Radio Broadcast (May 1928-Apr 1929)

154 RADIO BROADCAST JULY, 1928 FIG. I A picture diagram showing exactly where the apparatus goes and how to connect it socket 90 degrees in a clockwise direction (when viewed from the front). Such change will make a better arrangement of the high-frequency leads which should be well separated to minimize their mutual capacity. It is common practice to use a 250-mmfd. condenser to control regeneration. The reason for this is not quite clear since only half of the capacity is normally used. The amount of capacity necessary will vary with the layout and with the tube used but for the receiver described 100 mmfds. was found to be ample. For this reason a "midget" condenser was used resulting in a saving of several dollars. At the point where the condenser is used, that is, near maximum capacity, the rate of change of capacity per degree of rotation is lower than that for the usual logarithmic type of plate so that control is very good. THE J I2A AS A DETECTOR A112A type tube is recommended. This has a good detection coefficient, high mutual conductance and requires no more filament current than the 201 a type. It is a better oscillator even at high frequencies. The calibration curves (Fig. 2) were made with the antenna very loosely coupled. No universal calibration curve could be plotted with close coupling since the antenna characteristics theh affect the tuning. The calibration also varies slightly with different regeneration condenser Wenstrom's "The Cornet Multiwave Receiver" in the June Radio Broadcast, page 77. — Editor.] A standard set of three coils used in this receiver covers the bands between 17 and 107 meters. Other coils of the same type may be secured which cover the bands from 107 to more than 200 and from about 82 to 17 meters. There are several respects in which the design and construction of a short-wave regenerative detector circuit differ from the type used at broadcast frequencies. The principal problems arise from the difficulty of securing good, stable regeneration at the very high frequencies, say more than 10 megacycles (less than 30 meters.) This difficulty is due to a number of factors. To secure close tuning and more particularly to tune to the high frequencies with a reasonable value of inductance the variable capacitances must be small. This means that the stray circuit capacitance and the inter-electrode capacitances of the tube itself may be of the same order as those used in tuning. Dielectric losses increase with the square of the frequency and these become important at high frequencies. In many cases the advantage of "debasing" the tubes lies not so much in a reduction of the inter-electrode capacity as in a decrease in the dielectric losses. If a radio-frequency choke is to be a "universal" one it must be designed to have capacitative admittance, that is, act as a capacity at all the frequencies for which it is used. This follows from the fact that if it were inductive and used in a shunt plate-feed circuit this load would make the circuit unstable. In a short-wave receiver a tremendous frequency range must be covered and this makes it difficult to insure a high impedance at all frequencies. Since most of the frequencies to be received are very high, a low-capacity grid-condenser may be used and a high-resistance leak is then used to increase the "time constant" of the gridleak and condenser combination and therefore the sensitivity. In this receiver a neutralizing condenser of the "fixed-adjustable" type is used. This has a maximum capacity of 40 mmfds. At very high frequencies, the effective input capacity of the tube which shunts the tuning condenser is high and this decreases the maximum frequency that can be received. This may be offset slightly by reducing the grid condenser capacity at these frequencies. The reason for this lies in the fact that the grid condenser and the tube capacity are in series. The only point to watch in the construction is the mounting of the tuning condenser. The vernier dial has a sleeve which goes through the panel and this makes it necessary to mount the condenser itself on small bushings. The layout is clearly shown in the photograph. It may be slightly improved by rotating the COIL DATA TABLE Wavelength Sec. Space Tick. Wire Turns per inch Diam. 20 3* 1 3 No. 16 10 3" 40 7 1 4 No. 16 10 3" 80 16 2 6 No. 16 10 3" 160 32 3 15 No. 20 21 3" Primary 7 turns No. 18 18 2\" In the photograph accompanying this article, the spacing of the turns is clearly indicated. The information above is given for those who prefer to make their own coils and should be sufficient for the purpose. The copper wire used may be of any kind the builder may have at hand. > < Nl <^KDKAEi pcrimenlal WGY Exp erimental . PC J J ■ Holland — i-r Jam ■England CONDENSER DEGREES FIG. 2 Mr. Knowles's unit tunes as shown on this curve. Different tubes and slightly different wiring will naturally change these curves somewhat settings but this is not serious. Different types of tubes also alter the calibation due to the difference in effective input capacity. To use this set with the a.c. R. B. "Lab" receiver, the following should be done: Break the bulb and remove the elements from a ux or cx type "dud" tube. Remove the leads by heating the tips of the prongs with a soldering iron. Solder a lead to either of the large prongs. Hold the socket base in an inverted position with the large prongs, which normally go to the filament, toward you. The left hand small prong is then the one which goes to the P terminal and a lead should be connected to it. This lead is connected to the binding post which is connected to the choke, that is, post 2 on Fig. 1 and 3. The other lead goes to the binding post next to it, Post 1 on Fig. 1 and 3. If the unit is to be used with a Lab Circuit receiver whose filaments operate from batteries the posts marked A minus and A plus may be attached to leads which may be cabled with the leads from posts marked 1 and 2 and of course attached to the proper filament prongs of the plug which goes into the receiver's detector socket. Thus all connections to receiver and batteries will be made when the plug is in the correct socket. Unfortunately the a.c. tubes modulate signals too much to make it possible to use them in an oscillating receiver for code reception. With such connections utilizing the A