|
How the wheels go round in the Variable-Selectivity IV.The voltage and current distribution in modern receivers is not always clearly understood. In this article it is shown how it may be investigated with quite simple measuring apparatus. The receiver chosen as an example is one which, as far as the direct current circuits are concerned, is typical of practice.
A view of the Variable-Selectivity IV in which the chief components can clearly be seen.
It is always interesting to know exactly what is happening in every part of a receiver, and it is not difficult to find out in most circuits with the aid of quite simple testing equipment. It is especially important when a receiver is not functioning, for a comparison of what is happening with what should happen usually leads to the cause of the trouble in a few minutes. It is common knowledge that current and voltage tests are a very necessary part of all testing, but it is not always realised that there is a right and a wrong way of making them, and, of course, the readings must be properly interpreted if they are to be of any value.
Fig. 1. - The complete circuit diagram of the Variable-Selectivity IV. The current through the speaker field can be measured by inserting a milliammeter at the point indicated by a dotted ring. The Valves are:- X41, VMP4G, DN41 & MU12.
It is proposed, therefore, to describe in detail what should happen in every circuit of a typical modern receiver, and to show how tests may be applied to prove that events are really as described. The Variable-Selectivity IV has been selected to form an example, not only because it is representative of modern technique, but also because it is an extremely popular set. The complete circuit diagram is shown in Fig. 1, and it is proposed first of all to deal with the purely direct current aspects. Taking the receiver first of all with no applied signal, and for the time being, taking for granted that the mains equipment supplies and maintains a potential of 200 Volts across C24, it is easy to see that the current flowing in any circuit is dependent only on the total resistance of that circuit. This resistance includes not only the actual resistances employed, but, in addition, the resistances of other components and of the valves themselves. The valve resistance, moreover, is not a constant, but depends on the voltages applied to all the electrodes.
Fig. 2. - The circuits essential to the operation of the pentode section of the output valve are shown by heavy lines, while the detector circuits are indicated by light lines. In the case of the AVC circuit, dotted lines are used.
Fig. 2 shows the circuits of the output valve, which is a duo-diode-output pentode three valves in one. The purely pentode circuits are shown as heavy lines, the detector circuits as light lines, and the purely AVC connections dotted. The cathode of the valve emits electrons which are collected by the positive anode and space charge grid and constitute the current taken by the valve. This current flows from cathode to the positive electrodes in the valve, and thus from the negative of the HT supply to the positive.
Starting at the negative HT terminal in Fig, 2, therefore, the total current taken by the pentode (the cathode current) flows to the valve through R12 and R11. The current divides in the valve and a portion of it appears in the space-charge grid circuit, whence it flows straight to positive HT. The major portion of the current, however, appears in the anode circuit and flows to HT through R10 and the primary of the output transformer. As will be seen later, the control grid takes no current.
The Output Pentode
Now the anode current is the current which flows from anode to positive HT and the screen (or space-charge grid) current is that which flows from the space-charge grid to positive HT. The cathode current, however, is the current which flows from negative HT to cathode, and assuming that the other electrodes draw no current it is equal to the sum of the screen and anode currents. These currents can readily be measured by connecting a milliammeter in series with the various circuits, as shown at X, Y and Z, for the screen, anode and cathode currents respectively. In this case they should be 6.75 mA, 26 mA, and 32.75 mA.
The presence of the correct currents is prima facie evidence that everything is in order, but it is not proof, for they depend upon the voltage, and it is possible that normal currents might be obtained even with low anode and screen voltages if the grid bias also happened to be abnormally low. Now, care must always be exercised in the measurement of voltage, for the presence of the voltmeter changes the voltage in the circuit in some degree. With high-quality voltmeters of high internal resistance, this effect is usually small, but it can only be ignored when the resistance of the meter is high compared with the circuit across which it is connected.
The true voltages on the valve are those existing between its electrodes. The true anode voltage can be measured with a voltmeter joined between anode and cathode (B and C, Fig. 2) and the screen voltage by connecting the meter between screen and cathode (A and C). The anode voltage is less than the screen voltage. because of the resistance R10 and the resistance of the primary of the output transformer. In this case, a voltmeter between anode and cathode will read 160 Volts, but there is 185 Volts between screen and cathode. It is easy to see that 25 Volts is lost between A and B; from Ohm's Law (current (mA): 1,000 × Volts / Resistance (Ω)) the resistance between A and B must be 960 Ω, and as R10 is 100 Ω, the transformer primary must be of 860 Ω resistance. This can be checked by connecting an ohmmeter, with set switched off, between A and B; it should read about 960 Ω.
Now, in the cathode circuit, a voltmeter connected between C and E will read 20 Volts, the cathode potential above negative HT. The cathode current is 32.75 mA, so that the resistance should be 610Ω; the values of R11 and R12 actually total 600 Ω. This agreement is very good when it is remembered that the resistances may vary as much as ±15%, and no correction has been made for the loading of the voltmeter. In this case, the voltage drop across R12 serves no useful purpose as far as the pentode is concerned, and the voltage (16.375 Volts) across it has no effect on its operation. The potential across R11, however, of 3.275 Volts is actually the grid bias of the pentode, for the grid is returned to the point D through R9 and R8. It is not possible to measure the potential actually existing between grid and cathode with any ordinary voltmeter, however, for the grid circuit resistances are too high for any reasonably accurate indication to be obtained. It is quite easy to check that the bias voltage developed across the bias resistance R11 is reaching the grid of the valve by connecting a milliammeter in the anode circuit at Y and then connecting the grid directly to its source of bias - that is, short-circuiting the points DF. There should be no change of anode current.
The Detector
It must be realised that any marked difference from the correct figures that is obtained must be due to a defect, and from the particular differences which are noted it is usually a simple matter to locate the cause. Small variations are to be expected, however, for no two components and valves are exactly alike. While dealing with the question of voltages, it may be as well to remark that the voltage figures published in the constructional article dealing with this receiver are all measured from negative HT (chassis), since this is often more convenient than measuring the true anode and screen voltages. These true voltages are equal to figures headed, anode volts and screen volts if the cathode voltage of the particular valve is deducted from the latter.
Turning now to the detector circuits, current and voltage tests are of little use, for no voltage is applied from the HT supply, and a sensitive micro-ammeter would be necessary to detect the small current flowing through R6 as the result of the rectification of a signal. This applies also to the AVC circuits, for although there should be a potential of 20 Volts between G and C, it cannot be measured with an ordinary voltmeter on account of the high values of R13 and R14. Certain tests are applicable, however, and will be dealt with later.
Turning now to the IF stage, the anode is fed from positive HT through the primary of the transformer T2, but as the resistance of this winding is quite low the voltage at the point A is practically the same as that at B. The anode current of 6.85 mA passes through this winding. Now the screen requires a much lower potential than the anode, so that it is fed from a tapping on a voltage divider which is connected across the HT supply. The grid bias is obtained by the voltage drop across the cathode resistance R3. The matter is complicated by the fact that the currents consumed by the frequency-changer enter into it. In so far as they affect the present matter, they are marked on the diagram of Fig. 3.
Fig. 3. - The heavy lines show the main circuits of the IF valve.
The Voltage Divider
Measuring screen-voltage of the IF valve.
A voltmeter connected between F and E indicates a potential of 70 Volts between these points; as R18 has a resistance of 10,000 Ω there should be a current of 7 mA through this resistance, and this can be checked by connecting a milliammeter in series with it at X. When the voltmeter is connected across C E, however, it reads 100 Volts; there is consequently 30 Volts dropped across R17, and the current through this resistance should equal the sum of the current through R18 and the current drawn off at the point F, or 9.25 mA. This current can be measured by inserting a milliammeter at the point Y. The value of R18 can be evaluated from this data and is obviously 30 × 1,000/9.25 = 3,254 Ω this resistance is assigned a value of 3,500Ωand the calculation shows it to be somewhat lower, but well within the usual tolerance.
The screen grid of the IF valve passes a current of 4.25 mA, as can be seen by including a milliammeter at Z. The current through R16 is thus this current plus that flowing through R17, a total of 13.5 mA. A voltmeter across A and E registers 205 Volts, but connected between C and E it shows 100 Volts; the drop across R16 is thus 105 Volts and its resistance should be 105 × 1,000/13.5 = 7,474Ω. As the rated value of R16 is 7,500 Ω the agreement is remarkably close.
The anode current of the VMP4G can be measured by connecting a meter at the point W, not in the more convenient anode lead to the valve, since this would be likely to cause instability and this would in turn cause a change in the anode current. The total current consumed by the frequency-changer can be read on a meter joined in the cathode lead of this valve (V); it is 5.25 mA. The current through R3 is the sum of the screen and anode currents of the IF valve plus the current taken by the frequency-changer, or 6.85 + 4.25 + 5.25 = 16.35 mA. The bias can be read on a voltmeter joined to D and E and is 2.45 Volts; R3 should thus be 2.45 × 1,000/16.35 = 149.8 Ω, which is negligibly different from the nominal value of 150Ω.
The grid of this valve receives its bias through the secondary of T1 and R14, but the measurement of the bias directly between grid and cathode is just as impossible with ordinary equipment as in the case of the pentode. The connection of the grid directly to E, however, should in the absence of a signal cause no change in the anode current if all is well.
Fig. 4. - The circuits of the hexode section of the frequency-changer are shown by light lines, while those of the triode oscillator are indicated by heavy lines.
The frequency-changer can be dealt with in exactly the same way as the other valves, but is slightly more complex because of the oscillator section. The X41 is really two valves in one, and it is best to deal with the oscillator first, since its operating conditions affect the currents passed in the hexode section. The converse is not necessarily true, however. The oscillator anode current can be read on a milliammeter connected at X in Fig. 4, and it depends not only on the applied voltages but also on the amplitude of oscillation. It is likely to vary somewhat over the tuning range, therefore. when the valve is oscillating correctly the current should be about 1.7 mA, but if the valve be prevented from oscillating by short-circuiting terminals 4 and 5 of L5, the current will rise. If it does not, then the valve is not oscillating.
Measurements of voltage between AE and BE show respectively 205 Volts and 40 Volts. The drop across R5 thus appears to be 165 Volts, and as the current is 1.7 mA, R5 should be 165 × 1,000/1.7 = 97,000Ω. Its actual value is 75,000 Ω, however, and the discrepancy seems rather large. It is actually due to the resistance of the voltmeter which becomes important in this circuit because of the high value of R5. Actually, the anode voltage is greater than 40 Volts with the meter absent and the current through R5 is greater than 1.7 mA with the meter present. Assuming R5 to be really 75,000 Ω, the true anode voltage is easily calculated and is 205 - 74,000 × 0.0017 = 77.5 Volts. To be quite exact the grid bias of 2.45 Volts developed across R3 should be subtracted.
The Oscillator
The change of anode current between the oscillating and non-oscillating conditions forms a reliable test when there is any doubt that the valve is oscillating. The amount of the change, however, is no reliable guide to the amplitude of oscillation, and the efficiency of the frequency-changer depends very largely upon this. Fortunately, it is possible to measure the amplitude with reasonable accuracy by quite simple means, although a fairly sensitive milliammeter is needed. The HF potentials developed on the grid of the oscillator are rectified in the grid circuit by the normal action of a diode detector, and a steady current consequently flows through R4 in such a direction that the voltage set up across R4 by the passage of the current holds the grid at a potential negative with respect to the cathode.
The magnitude of this current is very nearly proportional to the amplitude of oscillation, and is approximately equal to the amplitude in volts divided by 1.2 times the resistance of the grid leak (multiplied by 1,000 for current in mA). The X41 requires an amplitude of 12 Volts, and as the leak resistance is 50,000 Ω the current through it should be 12 × 1,000/1.2 × 50,000 = 0.2 mA. This can easily be read on a meter giving a full-scale deflection for 1 mA, inserted at the point Y of Fig. 4, but is difficult to read accurately with a less sensitive instrument. If the receiver is tuned over the waveband with this meter in circuit it will be found that the current falls somewhat at the high wavelength end of the band. This is because of the change in the L/C ratio in the oscillator tuned circuit. On the long waveband the current is only about 60% of that on the medium, because the same reaction coil is used for both wavebands. The X41 is not at all critical as to the oscillator voltage, however, and quite wide variations from the optimum of 12 Volts are permissible without any important effect on the efficiency. It should be noted that these grid current variations of the oscillator are reflected in the anode current a fall in grid current corresponding to a rise in anode current.
The circuits of the hexode section of this valve are shown in Fig. 4 by the light lines. The anode voltage measured between A and E is 205 Volts, but the true anode potential between A and D is 205 - 2.45 = 202.55 Volts on account of the drop across R3. The difference, however, is negligible, as is also the drop in the transformer primary T1. The screen potential of 70 Volts is obtained from the voltage divider at the junction of R17 and R18, as already described. The control grid is returned to the earth line through a complex network comprising at first L3 and L4. There are then two paths, one through L2, L1, and R1, and the other through R2, R15, R13, and R14. In the absence of a signal, the grid should be at chassis potential, and this can be checked by short-circuiting the grid to the chassis and noting whether this causes any change in the anode current read by a meter connected at Z.
All these tests apply when there is no signal. When a signal which is strong enough to operate the AVC system is present, the conditions differ considerably and the distribution of voltage and currents is changed. When AVC is operating, the anode of the AVC diode becomes negative with respect to the earth line, and this potential is communicated through a resistance-capacity network to the control grids of the X41 and VMP4G valves, It is unnecessary to enter here into the precise arrangements of this distribution: for it cannot be checked with any ordinary meter on account of the high resistance of the circuits.
On tuning in a strong signal, however, the grids of the controlled valves receive an increased negative bias, and as a result their anode and screen currents fall. In the case of the X41, however, the change is much less than in the VMP4G, and as the oscillator anode current may rise slightly the total cathode current of the X41 may change very little. In any case, however, the total current through R3 falls and also the voltage drop across it. Because the screen currents fall, the voltage drop across R16 and R17 falls, and the screen potentials rise; owing to this rise in voltage there is a slight increase in the current through R18. The total current consumption of the receiver falls somewhat, however, and the HT voltage consequently rises, with the result that the output pentode will pass a slightly higher current and the voltage drops across the resistances in its cathode circuit will increase. The importance of checking a receiver without a signal will thus be appreciated, for it is easy to see that very misleading results may be secured if the set is tuned to a signal. It should, of course, be understood that some of the current and voltage changes which actually occur are of negligible importance and may be barely detectable by meter. The only important changes in current are those occurring in the valves controlled by the AVC system.
By totalling the figures already given, the total current consumption of the set can be computed; it is:-
The total current can be measured by inserting a milliammeter in series with the speaker field at the point marked with a dotted circle in Fig. 1. The measured current is found to be 56.5 mA; the difference of 0.4 mA may well be the leakage current of the electrolytic capacitor C24, but is, in any case, unimportant, since it is so small. It often occurs that the discrepancy is greater, and quite a large difference can be assigned to errors in the meter and errors in reading it.
The HT voltage of the receiver is measured across C24, and is 205 Volts, as already found, and the output of the rectifier can be measured across C25 it is 350 Volts. The difference of 145 Volts represents the drop across the speaker field, and as the current is 56.5 mA the resistance of the winding is 2,563 Ω - good, agreement with the rated value of 2,500 Ω.
There are, of course, many other tests which can be applied when a defect occurs in a receiver, but many of them do not show up so well just what is happening in the various circuits. For fault finding, resistance tests are often as useful as voltage and current readings, and in some cases they are better. It would not, however, be interesting to describe such tests in detail, since they consist merely of checking the resistance between various points in the circuit with an ohmmeter. A table is appended, however, giving the chief values of resistance between various points, all readings being taken with the set switched off.
|
