This was Suggested Circuit 201 and was originally titles Simple DC Heater Supply for AF Amplifiers but relates to DC for the first voltage amplifier stage.
By means of this simple rectifier circuit, the first valve of a sensitive audio amplifier may be heated by direct current (plus a small ripple) whilst the remaining valve heaters are supplied in normal manner. The two capacitors shown in dotted line may be required in some instances.
One of the major design problems problems with sensitive mains-driven AF amplifiers is the reduction of hum caused by the AC supply. If the amplifier employs valves, hum is particularly likely to result from unwanted couplings between the signal circuits and the heater wiring, and considerable care has to be taken with component and wiring layout to ensure that such couplings are kept to a minimum.
The first stage of the amplifier is, of course, that which is most susceptible to hum, and it is common practice to use a low-noise pentode, such as the EF86, in this position. The EF86 can exhibit a hum level of 1μV only when fed with a centre-tapped AC heater supply, but it is still necessary to pay careful attention to the valve-holder type, valve-holder wiring and component positioning if a figure as low as this is to be achieved in a practical working circuit.
Hum in the first stage due to unwanted couplings with the heater wiring can be completely eradicated if the heater of the first valve is fed with DC instead of AC. Whilst this represents an obvious solution to the hum problem, there is a tendency to assume that the provision of the necessary direct current involves the use of special mains transformers and expensive rectifier circuits, whereupon it becomes economically unattractive. Now that small silicon rectifiers and high-value electrolytic capacitors are available, however, the situation is changed, and the cost of a DC. heater supply need not be as excessive as would have occurred with earlier components.
This 'Suggested Circuit' presents a simple method of applying a DC supply to the first valve of an AF amplifier, the alternating voltage being obtained from a norm-al 6.3V centre-tapped mains transformer secondary.
The suggestion will be of greater interest from the point of view of application than of circuit operation, since the circuit consists quite simply of a standard bridge rectifier plus several trimmings. The circuit has the advantages that few components are required, that a standard mains transformer can be employed, and that the heaters of the remaining valves in the amplifier may be run in normal manner from the same 6.3V winding. It must be emphasised that the circuit does not provide a completely pure direct current, since the voltage applied to the valve heater carries a ripple having a peak-to-peak value of the order of 0.2V. However, this figure is much lower than the peak-to-peak value of an alternating 6.3V heater supply with the result that the circuit can offer a very significant reduction in hum pick-up in the first stage of an amplifier, even though it may not give the complete eradication of hum that would result from a fully smoothed DC supply. With the components specified, the circuit is suitable for feeding the heater of an EF86 or any other valve requiring a heater voltage of 6.3 at a current of 200mA.
The circuit of the DC supply is given in the accompanying diagram and it consists of a full-wave bridge rectifier running direct from a 6.3V heater winding of the mains transformer. For the circuit to function it is essential that the rectifiers have a low forward resistance and that the capacitor following the bridge has a large value. The rectifiers specified for the bridge are silicon types and thereby meet the first requirement, whilst the capacitor is an electrolytic component having a value of 3,200μF.
In order to keep switch-on surge currents to a low value it is desirable to insert limiting resistance between the transformer winding and the rectifiers. This limiting resistance must not, however, be too high or the required rectified voltage will not be obtained. In the present circuit the limiting resistance is 2.4Ω, this representing a compromise figure. With a limiting resistance of 2.4Ω the rectified voltage applied to the heater of an EF86 is 6.1V, and it was considered that this is sufficiently close to the nominal value of 6.3V to be acceptable.
The 2.4Ω limiting resistance is split up into two physical resistors of 1.2Ω, as shown, in order to maintain symmetry It will be appreciated that, on succeeding half-cycles, each of the DC. heater lines becomes effectively coupled first to one side of the 6.3V secondary and then to the other, whereupon it is desirable to keep the circuit symmetrical in order that the DC lines remain balanced about chassis potential. This approach ensures that the 0.2V ripple on the DC lines has minimum effect. In this context it is important to note that the 6.3V transformer winding must be centre-tapped, with the centre-tap connected to chassis. If, instead, one end of the 6.3V winding were connected to chassis, each of the DC heater lines would be taken through some 6.3V on alternate half-cycles and the whole object of the circuit would be defeated!
As is shown in the diagram, the four rectifiers, D1 to D4, are Lucas type DD000, these having a peak inverse voltage of 50 and a forward current rating of 500mA. The electrolytic capacitor employed for C1 in the prototype was a Mullard component in the C431BR series, with a value of 3,200μF and a voltage rating (which should not be exceeded) of 10V. It may be added that the rectified voltage across C1 will rise to nearly 9V should the EF86 be removed from its valve-holder, and this fact has to be borne in mind when deciding upon the working voltage of the capacitor. In general, it would be preferable to switch off the power supply if the EF86 is to be removed and re-fitted, since a slightly excessive current could momentarily flow in its heater if it were plugged in with the supply switched on and C1 charged to peak rectified voltage.
It will probably be necessary to make up R1 and R2 from short lengths of resistance wire, as their value is outside the range normally available for fixed resistors from home-constructor retail sources. Care should be taken to ensure that the resistors have the required value within ±5%.
When the DC supply circuit is fitted to an AF amplifier all the rectifier components should be mounted near the mains transformer. A tightly twisted pair may then be used to couple the rectified voltage across C1 to the heater pins of the EF86. A twisted pair is recommended since the DC supply still carries the 0.2V ripple voltage.
As was stated earlier, the prototype circuit provided 6.1V at 200mA for an EF86 heater using the components specified. The writer does not recommend that the circuit be employed for valves requiring 6.3V at 300mA since the greater current would necessitate reducing the values of R1 and R2, whereupon switch-on surges could become excessively high.
A final point is that the rectified voltage across C1 is 'floating' during the periods in the cycle when the diodes are not conducting and it is feasible that, under some conditions, this could result in a modulation hum being imposed on the AF signal being handled. Should this occur, both terminals of C1 should be coupled to chassis via 0.5μF capacitors, as shown in dotted line in the diagram.