The T.R.F. Set

Tuned Radio Frequency.

A TRF set is the simplest sort of radio to understand. It's simply a case of tuning in the signal you actually want, and then maybe amplifying it a bit, and then coupling this through another tuned circuit to the detector, which changes the radio signal into the corresponding audio signal. This audio signal is then boosted up by a power amplifier valve and fed to the loudspeaker.

Too Many Knobs.

That sounds fine, but the problem comes when ambition leads the designer to attempt to drag in weak and distant signals. More gain is required. And one valve won't give this. So he has to add one or two more, so that even weak stations can give a reasonably decent signal to the detector valve. But if you have all these valves, their circuits all have to be tuned. If a set has two tuning controls it's quite difficult to operate because they have to be kept in step as they are adjusted. If it has more, it turns into a nightmare which takes ages to tune up for each station. While it is possible to join all the tuning capacitors to one shaft and make the whole thing tune with one knob, it will still be necessary to fit small trimmer capacitors so that the various tuned circuits can be kept in step as the tuning is swung across the band. So the radio then becomes a bit of a nightmare to set up and keep tweaked to maximum performance.

This is the reason for the widespread use of the "superhet" circuit, and the widespread use of this sort of circuit is why our info pages have a feature on it.

Keep It Simple.

Having been rude about the simple Tuned Radio Frequency set, if you can keep things reasonably simple and have no more than two tuning controls to juggle, you can make a reasonable radio. And in the 1920s, almost every set made was a TRF of one sort or another.

It is a sobering thought that given even a few feet of aerial wire, it is quite possible to make a simple TRF set with a detector valve, an output valve, and a rectifier valve, which will receive local stations at loudspeaker volume. This simplicity meant that the TRF was ideal for the early days of radio when valves and components were very expensive. In fact, if you fancy building a valve radio for yourself in modern times, there is a lot to be said for starting with a TRF. They are quite simple and they work quite well with a minimum of fiddling around.

A Typical TRF Set.

The TRF was popular in the 1920s, and also in simpler, cheaper sets, right through to the mid 1930s. As an example we have included the circuit of the GEC A.C. Mains 4, which uses the mains for a power supply, and mains-type valves, and includes an RF amplifier stage, but is otherwise much the same as a 1920s set-up with battery valves.

GEC AC Mains 4 Circuit, 1935

The Input Circuit and RF amplifier.

The first job is to get the signal out of the end of the aerial, and to tune in the station required. The tuned circuit is made up of the main tuning coil (L2, with L3 added on LW), and C17. (C18 is a trimmer so that the tuning of the RF amplifier V1 and the detector V2 can be tweaked up to remain in step across the band.) This is applied to the grid of the pentode V1 for amplification. The cathode of V1 is taken down to chassis via a fixed resistor, R3, and part of the variable resistor R4. In this way, all of the current through the valve works with the resistors to make the cathode slightly positive. The extent to which it is positive is controlled by the setting of R4. Since V1 is a vari-mu valve, varying the cathode voltage slightly positive while holding the grid at 0V, means that the bias of the valve is altered, and therefore the gain is changing. So if a strong station is being received the bias on the valve can be made more negative and the detector will not be over-driven. To help with this, the other end of R4's track is connected to the aerial, so as the wiper of R4 is moved to increase the valve bias and cut the gain, the resistance between aerial and earth decreases as well, thereby cutting back the signal input from the aerial, and hopefully ensuring that at minimum volume, there is not much signal from the loudspeaker.

The Detector Stage.

The anode of V1, where the amplified signal appears, is fed from the HT power supply via L4, with L5 switched in on Long Wave. The inductance of these coils is such that the RF signal is developed across them, and coupled into the next tuned circuit, L6 (and L7 on LW) and C19, by C3. C20 is the other trimmer to keep the tuning in step across the band. The tuned signal from here is fed to the grid of V2 via C7. V2 is a leaky-grid detector, and is biassed with the 1 megohm grid leak, R5. This allows the space charge inside the valve to bias the grid a little negatively. C7 prevents this voltage from leaking away to earth through the tuned circuit. Being a mains valve, V2 needs a bit more bias than it can get from just the grid leak, so the cathode voltage is raised positive by R6 and R7. These resistors are bypassed by C8 so that they set up the DC conditions for the valve, but do not interfere with the way the valve handles signals.

The anode of V2 is fed from HT via R8 and C5, which provide a steady voltage supply of a voltage lower than the full HT voltage of the set, and through R9, where the Audio Frequency (AF) component in the output of the detector valve is developed. The anode is also coupled back through C21, a differential reaction control capacitor (two sets of fixed plates, and one moving set which swings between them) to L8, which gives reaction.

The advantage of the leaky-grid detector is that as well as recovering the Audio Frequency (AF) modulation from the Radio Frequency (RF) Carrier Wave, it can be arranged to give quite considerable gain. A simple diode can carry out demodulation quite happily, but it will not give any gain. So you would need a valve for gain as well.

The leaky-grid detector works by using the valve with very little negative grid bias, where it does not act as a very linear amplifier. It distorts. It amplifies, but it chews the signal up a bit as well. When the valve is set up properly, an RF carrier applied to the grid is amplified through the valve, and appears increased in amplitude at the anode. But also appearing at the anode of the valve are distortion products, which are signals resulting from the valve partly messing up the signal which it is amplifying. One of these distortion products is the audio component, which is the modulation information on the carrier wave, but at audio frequency and ready to be applied to an amplifier and speaker. This audio component can be obtained by making the fluctuating current flowing in the anode circuit of the valve flow from HT+ via R9. The audio signal can then be coupled away to the next valve via C10.

Another clever trick can be worked by virtue of the fact that the anode of the valve has a signal present which is the amplified version of the RF signal. This can be fed into a coil (L8) which is close to the main tuning coil (L6), and can couple back a little of the amplified signal into the input tuned circuit. If the coupling is arranged right, and tightly controlled, a little positive feedback can be set up, which has the effect of increasing the gain of the detector, and also of sharpening up the tuning of the main tuning coil a bit, and helping to filter out any unwanted stations at nearby frequencies. If this is set up carefully, good gain and nice tight tuning can be obtained. Overdo it, however, and the whole thing turns into an oscillator, and sqeals and whistles will be heard in the speaker. And since it is, at this point, a small transmitter sending a small signal back up its aerial, it could well interfere with nearby radios. The positive feedback here is commonly known as "reaction".

The Output Stage.

The resulting simple detector needs to be tuned in, have its reaction set up, and then tuned in tighter, and the reaction set up again. It sounds difficult to use but soon becomes second nature.

All that remains is to send the audio output signal to a simple power amplifier stage and hence to a loudspeaker, and the job is done.

The audio output stage uses a type of valve which is capable of throwing around a reasonable anode current. The output valve, V3, is required to work without too much distortion, so grid leak baissing is a non-starter here. In this case the cathode is biassed positively by the voltage developed across R14 by the current through the valve. Since the grid is referenced to 0V, the effect is of a negative grid, with respect to cathode.

The only slight trick in the output stage concerns R15 and C13. It is fairly obvious that C13 will provide a little fixed tone compensation, in the shape of a treble roll-off. However, R15 makes its effect more pronounced, and there is a good reason why R15 is present. The anode of the detector valve has significant RF components present, and these are fed to the output valve with the AF which is required. Hence quite a lot of RF will be present at the anode of the output valve. This, if not filtered out, could contaminate the power supply to the rest of the set, and thereby feed back to earlier stages. Uncontrolled and unpredictable feedback will lead to oscillations, or at least to ticklish operation of the reaction control, and therefore should be avoided.

Info Main