Radio Broadcast (May 1929-Apr 1930)

RADIO BROADCAST on the supply has been removed. The effect on the voltage at the remaining tubes is the same as that which would be obtained if part of the total load were removed by taking tubes out, or by any other means. The voltage rises, to a degree that depends upon the percentage of load removed and the regulating effect of the various drops in the supply system. In battery-operated sets, the rise will be very slight when the B batteries are fresh, but may be noticable if they are old. The degree of rise in voltage which occurs in socket-powered receivers is roughly a function of the cost of the supply system. The cheaper it is the higher will the voltage rise. When the thoughtful serviceman finds the condition described, he knows that the only trouble he has to find is an open. When more load is added to the normal load of a supply, the terminal voltage at the load is lowered. If a test for Ep at one branch of the total load does not show voltage, but the voltages at other sockets are lower than normal, there are just two possible causes of that lowered voltage. The most probable one is that the load on the supply has been increased, by the breaking down of the by-pass condenser across the section of the supply which does not show voltage. The other possible cause is partial failure of supply, which, in socket-powered receivers, would mean failure of the rectifier tube in ninety-nine out of a hundred cases. If that were the cause, lack of any voltage at the socket or sockets of one branch of the load would mean a coincidental open in that branch, having no bearing on the rectifier trouble. If the rectifier is thermionic, serious overloading will often — but not always — be visually evident by the fact that the plate is red. Gaseous rectifier tubes do not show that evidence, and as the overload caused by a broken down by-pass condenser may not be sufficient to redden the plate of a thermionic tube, the best way to be sure about it is to disconnect one side of the suspected condenser. If the trouble were there, voltage conditions would then return to normal. If the by-pass condenser is connected within the set proper, instead of directly across the voltage divider within the power pack, overload caused by its breaking down can be determined by measuring the total plate current with a milliameter in the minus-B lead between the set and power pack. The point to be emphasized, however, is that the evidence gained at the sockets themselves, without going further, is ample justification for the assumption that the sole trouble is the breaking down of that particular by-pass condenser. The alternative of an open in the part of the load circuit which shows no voltage, and a poor rectifier tube at the same time, is a rare .coincidence in actual practice. Assume that Ep tests at the sockets do not show voltage at any of them. In a battery-operated set it can mean only one of two things. The most probable one of those two is an open in the minus-B lead. The other, which occurs rarely, is totally dead B batteries. In his sixteen years of experience with them, the author has not seen more than a dozen times three, or even two, totally dead B batteries connected to a radio receiver. In a socketpowered receiver, the condition can mean an open in the minus-B lead, in the plus-B lead somewhere between the rectifier and the voltage divider, in the rectifier or the transformer secondary winding, or it can mean a broken-down filter condenser. The latter is far more frequent. The question of whether the trouble is an open or a short across the filter system through one of the condensers, can be determined often without further tests. Sometimes the shorting path does not immediately become a continuous metallic one, but is a practically continuous flash-over between plates within the condenser, which is clearly audible as a crackling noise. If the rectifier is thermionic its plate or plates will usually become red, when a dead short exists. If the tube is the gaseous type, a short across its output will sometimes be evident by accentuated humming of the transformer and also noticeable humming of the tube itself. The plates of a gaseous type rectifier tube do not get red hot even with the tube passing as much as an ampere of current. Tests for Short Circuits If none of those evidences of a short exists, then actual circuit testing becomes necessary. The easiest and quickest way to decide the question is to put a 4.5volt C battery in series with a voltmeter across the output bf the filter, from the negative to the positive side, with the power off. If a full or practically full reading is obtained, a short must exist. If a low reading is obtained it means that the circuit is completed only through the resistance of the divider. If no short exists, the trouble must be an open. If the C-battery test shows the shorted condition, then each condenser must be tested individually to determine which one is the offender. That may be done by disconnecting one side of each in turn. When one has been disconnected, turn on the power and test for voltage at the filter output. If approximately normal voltage appears, the disconnected condenser is obviously the bad one. The condensers may be further tested by subjecting them momentarily to the voltage supplied at the place in the filter where they are normally connected, and then shorting them with pliers or screwdriver. If a good fat spark occurs, the condenser is not broken down. If it is suspected that the condenser is leaky although not entirely broken down, charging it and letting it stand for ten. minutes or more before shorting it will answer that question. If the average filter condenser holds practically its full charge that long, it is in good condition. If it does not hold its charge that long, it should be replaced. Abnormal 7/ through any tube, with normal or low Ep and normal filament voltage, is a plate-circuit manifestation of trouble which may, or may not be an actual plate circuit trouble. In any set, it means insufficient C bias. In batteryoperated sets, and practically all sets using external B-power units, the trouble is exclusively a grid circuit one. But in all socket-powered receivers — with the exception of series-filament jobs and a very few others — the C-bias resistor is actually in the plate circuit in addition to being in the grid circuit. If that resistor is shorted there will be no C bias and the If will go up. If it is open there will be no lP at all, a case which we shall discuss further, later in this article. In battery-operated sets, if E^tests at all sockets show slight but irregular fluctuation accompanied by hissing and frying noises from the loud speaker, and the voltages are low, it is conclusive evidence, in most cases, that the batteries have outlived their useful span and must be replaced. If the same fluctuation and noise occurs when the voltage shows the batteries to be fresh, then the evidence points either to a varying resistive joint in the minus-B lead, or a defective B battery. It is fairly common to find brand new B batteries that are noisy, or have open sections, among the very cheap makes, but rare among the good makes. It never is economy to buy or use anything but the best batteries obtainable. If Ep fluctuation is observed at one socket, but the voltage at others is steady, that set of conditions is valuable evidence. If the fluctuation is at an r.f. tube socket, it points to a resistive joint as the most probable trouble. It rarely means trouble in the primary of the following r.f. transformer. If the fluctuation is at the detector socket or an a.f. socket, or an i.f. socket of a "super" with iron-core transformers, it points to the primary winding of the following transformer, or in rare cases to the loud speaker coils if no output transformer is used. In most cases loud speaker coils open completely when they do go, without any interval of intermittency. Transformer Breakdowns The study of the causes of breakdown in iron-core i.f. and all a.f. transformers is an interesting one. It is too lengthy to discuss in this article, but an excellent The up-to-date radio test panel and work bench pictured above was designed and constructed by G. R. Prell, service manager for the Sou thern General Electric Supply Company, Oklahoma City, Radiola distributers. • JUNE • 1929 • • 97