The BBC LS3/6

[Tony L]

Yours is a Quad serviced one isn’t it? Or did you modify it in some way?

I have two. One I rebuilt myself using modern Dada Electronics boards, i.e. very nice glass-fibre boards, all populated with modern audiophile components etc. The other one is a simply amazing condition late-60s original that has had its electrolytic capacitors and trim-pots replaced and has rather larger coupling and PSU caps, but is otherwise totally stock, i.e. has its original carbon resistors, driver board transistors. Both have the correct output transistors etc. The latter (mainly stock) one sounds quite noticeably better to my ears, and that is the one that met the lovely Radford. I don’t understand why, but that particular 303 is a bloody good sounding amp!

 

[tuga]

A bit Spendor of history:

Derek Hughes supplied a copy of a wonderful letter from 1980, sent by his father, Spencer Hughes, to . Part of the BC1 design team and founder of Spendor, the company that would manufacture it, 'Spen' was responding to a query about the origins of Spendor. Because it describes the speaker's origins 'from the horse's mouth,' so to speak, we're reprinting it here:


Dear Sir,

The Spendor BC1 was not, as it has so many times been described, a development of the BBC loudspeaker type LS3/6. Perhaps a short history of the lead into, and the development, of the two systems may be of interest.
From the very early days, even before hi-fi, the BBC has designed its own monitor loudspeaker systems as commercial systems were not, and most are still not, accurate enough for broadcast work. These designs were based on available units matched by, what were in those days, very complex crossover networks and mounted in custom designed cabinets.
During the mid-1960s, the development work carried out by the BBC had advanced to a stage which was beyond the capabilities of the available paper pulp cone bass units. The decision was taking to investigate the possibilities of using some form of plastic as a cone and surround material. It was assumed that plastic would be a consistent material unlike paper pulp, which to some degree seemed to depend on the mood of pulp stirrer. Over the years it has been found that it was not quite that easy.
The section of the BBC Research Department involved in this operation was headed by Mr. Del Shorter, now retired, with Mr. H.D. Harwood now of Harbeth Acoustics, second in command and myself completing the investigating team.
Some two years were spent making 12in unit cones in a variety of shapes and from a range of plastics; this could be a story on its own. The first successful unit was made from the now well-known Bextrene and used in the development of the BBC studio monitor type LS5/5. This loudspeaker was described in an article written by Mr. H.D. Harwood in the March 1968 issue of .
My part, as a laboratory technician, in the operation was to do most of the actual work both on the plastic investigation and the development of the LS5/5. With that experience I decided that it should be possible to make a loudspeaker from scratch in the home environment. With the aid of our electric fire, a compressor working in reverse and an iron bedstead the first vacuum former was built. Bins full of malformed cones were produced before any measure of success was achieved and the first 8in unit was produced. This unit turned out to be almost certainly the first commercial 8in Bextrene driver and still arguably the best.
The first pair of BC1s was constructed using these units and Celestion HF1300 units. The cabinets were smaller than the current model and initial listening tests indicted that the performance could be improved by an increase in size, hence the present design. At this point it was all being done for fun.
The second pair of BC1s was made for a friend who took them to Merrow Sound of Guildford. The third pair was sold to Merrow Sound and Spendor was on the way to a small niche in the audio world.
Now some difficulties were beginning to arise as under the terms of my contract with the BBC, the design had to be offered to them. Fortunately the 'Pop' era had just started and the main request was for more power, so the BC1 was turned down. Around about this time there was a special requirement within the BBC for one pair of speakers about the size of the BC1s. Being a kind soul, I suggested that my design could be used, so I was given the task of producing an official version of the BC1, later designated the LS3/6.
This design used an 8in unit made by Research Department, the Celestion HF1300 and a redesigned crossover. The main change in the crossover was the addition of a large multi-tap autotransformer to allow adjustment of levels between the two units, normal BBC practice at that time.
Some months later BC1s were fitted with an amplifier mounted in the back panel and the 4001G super tweeter added. This addition was for purchase tax reasons, but it did have two extra gains. Firstly, it improved the overall dispersion characteristics, secondly, from the broadcasting angle, it made any 625-line breakthrough to be more easily detected.
Now the LS3/6 was offered to a number of commercial companies and eventually taken by Rogers, then under the control of Jim Rogers. With approval, and a little assistance from the BBC, Rogers added the Celestion HF2000.
As Spendor was now a commercial company it was agreed that a royalty should be paid to the BBC for each BC1 produced. This was in recognition of the work I had done on the loudspeaker whilst still employed by the BBC.
To perhaps prove the order of development of the two systems, it is of note that out of over two thousand BC1s supplied to the professional market there are over six hundred in operation with the BBC and as far as I know very few, if any, LS3/6 speakers.
In addition to the above, the name Spendor is derived from the first names of myself and my wife Dorothy. Mrs. Hughes provided practical assistance in the early days with her coil winding expertise and now as Managing Director is responsible for all accounting, sales and general management. Derek Hughes, the son, another ex-BBC employee, deals with an amplifier design and assists me with research and development and general running of the factory.

Yours,


Spencer Hughes

 

[tuga]

The BBC had it's own research team of engineers, dedicated facilities and specific design objectives.

A small selection of old papers by the BBC Research Department.

A Survey of Performance Criteria and Design Considerations for High-Quality Monitoring Loudspeakers
D.E.L. Shorter - January 1959

Loudspeakers used for monitoring purposes in broadcasting and recording studios are designed to give the nearest practicable approach to realistic reproduction. The paper discusses the various criteria which can be applied to the performance of such loudspeakers, together with the relationship between the measured free-field characteristics and the response as subjectively assessed in the working environment. While the degree of realism achieved in sound reproduction can only be judged aurally, even subjective assessments can be misleading unless carried out under controlled conditions and with clearly defined terms of reference; the precautions necessary in such tests are discussed. Some of the less obvious design considerations are reviewed and illustrated by examples.

http://www.aes.org/e-lib/browse.cfm?elib=579 ($)


Design of a new free-field sound measurement room: Specification and performance
D.E.L. Shorter, H.D. Harwood (Harbeth), C.L.S. Gilford, J.R. Chew, R.L. Deane - 1965

The various requirements laid down in the design of the new free-field room at Kingswood Warren are discussed. Details are given of the building, the acoustic treatment and the associated technical equipment. Acoustic tests on the completed room are described; the frequency range over which free-field conditions can be obtained depends to some degree on the direction of sound propagation, but in favourable circumstances extends to below 50 c/s.

http://downloads.bbc.co.uk/rd/pubs/reports/1965-17.pdf


The design of a low-frequency unit for monitoring loudspeakers
H.D. Harwood - 1966

The present state in the design of low-frequency loudspeaker units is reviewed and the areas where improvement is desired are indicated. Experimental details are given leading to the design of 12 in. (305 mm) unit incorporating a vacuum-formed cone of toughened polystyrene with a p.v.c. (poly-vinyl chloride) surround and it is shown by objective and by listening tests that this design is superior to existing units. An analysis of the price indicates that the new unit should not cost any more than those at present in use.

http://downloads.bbc.co.uk/rd/pubs/reports/1966-28.pdf


The design of studio monitoring loudspeakers Types LS5/5 and LS5/6
H.D. Harwood, S.A. Hughes (Spendor) - 1967

Details are given of the various factors which have led to the design of two new monitoring loudspeakers suitable for use in studios and outside broadcasts. The loudspeakers are much smaller than those of the present type; a floor standing model is designated type LS5/5, and one intended for hanging is called type LS5/6. In the course of the design, the questions of bass pre-emphasis and of directivity have been examined in some detail.
The quality of reproduction and the directional properties are appreciably in advance of those obtained for the LS5/1A and the maximum sound level is also higher. The spread in frequency characteristics between development specimens is extremely small, and the level of non-linear distortion is low.

http://downloads.bbc.co.uk/rd/pubs/reports/1967-57.pdf


Aspects of High-quality Monitoring Loudspeakers
H.D. Harwood, C.L.S. Gilford, S.A. Hughes - September 1969

In Part I of this monograph a description is given of the design and performance of a compact two-unit loudspeaker originally intended for use in mobile control room.
Part II describes investigations to determine why the reproduction given by the latest hanging studio loudspeaker type LS5/6 became noticeably coloured when it was suspended above the picture monitors in a television sound-control room, as dictated by considerations of layout. It is concluded that the colouration was due to the build-up of multiple reflections from the walls and windows when the loudspeaker was suspended symmetrically in the corner of the control room and that the colouration was more noticeable in the case of the LS5/6 primarily because of the higher quality of this loudspeaker.

http://downloads.bbc.co.uk/rd/pubs/archive/pdffiles/monographs/bbc_monograph_78.pdf


The design of the LS3/4 loudspeaker
H.D. Harwood, S.A. Hughes - 1969

A description is given of the design and performance of a compact two- unit loudspeaker for use in mobile control rooms. Both the frequency range and the quality of reproduction are similar to that of the studio monitoring loudspeaker type LS5/5 from which this design was developed; the power handling capacity, whilst less than that of the LS5/5, is quite adequate for the purpose.

http://downloads.bbc.co.uk/rd/pubs/reports/1969-05.pdf


New BBC Monitoring Loudspeaker
H.D. Harwood - March 1968 (Wireless World)

1. Design of the low frequency unit
2. Bass equalisation: The cabinet: Frequency response characteristics of the units

http://www.keith-snook.info/wireless-world-magazine/Wireless-World-1968/New B.B.C. Monitoring Loudspeaker.pdf


Stereophonic Image Sharpness
H.D. Harwood - July 1968 (Wireless World)

Results of experiments to discover how the width of the sound image is related to the image position; also how image position is related
to inter-channel level difference.

http://www.keith-snook.info/wireless-world-magazine/Wireless-World-1968/Stereophonic Image Sharpness.pdf


Loudspeaker distortion associated with low-frequency signals
H.D. Harwood - January 1972

Three differing forms of distortion, which are associated with low-frequency signals in loudspeakers, are investigated.
It is shown that distortion due to the Doppler effect can be compared with that due to wow and flutter in recording machines and subjective data obtained for this purpose can be applied to loudspeakers. Generalised design limits for loudspeakers are calculated and it is seen that present BBC designs lie within these limits.
In loudspeakers designed to reproduce low frequencies, the voice coil is made longer than the magnetic field. At low frequencies, when the amplitude of vibration of the cone exceeds the difference in length, it is shown that instead of the peaks of the waveform being clipped, expansion of the input/output curve takes place. This effect, with its associated distortion, can be compensated by employing an appropriate non-linear suspension and thus a much greater useful output can be obtained than by using a linear suspension.
Finally, a vented cabinet is often used to reduce the magnitude of the undesirable effects previously mentioned as well as to extend the bass response. However, a vented cabinet is a resonant system and high sound pressures and particle velocities are produced in the vent. These are liable to give rise to distortion from the inherent non-linearity in the air and from turbulence at the orifice and in the pipe. Existing data are used to estimate the sound levels which may be generated in a typical listening room before distortions from any of these causes is audible. It is also shown that this form of distortion is not a troublesome factor in the design of the BBC Studio monitoring loudspeaker.

http://downloads.bbc.co.uk/rd/pubs/reports/1972-25.pdf


The design of the miniature monitoring loudspeaker Type LS3/5A
H.D. Harwood, M.E. Whatton, R.W. Mills - October 1976

This report describes the design of a miniature two-unit loudspeaker of adequate sound quality and loudness to serve as a monitor in conditions where larger existing designs would be unusable.
Details are given of the construction and performance of the loudspeaker which is shown to be equally suitable for monophonic, stereophonic and quadraphonic purposes.

http://downloads.bbc.co.uk/rd/pubs/reports/1976-29.pdf


Factors in the design of loudspeaker cabinets
H.D. Harwood, R. Matthews - January 1977


The mechanical properties of timber, wood products and other materials potentially suitable for the construction of loudspeaker cabinets have been measured and details of the results are given. Various commercially available damping materials have also been assessed and their relative efficiencies are listed. A new method of test for the cabinets of completed loudspeakers has been devised and a tentative performance specification has been produced.

http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf


Some factors in loudspeaker quality
H.D. Harwood, R. Matthews - May 1976 (Wireless World)

Some of the factors in loudspeaker design have been dealt with in the technical press many times but there are others which have received comparatively little attention, at any rate quantitatively. In this article it is proposed to deal with a few of the latter and to add some subjective data which is new.

https://pdfhost.io/v/Ff4nP~MG_Some_...arwood_R_Matthews_May_1976_Wireless_World.pdf


A new distortion measurement - Better subjective-objective correlation than given by THD
R.A. Belcher - May 1978 (Wireless World)


This article describes a new technique for measuring non-linearity distortion which gives much better correlation with subjective assessment of sound quality than does the conventional total harmonic distortion measurement.
 Known as the double comb-filter method, it uses pseudo-random binary test signals and largely digital processing, so the cost of instrumenting it can be expected to fall with the increasing availability of l.s.i. circuits. Ultimately, the hardware may cost less than that used at present for THD measurement. Ways are suggested for using the new technique to measure cross-over distortion, transient intermodulation distortion and other parameters such as "wow" and "flutter" and linear distortions. The BBC is now testing it on sound-signal transmission circuits and studio equipment.

http://www.keith-snook.info/wireless-world-magazine/Wireless-World-1978/A new distortion measurement.pdf


Design of the high-level studio monitoring loudspeaker type LS5/8
C.D. Mathers - November 1979

A new studio monitor is described which can provide high-quality sound at sufficiently high levels to fulfil the complete range of monitoring requirements from pop to serious music. The design principles and criteria are outlined, and the performance of the prototypes is assessed both subjectively and objectively.

http://downloads.bbc.co.uk/rd/pubs/reports/1979-22.pdf


The design of the prototype LS5/9 studio monitoring loudspeaker
K.E. Randall, C.D. Mathers - August 1983

The prototype of a new medium-power studio monitor is described which is physically small enough to be accommodated in areas where a large high-power monitor would be inappropriate.
The main design and performance features are as follows:- A vented cabinet having a volume of 28 litres (1 cubic foot). Two drive units; a proprietary 34mm soft-domed high-frequency unit and a BBC designed 200mm low-frequency unit having a polypropylene diaphragm and a high temperature voice-coil. A low-level cross-over circuit feeds a 50watt stereo amplifier which drives the units separately; the cross- over frequency is 2.4 kHz. The axial frequency response is +/-3dB from 56Hz to 16kHz and the maximum sound level is 100dB(A) at 1m on axis measured in a typical listening room using light music.

http://downloads.bbc.co.uk/rd/pubs/reports/1983-10.pdf


On the design of loudspeakers for broadcast monitoring
C.D. Mathers - December 1988

By designing its own monitoring loudspeakers for about the past forty years, the BBC has achieved a degree of continuity and consensus in a subject where there are almost as many opinions as there are designers. This report describes the approach that has evolved within the Corporation towards design and assessment, and indicates how calculation and objective measurement are supplemented by experience and subjective judgement to arrive at a design that is acceptable to users in the broadcasting service. Each component of the loudspeaker is considered, and the problems of achieving consistency and reliability are addressed.

http://downloads.bbc.co.uk/rd/pubs/reports/1988-14.pdf

 

[Tony L]

Was the Quad ESL(57? 63?) a sort of BBC benchmark which manufacturers tried to achieve with box speakers?

I can’t remember the exact quote, but Quad’s Peter Walker was certainly very impressed by the BC1.

 

[mandryka]

. I’d have liked to compare it to the S20 in the 149 system too, but there wasn’t time.

That is a shame.

I wonder whether the Leak is better than the quad with the Tannoys. If I'm remembering correctly you prefer the Leak to the Quad with the Rogers.

Re the Radford, I'm convinced that it's a much more rewarding combination with my JR149s than my SP1.

 

[Tony L]

I wonder whether the Leak is better than the quad with the Tannoys. If I'm remembering correctly you prefer the Leak to the Quad with the Rogers.

Yes, I think I have them in the right systems, but to be honest the Quad and Leak are far more similar to one another than are the Audio Synthesis passive pre and the Verdier valve pre. Really I’m comparing two synergistic combinations; the passive pre suits the very high-gain Leak perfectly, the Verdier sounds wonderful with the Quad, and as the Verdier has the phono stage that dictates which combo is based in which system.

 

[Martyn Miles]

I can’t remember the exact quote, but Quad’s Peter Walker was certainly very impressed by the BC1.

I understood he rather dismissed moving coil speakers.
Looks like I’m wrong...

 

[Tony L]

I understood he rather dismissed moving coil speakers.

He had until hearing the BC1s. There is a quote that I can’t remember that he said to Spencer Hughes after hearing them at a show they were both exhibiting at, something along the lines of “if I knew you could get moving coil speakers to be that good I’d not have bothered...”, something of that flavour anyway. I suspect he was being kind, but still a good indicator of how highly he rated them.

PS As much as I like the BC1 the ESL is a better speaker IMHO.

 

[tuga]

It is perhaps not a coincidence that it was Derek, Spencer Hughes' son, who design the first Quad moving coil box speakers, the 77-10L.

 

[mandryka]

Yes, I think I have them in the right systems, but to be honest the Quad and Leak are far more similar to one another than are the Audio Synthesis passive pre and the Verdier valve pre. Really I’m comparing two synergistic combinations; the passive pre suits the very high-gain Leak perfectly, the Verdier sounds wonderful with the Quad, and as the Verdier has the phono stage that dictates which combo is based in which system.

Have you ever tried taking the Verdier out of the loop and playing the Quad into the Tannoys with no pre or the (presumably) more or less transparent Audio Synthesis?

 

[Tony L]

Have you ever tried taking the Verdier out of the loop and playing the Quad into the Tannoys with no pre or the (presumably) more or less transparent Audio Synthesis?

Yes, many, many times using the Audio Synthesis, which is about as good as a stepped attenuator pre can get (I actually have a pair of them, a Passion and a ProPassion). It sounds great. The Verdier sounds very different though. It is unashamedly a valve preamp and does bring all the things those of us who like valves like to the sound. It is big, bold and has a real presence that I like along with a total lack of grain or digital ‘sheen’. Basically Miles just sounds more ‘there’. That is what matters to me. Anything that gets me closer to that suspension of disbelief is what I’ll choose.

I’m sure detractors will argue it can only possibly be bringing colouration and limiting, and logically they may be entirely right, but in this system to my ears I just prefer it. I have a real interest in the technology and science behind audio and understand a fair bit of it, but I will always, every single time, trust my ears. I’ll still do the measurement bit as I’m always curious why something sounds better!

The ultra-minimalist single-input dual-mono Audio Synthesis ProPassion sounds staggeringly good with the Leak, so that is where it gets used. The Leak has all the valve magic one could want and is so high-gain/high-impedance it just doesn’t need anything more than an attenuator upstream. It is a great combination, as is the Verdier/303.

PS I can’t use the Verdier/Leak as a combo, even on the Verdier’s lowest gain setting it is far too loud for the Leak’s input and as such the Leak finds noise in anything upstream. I’m not prepared to modify the Leak to reduce input gain, and don’t feel I need to as the AS passive seems to suit it perfectly.

 

[Electrostat]

The crossover and response on the SHL5 "Plus" has been revised and the problem that you mentioned may no longer be audible, but I am also very partial againts metal domes and cones, which is why I went for the Stirling instead.

You hear zero crossover problems with the Stirling LS3/6? I wish they were easier to audition in the US :/

 

[Martyn Miles]

It is perhaps not a coincidence that it was Derek, Spencer Hughes' son, who design the first Quad moving coil box speakers, the 77-10L.

A friend has a pair of these 10Ls and I had a listen to them on my system.
I could certainly live with them.
He prefers his various pairs of LS3/5as.

 

[essgee]

Some interesting documents giving valuable background information for many of us to understand how loudspeaker development in the "early" years of "Hi-fi".

 

[Martyn Miles]

Yes, very interesting.
I’ve seen some before and also own some BBC info. collected over the years.
The ‘BBC approach’ has always interested me.

 

[Linus]

I’d personally put the Harbeth in another ‘influenced by’ category as the metal dome tweeters and Radial cone give it an entirely different sound. I know I’ll annoy many here, but they just don’t have the BBC’s midrange magic to my ears. IIRC they have MDF rather than ply cabs too. They just aren’t a BC1 or LS3/6 type speaker. They certainly have their fans, but I see them as a modern alternative/competitor in the same way say a Kef LS50 is to an LS3/5A. The only thing in common is the box size.

I couldn’t agree more having gone from a Spendor sp1/2e to a Harbeth HL5 and being. The Spendor had some magic to it and I regret trading it. Never got on with the Harbeth. To me there was something wrong and I was only too happy to see it go.

 

[Radfordman]

I have two. One I rebuilt myself using modern Dada Electronics boards, i.e. very nice glass-fibre boards, all populated with modern audiophile components etc. The other one is a simply amazing condition late-60s original that has had its electrolytic capacitors and trim-pots replaced and has rather larger coupling and PSU caps, but is otherwise totally stock, i.e. has its original carbon resistors, driver board transistors. Both have the correct output transistors etc. The latter (mainly stock) one sounds quite noticeably better to my ears, and that is the one that met the lovely Radford. I don’t understand why, but that particular 303 is a bloody good sounding amp!

Just tidying up here, found these, almost gold dust! forgot that I had them, bought them about 12 years ago from a guy who worked for Net Audio. They are used, but test fine. Correct output transistors for a Quad 303.

DSCF1107 by A60man, on Flickr

 

[Tony L]

I have a stash of them too! The ones I have no spares for are the ones in the little heatsink in the driver boards, I forget the number. I’d like to have a couple of those in ‘stock’ too.

 

[Radfordman]

I have a stash of them too! The ones I have no spares for are the ones in the little heatsink in the driver boards, I forget the number. I’d like to have a couple of those in ‘stock’ too.

Yes, I remember, I also bought from Langrex, I had forgotten that I already had stock! My age you know.

The driver transistors on the small black heatsinks, 38495 NPN and 38496 PNP, can be replaced with 2N5320 and 2N5322, that is what Quad supplied me with a few years ago.

 

[mandryka]

As far as I know only Tony has heard both an STA 15 and a Quad 303 with a Verdier pre. I don't own either an STA 15 or a 303 so I can't check whether, without that particular pre, the two amps are "different rather than better/worse" and whether the 303 has more slam with the Tannoys. It makes it so much more complicated that hi fi is holistic!

(I just note that my STA25 into JS149s doesn't feel short of SLAM in the music I listen to.)