Revox Model 40 integrated stereo amplifier (two reports)
amplifier menuMain Menu



Taken from 'The Gramophone', December 1962

It so happens that at the present time the store of equipment awaiting test and assessment for these columns would form the nucleus of a very international Audio Fair. Items from Japan, America, Denmark, Norway and Germany as well as Switzerland wait alongside many native products and in due course we shall investigate them all: some may fail to make the grade but the subject of this report passes with honours.

Some time ago I was able to give a glowing account of the Revox tape recorder and I am pleased that no letters have arrived questioning my judgment, only a few bemoaning delay in delivery! The import of an amplifier from the same source cannot be of such great interest because we have a good crop of well-designed and constructed British amplifiers, and present duties inflate the price of the foreign articles. However, their methods and scale of values are sufficiently different to be interesting to many people and those who take the trouble to investigate will not be disappointed. In spite of this, the virtues of this amplifier can be simply stated. It is a good straightforward design, impeccably built with first-class materials and tastefully presented.

A stout, light green, wrinkle-painted metal cover, a dark grey panel, good knobs and clear markings suggest quality and precision. The novel Revox method of horizontal construction wherein all small components are mounted on special turret valve holders has been joined by an interesting new arrangement in the tone controls and their associated many small components and wiring. Each control is of the stepped switch variety and each switch wafer is enlarged to form a small printed circuit board embodying all the necessary and often messy bits required in this sort of circuit. These controls work silently and with great precision and are arranged in a feedback chain over the first ECC81 twin triode stage in each channel. The method, a familiar one in tape recorders, is perhaps not so common in amplifiers: the feed- back path is from the anode of the second triode to the cathode of the first. A similar arrangement is used to provide fixed RIAA equalisation in the ECC83 gramophone pickup stages which are selected, as are the other inputs, by a robust push button switch: an additional button parallels the channels for monophonic use and a sixth turns the equipment off. The remainder of each channel is built around a pair of ECL86 triode pentode valves and consists of an audio stage and a split load drive to the pentode connected push pull output stage.

The output transformers are decidedly adequate and there is overall negative feedback. Power supply is from selenium rectifiers with resistor and capacitor smoothing, the latter totalling 300 microfarads! A DC supply provided for the gramophone pre-stage heaters is also used to give a fixed bias to the output valves.

The results from this excellent design are equally acceptable to ear and meter: indeed my measurements are just a repeat of the manufacturer’s specification and so I will only add a couple of footnotes. In particular, the exactness of the tone control steps and the accuracy of the RIAA compensation (except below 40 cycles where it is deliberately chopped to avoid rumble) are exemplary, repeating as they do the curves given in the sales literature like a carbon copy. To the ear a cleanness of sound and complete freedom from hum or background at all likely listening levels is the most remarkable characteristic and the only real criticism is the all too common one concerning the lack of a low pass filter and even this is not too serious in this case as the type of treble tone control used is more helpful than most.

To say the Revox 40 is built like a battleship is perhaps to mock the Swiss navy. Nevertheless I have a feeling that this will prove to be an extremely reliable amplifier and its performance is in keeping with what we in Britain have always expected when we see the label "Swiss made".

Power/Frequency response:
The power was measured with 220 volts applied to the highest (220v) mains tapping. The makers are not concerned by the apparent overload if normal British 240 volt mains are used, when the output rises by almost 15%.

11 watts output between 60 c/s and 6 kc/s.
8 watts output between 40 c/s and 15 kc/s.

It was not possible to provoke oscillation into any output load although severe ringing of a 1 kc/s square wave was produced. There was in fact noticeable overshoot and traces of ringing when a 15 ohm resistive load was used. Investigation showed supersonic resonances of between 2 and 3db at 37 and 90 kc/s. No trace of any trouble due to them could be found in listening tests, which were repeated after they had been discovered.


Taken from 'Hi-Fi News' , December 1963

This stereo amplifier has the same high-grade workmanship as the E36 tape recorder, and the same meticulous attention has been given to the controls so that they are completely functional and easy to use. The metal case itself is finished in a light green wrinkled paint contrasting with the darker green enamelled panel with its grey knobs. On the left-hand side is a six-button switch, which selects the various inputs, and on the right are arranged the four knobs: bass, treble, balance and volume. The on-off switch is combined with the push-button unit, and selecting radio, tape or gram will automatically switch on the amplifier.

Inside, the valves are placed in an orderly row with the transformers mounted at one end. The wiring is rather unusual for these days as no printed circuits are used: instead, each stage is built around its valve holder in a form of a components module. This type of construction is similar to that used by the tape recorder and simplifies mass production as well as giving layout advantages. It was noted that the highest mains lapping is at 220 V. but presumably the makers are unworried by any overload if the amplifier is supplied with 240 V.

The first valve is one half of an ECC83 which is coupled to another triode, this lime a half of an ECC81. Equalisation is obtained by a feedback loop connected from the anode of the second valve to the cathode of the first. Then come the high level inputs, and the volume control, which is followed by an ECC8l double triode The tone controls .are connected in a feedback loop from the second anode to the first cathode, as in the preceding stage Next comes the balance control, driver, phase-splitter and output stage, using two triode-pentodes ECL86’s. These are pentode-connected and are biased from a separate supply, which also provides DC for the first stage valve heaters. Negative feedback is applied in the normal manner from the output transformer secondary to the cathode of the driver triode Rectification is by bridge-connected metal rectifiers; two for HT and one for bias-cum-heater supply.

These figures are rather higher than claimed due to the higher voltage input (240 V). The feedback loop measured I7dB and this could be increased by 12dB before instability became evident. The frequency response is shown in the diagram, and was measured from the radio input with the tone controls in the central position. The peak of some .2 5dB at 36 Kc/s is probably due to the transformer resonant frequency being too low.

The gram equalisation follows the RIAA curve within 1dB down to 40 c/s. where the high-pass filter takes over and reduces the response to -9dB at 20 c/s.

The tone controls are switched, with three stages of lift and cut. The bass control gave a boost of 6, 10 and 13.5db and a cut of 5. 5 and 14db at 40 cycles: and the treble gives a lift of 4.5, 9.5 and 13db with a cut of 3, 7 and 11db at 10 Kc/s. The balance control has a range of 10db for each channel. Crosstalk measured 47db at 10 Kc/s and 55db at 1 Kc/s. Hum and noise were very low at -62db gram, 65db tape, -62db radio 1 and -65db radio 2. The tone controls were very well matched, as might be expected, and showed a divergence of less than 1db. The volume controls were also matched within 1db except at very low settings where the error increased to 2db.

Finally, a square-wave test was applied using a Tektronix 105 generator and type 545 oscilloscope. As expected from the peak in the frequency response, both the 1 Kc/s and 10 Kc/s waveforms showed signs of ringing, and the overshoot at the lower frequency is definitely noticeable (see sketch). It might be asked: just how important is such a test? Well, it must be said right away that a good amplifier should reproduce a 1 Kc/s square-wave without any appreciable signs of "ringing" or overshoot. Some "rounding" of the waveform al 10 Kc/s would be perfectly acceptable, but the waveform should show little visible ringing. If an amplifier passed these tests and had a good stability factor then it is reasonably certain it would work with any speaker system and with any crossover arrangement without trouble.

The square-wave tests were taken with a non-inductive load, and in actual practice the transient response would be further affected by capacitive and inductive loads, thus the addition of purely capacitive loads degraded the performance still further, but the connection of a three-speaker system using a ¼ wave cross-over improved the waveform and no transient distortion was noticed in the actual listening tests.

The hum and noise levels were commendably low and the controls worked very well; I had imagined that the "steps" might be too steep, but in practice I did not really feel the need for any intermediate positions. No low-pass filler is fitted, but the shape of the treble feedback tone control characteristic helped to remedy this defect rather better than the passive type, which tends to swing the response about a central "hinge". The styling and presentation is very good indeed, although some people may consider the appearance too clinical and severe. As stated before, the workmanship is excellent and apart from the reservations concerning the square-wave performance, the Revox 40 is a fine amplifier worthy to be associated with the Revox tape recorder.

PS. One minor grouse - no speaker plugs or indeed any plugs were supplied, and the former are non-standard!