R3 1-45 V. -B(+C) R4 — O -C(1-12V.) ■C (12-40 V.) We shall consider a receiver which comprises three r. f. tubes, detector, and two audio tubes, the last of which should be a power tube functioning directly from the highest voltage. So the first five tubes are all we have to consider now in designing the resistor bank. Refer once more to Fig. 4. The current through Ri is 10 milliamperes loss current, plus 12 milliamperes r. f. current, plus 1.5 milliamperes detector current, plus 2 milliamperes 1st audio current, or a total current of approximately 26 milliamperes. The voltage drop is 200 volts less 90 volts, or 1 10 volts. By Ohm's Law, the value of the resistor will be: 1 10 X IOOO Ri= 26 -=4200 Ohms R2 carries a loss current of 10 milliamperes plus the detector plate current of 1.5 milliamperes. The voltage drop is from 90 volts to 45 volts. The calculation here is: — =3900 Ohms 1 1. ? Only the value of R3 remains to be calculated. As the sum total of all the other resistors is 10,100 ohms, R3 obviously should have a value of 9900 ohms — or, say, 10,000 ohms. The resistors should always be capable of dissipating sufficient wattage. The wattage is equal to the voltage drop across any particular resistor times the current in milliamperes divided by therefore is: Wi = FIG. 3 1000. The wattage in Ri 1 10 x 26 _, =2.86 be used, thus To B on B Device FIG. 4 IOOO A five-watt resistor should insuring perfect safety. Similarly Rj should be a 1 or 2-watt resistor, and R3 the same. In the majority of cases the reader will not have to make the calculations outlined above, for the resistor bank „hown in Fig. 5 is well adaptable to the average B power device and receiver. The resistor bank is merely connected to the old B power unit at the high-voltage and negative posts. No other posts on the B device are connected, the various voltages now being drawn from the new resistors. John J. Willis Batavia, Illinois. STAFF COMMENT \A R. WILLIS suggests a simple way of ob' . taining various B and C voltages from B power device other than those for which the device was originally designed. No changes whatever are made in the B unit itself. Thisarrangement, however, is practical only in the case of B device supplying voltages of 220 or more with a drain of 35 milliamperes, since the obtainment of C bias will reduce the voltage of the maximum B tap by an amount equal to that of C voltage. Thus, a device ordinarily capable of delivering 220 volts will still deliver sufficient voltage (180 v.) for best operation of a 171 type tube after a reduction of 40 volts is made to provide for C bias. The output of your present B device can be determined from the manufacturer's specifications accompanying it. The resistor unit should be bypassed with 1mfd. condensers, as indicated by the dotted lines in Fig. 4.