Radio Broadcast (May 1928-Apr 1929)

RADIO BROADCAST lto5 mfd. V.T. Voltmeter § 1000 ohm t Potentiometer : ! // — — ' 1 tr Fig. 2 — Test circuit for trap tuning -r/W5> VL = Vo Zl+LZs (A) v. -3 *t: ^ Series lm O pedenceZs Shunt lm-' lis ; pedenceZL (B) Fig. 3 — Method of calculating voltage reduction. (A) Actual circuit, (B) Equivalent circuit. an annoying "yoop" when keyed with a "brute-force" filter. The drop through the choke was 175 volts. Filter circuits such as shown here cured this almost completely without any change in the r.f. circuits. TUNING THE TRAP The question will naturally be asked, " How am I going to be able to tune the trap without some sort of analyzer?" Fortunately this is simple. If you have, can make, or will borrow, a vacuum-tube voltmeter, connect it to the output of the loaded filter as shown in Fig. 2 and juggle the trap condenser until the vacuum-tube voltmeter shows a minimum. The 1000-ohm potentiometer is used as a sensitivity control and enables you to keep the meter deflection within reason. The isolating condenser Cs must be capable of withstanding the full voltage of the output. Always do the testing with the rectifier and filter operating at their normal load as the inductance of the chokes vary with the d.c. flowing through them. If a vacuum-tube voltmeter is not available, a pair of phones may be used and the filter adjusted for minimum sound. Leave the phones on the table; the condenser Cs might puncture and electrocution is reported to be unpleasant. The rectifier shown is of the "centertap" variety. If the "Bridge" connection is used for any reason a larger 60-cycle voltage may appear and may call for a 60-cycle trap in addition to the 120-cycle trap. Such a trap will require about 4 times the capacity (in the trap circuit) as is necessary for 120 cycles. No tests were made on a half -wave rectifier but in this case also a 60-cycle trap is necessary, although the 120-cycle trap may probably be omitted. Incidentally, in the fullwave rectifier discussed some 60-cycle voltage appears because of an unbalance. If for any reason it is objectionable a 60-cycle trap is the cure. Some measurements of the harmonic currents flowing in various parts of the circuit were made by inserting a resistor in the circuit and measuring the alternating voltage drop across it. The 120-cycle current flowing in the branch C3 of diagram c, Fig. 1, was 120 niilliamperes. This is practically the total 120-cycle output current of the rectifier, since the 120-cycle current flowing through the trap to the load was only about 4.7 milliamperes. From this latter current and the voltage across the trap it is easy to calculate the impedance of the trap to 120 cycles: 0.53 I 0.0047 = 11,300 ohms. The impedance of do to 120 cycles is 132 ohms, hence we should expect a 120-cycle voltage at the load of + " =0.64 volts. The voltage measured was 0.60, a fairly close check. The method of calculating the last voltage is explained in Fig. 3. The combination of series and shunt impedances may be considered as a potentiometer, which permits a fraction of the impressed (or rectifier) voltage to appear across the load. The ratio of these two will be the ratio of the total impedance, Zs + Zi, to the shunt impedance, Zi_, which latter is made up of the load and the last condenser, in parallel. By varying these impedances we can make the a.c. voltage reduction in our filter anything we please. For instance, suppose that the final condenser, Ci0, is reduced to 1 mfd. Its impedance will then be 1320 ohms, and the 120cycle voltage appearing at the load will be (neglecting load impedance) 53 x 1320 11,300 + 1320 = 5 5 V°ltS This simple way of looking at filter action permits you to calculate quickly what reduction of voltage you may expect from a given set of impedances. The impedance of a choke coil is given approximately by Z = 6.3 x (inductance in henries) (frequency) The impedance of a condenser is given by z= 160.000 (capacity in mfd.) (frequency) The impedance of the trap was discussed before. Of course, the method of calculating drops does not work out if any choke is in resonance with one of the harmonic frequencies by reason of its distributed capacity or if one of the chokes and one of shunt condensers makes a series resonant circuit. COMMENT AND SUGGESTIONS The difference between this filter and the "brute-force" type is that it does a very neat piece of work for the particidar conditions to which it is adapted, while in the "brute-force" case we squirm out of the necessity of adjustment at the expense of tolerating rather bad voltage drop and rather large expenditure for apparatus. Probably everyone has found it out before now that choke inductances and paper condenser capacities usually vary considerably from the marked values so that a little cutting and trying is useful. This means changing the condensers around and adjusting the air gap of the choke. A 1.2-HENRY CHOKE In Fig. 4 are shown a pair of suggested chokes which can be varied from about J to about 1.2 henries by changing the air gap. In both forms (a and b) the core legs are 1 inch square and butt joints are used throughout. The air gap is adjusted by laying different thicknesses of cardboard in it after which the core is again clamped or taped to reduce humming. About If pounds of No. 24 enameled wire will fill the winding space of a while No. 23 enamel may be used for b. The chokes will handle about | ampere without Table I ANALYSIS OF FILTERING ACTION OF CIRCUITS SHOWN IN FIG. 1 3k- -1 -3Js" ■*1 Frequency Voltages at Load A B C D d.c. 530.0 750.0 750.0 750.0 / 60 25.0 7.4 1.6 1.1 / 120 245.0 53.0 0.60 0.3 Nl 180 1.2 1.06 0.01 0.01 Ol240 17.6 14.2 0.90 0.075 T \ 1300 0.16 w A (360 12.5 1.9 0.13 0.01 N \ T 1480 7.5 1.0 0.075 E / Df 600 6.7 0.67 0.05 I 720 3.5 0.151) 0.02 1 840 1.25 0.19 0.15 \900 0.51 0.085 A blank means that the voltage was less than ijij -Air Gaps -± ? -Butt ! Joints 1 volt. t.-1800 TurnWindmg....v|||| (A) <B> Fig. 4 — Details of an efficient filter choke. (A) Usual form, (B) Simplified form u-ilh all legs alike. saturation and may, of course, be wound for any other inductance, keeping in mind that this changes approximately with the square of the number of turns and that the ampere turns must not exceed 800. Unlucky New England IT HAS been said that if a set will work through the static at New Orleans it will work anywhere except in the Caribbean. To this let us add that if a receiver will give a good account of itself in southern New England it will be an amazing success elsewhere. Here at Hartford we sit in the world's largest nest of high-power radio stations and with rather decent receivers hear materially less than does the Kansan with the most ordinary of home-made apparatus. Weaf is not amazingly strong here, though only 90 miles away over seemingly favorable country, and this is the cause of disparaging comments from our Kansas relatives, who, at 1000 miles, are accustomed to hear the station comfortably in daylight. Wgy's high-power commissiondefying transmitter fades horribly and is of little use to us, while wjz fluctuates in a ratio of as little as or as much as 5r§-Tj if one speaks of audibilities. As to other stations, and other types of signals, matters are not materially better. Television and rayfoto signals have a rather terrible time of it. Altogether one is not surprised that the American Radio Relay League was organized here to relay messages over the 27 miles to Springfield while at the same time amateurs in Illinois, Ohio, Kansas and Missouri conversed freely and easily over distances well up toward 700 miles. From Here Forward OUR time for this month is up. The Editors are leaving the nature of the future material to our judgment. This is not as reckless as it seems — for the copy passes under the blue pencil before reaching these pages! Since this encouraging "department" was suggested a pleasant number of letters have arrived, bringing material and comment of the most varied sort. We can, so to speak, take from stock a conversation on vacuumtube voltmeters, variable a.c. supplies, an extremely queer short-wave effect, some ingenious tuning devices, or several other things. These letters assure our immediate future but they are not a "preferred list." Whatever you have done, are doing, or may be planning is of interest. Do not be too serious about it, this being neither a literary society nor the Franklin Institute. The main thing is to learn of each other's doings, whether they concern a simple mechanical makeshift, an odd effect or large plans for investigation of one of the two score unsettled problems. Letters — or informal papers — will be welcome. They should be addressed to Bobert S. Kruse, care Radio Broadcast, Garden City, New York, and should refer to these pages by title. • april, 1929 page 381