Reviews and Recollections - The Early Silicon Transistor Era

For Armstrong the first public signs of the ‘Silicon Transistor Era’ were when franchised retailers received an Armstrong ‘Bulletin’ which announced that the new 600 Range would make its public debut at the ‘SONEX’ show that was open from March 28th to April 1st 1973.

The April 1973 issue of ‘Hi Fi News’ magazine – which would have been on sale both before and at the show – included a full colour foldout bound between pages 702 and 703, This displayed a large photograph of the 626 AM/FM tuner + amplifier (receiver) with the headlines: “We know it’s the best we’ve ever made.” and “We also think it’s the best you’ll have ever heard.”

Both the appearance and circuitry of the 600 range were a real advance over the 500 range. At that point Armstrong were riding on the success of the 500 range and felt confident that this was a significant improvement upon it. At the time Hi Fi News had a full colour cover, but all the internal pages were, by default, black-and-white. So the double-size colour photo foldout was a really dramatic introduction. The magazine was selling over 60,000 copies per month in 1973 and was the best-selling UK Hi-Fi magazine. It had become established as the ‘paper of record’ for the UK Hi-Fi scene. So the bold advert was a real event for Armstrong and the magazine.

The following issue (May 1973) of Hi Fi News followed up by featuring the 626 on its cover. In fact this cover photo showed the 626 in the living room of Barrie Hope’s then-home in East Grinstead. Barrie was one of the Armstrong directors at the time and had supplied the photo for the magazine. The editorial page comment on the cover picture was:

“...our real point here is to show the new Armstrong 626 AM/FM tuner-amplifier in a domestic setting. Supplemented here by Revox tape recorder, Thorens turntable, SME arm and Shure cartridge, all comfortably housed on Beaver Shellvex fittings, this fine piece of British electronics sounds splendid and should be very popular.”

The April 1973 issue of “Hi Fi Sound” also showed a close-up of a 621 amplifier and 624 tuner.

However, despite featuring on the covers of these magazines, the first review was actually in another UK magazine, “Audio”...

600 Range

1973 May “Audio” – 626 Receiver (Amp + AM/FM tuners) review by Fred Judd

Given the keen interest the magazines clearly had in the new Armstrong range it’s understandable that Fred Judd began his review by saying:

“We are privileged to be the first of the audio journals to review what might be considered the ultimate of the new Armstrong Audio 600 series range of high fidelity radio receivers and amplifiers.”

He went on to explain the modular nature of the 600 range units and that by reviewing the 626 the magazine could also cover the performance of the tuners and amplifier when sold as separate units.

“The model 626, like its sister models, is the result of five years intensive research and utilises many new techniques in both electronics and audio circuitry. The 'mechanics' of the series also include many new design features and materials guaranteed to ensure strength in construction as well as continued reliability in use. Styling is yet another special feature and for this a low, slimline look, has been adopted.”

And then summarised the controls, features, and facilities provided, adding:

“... among these are tabs for the choice of three pre-tuned FM stations and three pre-tuned AM stations. The normal radio tuning control otherwise covers the full VHF band and the medium and longwave bands. It is interesting to note, however, that tuning is accomplished by means of varicap diodes so the usual ganged tuning capacitor we have been accustomed to in radio tuners is replaced by a variable potentiometer operated from a flywheel assisted tuning control. Because of the special HF circuitry used in the 626, the whole of the medium and longwave band has continuous tuning so one does not have to switch from one waveband to the other.”

and:

“One particularly novel idea - and a very good one - must be mentioned. To overcome the audible clicks and bumps which usually occur when circuits carrying audio signals are switched, the makers have introduced diode switching. This eliminates switch noises and also provides a quick but quite automatic and smooth fade-in of any selected input signal.”

Details like these showed that the circuits in the 600 range were unusual for the period and incorporated some quite ingenious ideas. When it came to performance in use he commented that:

“The 626 has a superb performance on radio which could not be faulted and was moreover quite up to that specified... There is a most noticeable absence of spurious whistles and other unwanted signals usually picked up on AM bands and even on the built-in ferrite aerial many stations could be received free of the usual howls and whistles that normally accompany otherwise good programme material. With an external aerial reception on the AM bands is phenomenal.”

It is worth recalling that Ted Rule (who designed the 600 range) was a radio amateur who had served in the RAF, and that Armstrong had – for many decades before the term “Hi-Fi” was known – developed and sold high performance radio chassis to users who wanted outstanding radio reception quality.

Fred Judd then concluded by listing the measured results and saying:

“I have described the main features and facilities but admit that, space permitting, more could be said in praise of these. The test figures prove the product and its worth so if you are looking for and can afford a tuner/amplifier, or as Armstrong call it, an AM/FM stereo receiver that will match the performance of any high fidelity auxiliary equipment you care to use with it, then go out and buy the 626.”

1973 October “Hi Fi News” – 626 Receiver (Amp + AM/FM tuners) review by Gordon King

Hi Fi News published their review of the 626 AM/FM receiver on pages 2033-2039 of their October 1973 issue. Gordon King started with some general comments on the 600 range, including:

“The printed circuit boards of the preamplifier, power amplifier, FM receiver, decoder and AM receiver sections are designed as common building blocks, which not only facilitates the factory assembly of any model, but which will undoubtedly ease any subsequent servicing requirements. This technique, of course, also makes it possible for the manufacturer to continue with state-of-art developments without having to start at square one, so to speak.”

“The 626 is good to look at, and even though it yields a true 40+40 W into 8-ohm loads and has sensitive radio sections it is no larger than 3¼ H by 9¾ W by 11¼ D in., so would not be difficult to place on a small shelf or hi-fi cabinet. Substantial rear heat sinks look after the power transistors, and although the temperature here rises mildly on quiescent current, the thermal capacity is adequate for protracted full-power music signals - and sinewave signals, come to that, as my tests proved! Facia is black and white print with contrasting white control knobs, and the woodwork can be either teak or rosewood at the same price.”

He then outlined the many controls provided via the ‘tabs’ along the front panel, etc, before focussing on the tuner:

“The meter registers the reverse-bias applied to varicap tuning diodes, and since diode tuning is used on AM, which is unusual, as well as on FM, AM station preselection becomes possible.”

“With the AFC switch up the meter deflects to the centre circle when the tuning is adjusted accurately (thus indicating correct balance of the FM detector). With the AFC switch down, the AFC becomes active and ‘locks’ the tuned station while at the same time the meter records in terms of aerial signal strength. Moreover, the signal-indicating sensitivity of the meter is affected by the distance press-key, the deflection for a given signal level being greater in the ‘down’ position. This switch also works the FM muting.”

“This is the only hi-fi receiver that I know of which tunes continuously - without switching— over the LW and MW bands from one end of the scale to the other. This is truly remarkable, particularly when it is known that a solitary varicap diode is the only variable tuning element involved! This is achieved by double conversion.”

“During the period of the trials I had no trouble in receiving stereo signals over a path of some 100 miles from a six-element aerial, and for most of the time the signal was well above noise, reasonable reception holding even when the signal strength had diminished so as to put other tuners well into noise.”

He then discussed the audio sections:

“The audio department came well up to specification - in fact, it was virtually impossible to catch out the specification on any count!”

Overall the review measurements met or exceeded the specifications published by Armstrong. However the reviewer did encounter one result which puzzled him:

“With the resistive loads I normally use for testing amplifiers, I was unable to obtain the same low distortion at 10 kHz as measured by Armstrong. However, by employing the loads used by Armstrong, it was later found that the 10 kHz distortion factor was hardly any greater than that at I kHz. The later values are included in parenthesis under the distortion parameter in the test results. The Armstrong loads also made it possible to obtain a greater 10kHz (ref. 0·5% distortion factor) power capacity.”

“It is curious why this should be so, but detailed tests gave the impression that the RL Zobel network used in the output stages for ‘neutralising’ the impedance effects of loudspeakers has a bearing on the case, since by shorting out this network the 10 kHz distortion fell when my normal loads were used. It thus appears that by some coincidence the Zobel network cancels any reactive components of the Armstrong loads, while having minimal cancelling effect on these components of my own loads. All this makes me wonder just what happens to the distortion of any amplifier when it is driving power into real loudspeakers!”

Engineers experienced with designing and testing audio power amplifiers may recognise this type of behaviour where the precise levels of distortion at high frequency may depend upon the test loads and even the wiring layout!

In his table of results Gordon King listed the values he measured using both his original test loads and the ones used by Ted Rule. The values obtained using Ted Rule’s loads were given in parenthesis to show the difference the loads could make. The review then concluded with a summary:

“Well, that’s about it: a very well designed piece of British hi-fi equipment, and although this review has exposed one or two small curios, these are small compared with the excellence of the overall quality; and in any case, anything which is improvable will be improved - with Armstrong you can bet your bottom dollar on that. A great deal of design effort has undoubtedly been injected into the ‘600’ series, and Ted Rule and his small design team up there in London N7 have much to be proud of. How encouraging to see that British hi-fi is taking the lead once again...”

1973 November “Luister” [Dutch] – 621 Amplifier review by Jan Kool

I can’t help wondering if “Jan Kool” was a pen-name! However this review was quite short and to the point:

“The Armstrong 621 is definitely a good amplifier and in addition it is remarkable for its very good looking styling. It is certainly at a very low price for its power output and furthermore it has excellent sound quality and very useful facilities. There are two treble filters with a slope control of 6db and 12db per octave, with both filters pressed one gets a third crossover frequency. There is a variable tape recording output of between 0 and 250 mV and this is a facility I would like to see on all amplifiers so that all the matching trouble with American, Japanese and European tape recorders would end.”

“In nearly all cases the specification of the 621 is exceeded and stability is unconditional, one of the most important features of this amplifier I feel. So the Armstrong is usable in any circumstances with any loudspeakers, including electrostatics. Also the headphone socket will accept all types of headphones, a feature one finds with only a very few amplifiers.”

“Conclusions. A 2 x 40 W. at 8 ohms (2 x 50 W. at 5 ohms) amplifier with negligible distortion, absolute stability, two pairs of loudspeaker outputs, good filters and nice details like variable tape output for less than 800 guilders and on top of that an alternative appearance.”

“Unconditionally recommended.”

Amongst the measured result were:

“THD at rated power: 0·054% at 1 kHz, 0·14% at 10kHz, 0.13% at 40Hz

THD at 1 Watt 8 Ohm: 0·038% at 1 kHz, 0·09% at 10 kHz, 0.08% at 40Hz.

SNR Aux 100mV -67dB, Disc 2·4mV -67dB

RIAA +/- 0·5dB: 20Hz - 20 kHz

Damping factor: better than 80 at 30 Hz

Stability: unconditional”

1974 February “Records and Recording” – 626 review by John Earl

John Earl began his review with a general description of the 600 range before adding:

“The various configurations are based on a pattern of printed circuit board modules which facilitate not only the production of the series but also any subsequent servicing requirements – particularly from the dealer’s point of view. Moreover, each item is designed for ease of electronic assembly exposure; on the 626, for example, it needs only the sliding of a couple of catches to release the wooden cover and make accessible the printed circuits boards, which themselves are conveniently arranged...”

That was correct in 1974, but it is worth adding that a few years later a small metal clip was added which required a small bolt to be undone before the cover could be removed. This was because of changes to the regulations on the safety of mains-powered devices. For engineers and enthusiasts being able to ‘pop the lid’ with the flick of two catches was handy. But it also meant there was a risk that tiny hands might find they could do the same! Adding the bolt and clip meant a screwdriver was also needed to free the lid!

As with many other reviewers, John Earl took particular interest in the radio sections.

“A good deal of design thought has been injected into the radio sections – some revived from days past but in contemporary guise, and some state-of-the-art.”

He then commented on the usefulness of the ability to tune right across the Long and Medium wave bands and use electronic presets for these as well as the FM band. Features made possible by the use of varicap diode tuning and double-IF conversion. Virtually unique in modern HiFi tuners! Regarding the FM tuner section he wrote:

“Varicaps are also used for the FM, there being three prior to the mixer and one for local oscillator tuning. The net result being an FM section of desirably high front-end selectivity, and hence high rejection ratios and immunity from input overload – the requirements for operation in areas of high signal field without breakthrough of a strong station when a weak one is tuned.”

He then turned his attention to the amplifier sections:

“In fact, referred to 0·5% distortion the 1 kHz power was 48+48 Watts into 8 Ohm loads. Distortion factor at 40+40 Watts was a mere 0·05% at 1 kHz, falling with decreased power into the noise. The distortion at 10kHz was not much greater.”

“The negative feedback loops are arranged so that the source impedance remains at a low value at low-frequency, thereby ensuring a good bass damping factor which, particularly with certain speakers, helps to achieve a ‘tight’ and well controlled bass reproduction, and this without resource to direct coupling and its attendant problems.”

“Half-power bandwidth was 25Hz to 50kHz (input auxiliary). Hum and noise referred to 40W was 66dB on pickup and about 70dB at the other inputs”.

He then pointed out the adjustable tape output levels, headphone socket, speaker switching for two pairs of speakers, and other features. In particular, mentioning the use of electronic input switching to avoid unwanted noises. He concluded:

“The 626 is also pleasantly presented and having in mind the power yield and facilities provided is relatively small.”

“After extensive laboratory tests I have no hesitation in proclaiming the Armstrong 626 one of the best British tuner/amplifiers currently available, and anyone requiring a good item of British hi-fi of accurate specification and realistic price should certainly not overlook the new Armstrong range of equipment.”

1974 October “Audio” – 625 review by Gordon King

This review began with :

“GORDON J. KING re-examines this unit in its modified version and in the light of a year's experience with a 625 in home use.”

This ‘re-examination’ followed on from the May 1973 review in “Audio” of the 626 by Fred Judd. Now, Gordon King – who had reviewed the 626 for Hi Fi News the previous year – began by giving the background to his review:

“Shortly after the launching of the '600' series of Armstrong equipment - consisting of an amplifier, FM and FM/AM tuners and FM and FM/AM tuner-amplifiers - Fred Judd presented a very encouraging review on the Armstrong Model 626 which is the FM/AM tuner-amplifier, or receiver as it is now sometimes called. For pretty well twelve months I have been operating a different sample of this same model under typical domestic conditions, and just recently have had the opportunity of comparing it both objectively and subjectively with a new version of Model 625, which is the FM-only receiver. This article, therefore, is partly a review on the new 625, partly a user's report on the 626 over a twelve-month period, but with technical comment, and partly a comparative report between the early 626 and the latest 625.”

“It should be noted that the various Armstrong models are based on 'module' assembly, which means that any updating in the audio section is reflected into the two tuner-amplifiers as well as into the amplifier, while any updating in the radio departments vs also reflected into the tuner-amplifiers as well as the tuners”

He went on to write:

“Let us see first, then, how the 626 fared over the twelve months. It was operated day in and day out, being used for both background music from radio and hi-fi-scaled sessions from radio and other sources, and on average it would be fair to say that the tuner-amplifier was running for at least eight hours each day. At no time during the twelve-month period did it give trouble or show any sign of distress.”

“The power yield was found to be adequate for hi-fi-scaled music sessions in conjunction with loudspeakers of circa 0·3 per cent efficiency in a domestic environment of some 60 cubic metres and rms peaks could easily be measured within the amplifier's dynamic range sandwich; that is, without ppp being masked by noise or fff running into clipping.”

“Hum and noise with the volume at minimum was measured as a mere 0·4mV across eight ohms, which corresponds to the incredibly low power of 0·02 micro watt! In an early report I also referred to the distortion at 10 kHz obtained with certain load resistors used for the measurement. It has since been proved that these loads suffered from a curious kind of nonlinearity and are now no longer used. The distortion at 10kHz with linear loads is less than 0·1 per cent at 35+35 watts.”

Having discussed the amplifier sections he wrote about the tuners:

“Although sited some 100 miles from the nearest stereo station, over the twelve months I have been able to receive entertainment quality stereo from this station in conjunction with a six element array located in clear roof-space for at least eighty per cent of the time.”

“Reports have been received of a high-pitched whistle from some receivers (NOT Armstrong) employing a phase lock loop decoder IC when tuned to certain transmitters using slave switching. The mechanism producing this is different from that responsible for stereo 'birdies'. Slave transmitters may use a 23kHz tone for switching operations, etc. which, at the receiver, beats with the 76kHz divided-down 19kHz signal generated by the decoder thereby yielding a 4kHz whistle. The Armstrong tuners overcome this by input filtering to the decoder; but more recent versions are using an improved decoder IC which not only gives better separation but also much lower spurious outputs.”

At the time stereo ‘birdies’ was an effect suffered by some poor stereo FM tuners. it occurred if a harmonic of the 38kHz subcarrier used for FM stereo could ‘beat’ with other FM transmissions that were in adjacent (frequency) channels. The result was high frequency ‘twittering’ noises like distant birdsong, hence the name. Gordon explained that:

“A receiver with poor IF selectivity is more prone to the effect than others, and since the selectivity of the Armstrong is upwards of 60dB, provided by ceramic filters, the effect is rarely experienced.”

Gordon King was an amateur radio enthusiast. He had a keen interest in tuners being able to produce good results from distant, weak, radio stations in the presence of possible sources of interference. So the fact that he was impressed by the ability of the FM tuner to pull in a distant station and reject interference was quite a good mark for the 600 range’s tuner.

He then examined an area which is rarely considered in Hi-Fi reviews:

“The recent version of the 625 reflected some of the developments injected by the Armstrong engineers from field feedback. This philosophy of forever striving towards improvement is highly commendable, for once a good design has been attained neither the customer nor the manufacturer benefits from a completely new 'start-from-scratch' design. With the Armstrong method of 'module' construction it is a relatively simple matter to introduce slight changes in parallel with the production.”

These comments are significant because they illuminate Armstrong’s long-established policy of ‘continuous development’. This mean that units sold over the years with the same model number, appearance, and published specifications actually tended to gradually be improved during production. As a result it was the case with the 600 range that examples made later on were improved in many detailed ways compared with the version originally released.

“One area that has received the attention of E. A. (Ted) Rule, Armstrong's designer and chief engineer, is in the amplifier and is concerned with the low-frequency response and phase linearity. It might be recalled that the original design was based on capacitor coupling to the loudspeakers, and with all the recent talk about the merits of direct coupling, he decided to find out for himself whether the amplifier would reflect any significant advantage by changing over to this technique. A laboratory programme was set into motion, and it was found that the prime features of direct coupling could be achieved by retaining the original capacitor coupling, keeping this in the feedback loop, with one or two small changes, which have now been incorporated in the new versions.”

Again, the above comments show one of the unusual aspects of the 600 range’s design. The power amplifier employed output capacitors between the amplifier and the speaker. This was usual for ‘single rail’ amplifiers to prevent d.c, being continually applied to the speakers. However the capacitor tended to introduce unwanted side effects. In particular it reduced the ability of the amplifier to control the speaker at low frequencies and also tended to limit the low frequency response.

The 600 power amplifier employed two feedback loops. One taken for the loudspeaker side of the output capacitor. This allowed the design to reduce the unwanted effects. It could extend the low frequency response and improve the amplifier’s ‘damping factor’.

With this in mind the reviewer went on to report:

“From 50Hz to 1 kHz the damping factor holds around 55; from about 2Hz to 50Hz it tends towards slightly negative and infinity, while a little below 2Hz it is still not less than about nine. It is at these low frequencies where some loudspeakers call for maximum electro-magnetic damping control (ie, very low source impedance, tending towards negative), and this is, in fact, one of the arguments put up in favour of direct coupling.”

“The L.F. response of the Armstrong power amplifier was measured and at 5Hz the response was less than 2dB below the full power at 1 kHz. Incidentally, it needs an audio millivoltmeter extending down virtually to dc to measure this. Most run-of-the-mill meters of this kind are themselves around 4 or 5dB down at 5Hz. It must be admitted that this is jolly good performance.”

“The 25Hz distortion factor at 35+35 watts into eight-ohm loads was less than 0·1 per cent, and this included ripple components.”

So overall the low-frequency performance was felt to be good even when compared with ‘direct coupled’ amplifiers that didn’t need an output capacitor!

Comparing the new 625 with the original 626 he mentioned various detailed improvements - higher FM sensitivity, lower levels of distortion in the preamplifier section, etc before concluding:

“It is pleasant to be able to review after twelve months of use a British design so well evolved and to report on the improvements which have been so carefully filtered from field feedback. Not many people are aware that it is the policy of Armstrong to keep complete statistics of all the service faults that they encounter so that any trend becomes apparent before it can generate out of proportion. I suppose this is one big advantage of the smaller enthusiast firm which undertakes its own servicing Any small abnormality is immediately analysed, thereby making it virtually impossible for it to develop into a real problem. There are not many firms nowadays who can claim the aimed at 48-hour technical service of Armstrong.”

1974 December “Hi Fi Answers” – 626 Receiver (Amp + AM/FM tuners) review.

This was part of a “Top Systems Spotlight” item. The comments included the following:

“The next task was to choose a suitable tuner-amplifier, and this task was made quite easy by the appearance on the market of the Armstrong 600 series. The 626 in particular, is a very fine piece of equipment, and offers a number of unusual facilities.”

“The 626 may be used with 3-head tape decks for ''off-tape'' monitoring and we were pleased to note the inclusion of two high frequency filters and a slope control, giving one a choice of two roll-off rates. This is much better than the statutory "scratch" filter that so many manufacturers include as an afterthought and they certainly seemed effective in cleaning up "less-than-perfect" material.”

After mentioning various features of the design the review went on to say:

“The instructions supplied were more than adequate, and, a particularly nice touch, Armstrong seem to have thought of every contingency in the "leads and plugs" packet supplied, even including adaptors for phono to DIN plugs on any input, and suitable 300 or 75 ohm aerial plugs etc. The unit performed extremely well in general use, picking up quite difficult FM stations with remarkable ease, and providing reasonable entertainment value on AM. Power output was rather in excess of that claimed, at 46 watts per channel, into 8 ohms, and the signal to noise ratio on the disc input was of a better than average value. The sound produced through our chosen speakers, was of a high quality level, and we can strongly recommend their [i.e. Armstrong] latest offering as excellent value for money.”

It is perhaps worth adding to the above that the ‘leads and plugs’ pack provided with 600 tuners and tuner-amplifiers also included a small screwdriver. This was to ensure the owner had to hand a screwdriver suited to adjusting the FM/AM presets when they first started using their new set.

Another small touch was that the wooden sleeves of the 600 range units were polished at the factory just before they were sealed into their delivery cartons. This was to ensure they looked good, and smelt fresh and new when the carton was opened by the eager new owner!

During the 1970s the quality and performance of silicon transistors improved rapidly. As that happened Armstrong introduced many changes to the 600 range to exploit the better devices and deliver increased performance and reliability. Unfortunately, during the same period, changes were also taking place in the UK economy in general. And the behaviour of reviewers was also changing. Despite the popularity and good performance of the 600 range, these factors contributed to Armstrong eventually ceasing to make Hi-Fi equipment during the next few years. Armstrong didn’t actually ‘go bust’ as some assumed later. And what happened next may in some ways be a salutary lesson, now, from history. But that’s a story for another page to tell...