Quadraphonic Amplifier 4 x 14 watts [Solid State Audio hi-fi Construction Projects]

PLAYMASTER 140--Quadraphonic Amplifier 4 x 14 watts

WE BEGIN THE DESCRIPTION OF A HIGH FIDELITY QUADRAPHONIC AMPLIFIER, WITH IN-Built SYNTHESISING AND DECODE FACILITIES, THE NEW PLAYMASTER 140 ALSO HAS FACILITIES TO ACCEPT "DISCRETE" 4-CHANNEL SIGNALS FROM TAPE OR CD-4 DISCS.

SPECIFICATIONS

Power Output (8 ohms): 16.5 W RMS with one channel driven; 15 W per two channels driven; 14W per four channels driven.

Power Output (16 ohms): with one, two and four channels driven - 10.5 w RMS, 10 w RMS, 9 w RMS.

Frequency Response: As per curve, within *2 and -2dB from 20Hz to 20kHz with tone controls at appro* center. Power amplifiers flat to 60kHz, then deliberately rolled off.

Compensation: RIAA for phone input. Other Inputs flat.

Sensitivity: Magnetic phono, 2mV into 50K nominal for 15W RMS output.

Other inputs, 150mv into 500K nominal.

Signal/Noise Ratio: Better than 60dB for all inputs, tested with input circuits open.

Cross-Talk: Better than 44dB at 1kHz for all channels with typical sources connected to the inputs.

Distortion: THD at 1kHz and ma* rated power 0.6pc. At typical listening levels (incl. noise component) 0.4pc.

Bass, Treble Controls: Nominally +14dB and -18dB at 50Hz and 10kHZ. (see curves).

Filters: -14dB at 20Hz and 10kHz.

Stability: Tested and stable into capacitance values across load up to 2uF.

We have been tossing about the idea of a do-it-yourself quadrophonic amplifier.

But, as often happens, we ran into problems of supply and approach which caused delay. Commercial amplifiers often use a master 4-gang volume control in association with a "joystick" pot system for balance. We could have obtained such items for our prototype but the supply position turned

Out to be dubious in the months ahead. Similarly for some of the other items. Weeks slipped by as we -- and the various suppliers -- waited for information.

Again, the more we looked at the project, the more we became convinced that easy answers were not necessarily the most appropriate ones. Constructors, facing up to a four channel amplifier, might want to avoid undue complication and expense but, at the same time, may not welcome too many compromises. A worthwhile design would have regard to future as well as present needs.

A variety of control and access and layout configurations were devised and discarded before we settled upon the one finally adopted.

As far as possible, it avoids specialized components, devious circuitry and constructional methods that would present problems to the homebuilder. At the same time, the approach and styling is modern, and there is opportunity to adapt to changing ideas and techniques.

---

At this point 1n the article, we would normally refer to the main circuit diagram and use it to explain our general approach. We can't do that on this occasion, because the circuit diagram of a complete quadraphonic amplifier is hard to draw and hard to read, particularly if compressed into the available page area.

Instead, we plan to rely on block and functional diagrams which, hopefully, will contain a lot of information of constructional value. The basic circuit lumps around which the amplifier is built are regarded as modules, each one to be the subject of separate presentation.

Not only should this simplify construction of the amplifier as a whole, but it will meet the needs of those who may want to tackle the project in stages. For example, by providing just the preamp, board, two power modules and the power supply, the amplifier could be operated as a normal 2-channel stereo system.

The additional channels and facilities could be added, as they are needed, or as finances permit.

But, before proceeding further, the simple diagram indicates the positions assumed for the four loudspeakers and the coding , adopted to identify them. When ultimately building the amplifier you will need to be constantly alert to avoid getting the channels mixed up. Different colored wiring will help, but we have also tried to build into the layout and diagrams a logical clockwise / progression which is: front left; front right; back right; back left.

Perhaps we should add that the listening room need not be as regular or as bare as the diagram seems to suggest. One of the attractions of quadraphonic sound is that it can fill a listening room with pleasantly dimensioned sound, without being over much inhibited by the-shape of the room or the disposition of other furniture. It is, in fact, a good deal less demanding in this respect than ordinary 2-channel stereo.

The general approach adopted 1s best appreciated by examining the accompanying block diagram.

A series of DIN sockets on the rear of the chassis accepts inputs from five possible sources: magnetic phono pickup, feeding directly into a preamplifier compensated for RIAA characteristic; radio, either mono or stereo; external stereo tape player, providing it has its own preamp / compensation; auxiliary, either mono or stereo; external 4-channel source, providing it has its own low-level circuitry. i We shall have more to say later about the external 4-channel input.

The precise sensitivity at the various inputs will vary somewhat with component tolerances and with operating mode. However, in general terms, the specification for the phono input is 2.0aV minimum into 50k Ohms for full output. This will more than meet the needs of current magnetic cartridges.

Other Input channels present an impedance of approximately 0.5 megohm and require an input of nominally 150mV for full output. These again are convenient and conventional figures.

On this occasion, we have not sought to make specific provision for ceramic phono cartridges, since magnetic types are now virtually universal in the high fidelity field.

The desired input is selected by a 5-positton "Select" switch marked as Sla and Sib. The switch has two other poles which we shall also refer to later but, otherwise, this much of the circuitry is conventional.

Conventional also is the left/right balance potentiometer which follows the Select switch, and the "Tape Out" provision from this same part of the circuit. Hade available via the Tape DIN socket already referred to, the tape output allows material to be copied from the amplifier, without being subject to the amplifier's volume and tone controls.

The tone controls employ conventional bass and treble cut and boost circuitry which can modify signals arriving via any one of the 2-channel stereo inputs. They are therefore effective for anything derived from these sources, including normal 2-channel stereo, simulated quadraphonic, or matrixed quadraphonic.

Output from the tone control stages passes to two poles of a "Mode" switch, designated as S2a and S2b. Some of our earlier effort was aimed directly at avoiding the need for such a switch and we had a scheme worked out involving interchangeable plugs at the rear of the chassis, which would have been readily adaptable to future needs. 8ut who wants to fiddle with plugs behind a chassis, when changing from one mode of operation to another? Whether by a socket system or a mode switch, the operating mode will almost certainly need to be changed, perhaps more than once in a typical listening session.

Output from the various signal sources may need to be directed to the power amplifiers for straight-through 2-channel stereo, or it may need to be diverted to one or other of the synthesizing or decoding systems, either in-built or external.

The Modes switch therefore needs to direct the available signal the circuit which is to receive it while, at the same time,

.the power amplifiers have to be switched to pick up their signal from that source. Poles S2a and S2b perform the first function, while poles $2c, d.e S f operate at the respective inputs of the power amplifier modules.

Before proceeding, it would be appropriate here to refer to the additional poles on the Select switch Sic and Sid, which relate to the provision for external 4-channel input. Such a provision is necessary if the amplifier is to cope with the discrete 4- channel signals which might typically be derived from a tape deck or from a future CD-4 record playing deck and demodulator.

The front channels present no problem, since they can be handled by Sla and Sib and fed through to the appropriate left and right front power amplifiers.

The back channels, on the other hand, have to be fed to the respective power amplifiers which, in all other modes, derive signals extracted from the "front" 2-channel sources. Thus the Select switch has to carry two extra poles so wired that, in the "Ext 4" position, the power amplifiers driving the back loudspeakers are connected through to the "Ext 4" signal source.

But the problem doesn't quite end here. The "front" signals pass through the tone control board and we would not want it any other way; it retains the bass/treble control facility, but it does introduce a 180-degree phase change in the signal.

A seemingly obvious course would be to duplicate the tone control facility, but this is one area where we accepted compromise.

Duplication would apply only for the back channels and only with discrete input signals. And, coming from the latest technology sources, these would hopefully be the signals least likely to need "doctoring".

The price would be duplication of the entire tone control circuitry, including the provision of additional and expensive ganged potentiometer elements.

Rather than commit constructors to this course, we suggest that the back channels be operated level in the "Ext 4" mode. The problem of the phase change can be handled simply by incorporating in the signal lines a relatively simple stage operating at slightly less than unity gain and virtually duplicating the tone control stage in this respect.

The difficulty might alternatively have been tackled by reversing the connections to the back loudspeakers in this mode only, but this might have led to other problems.

As it is, the phase inverting amplifier can be added much later, and only when the "Ext 4" facility is needed. Since it uses only a couple of small signal NPN transistors and a few standard wiring components on a piece of Veroboard, there will be no risk of the parts becoming unavailable.

At this point, we faced the decision of what modes to provide for in the prototype amplifier, having in mind present and future needs and the fact that firms selling kits and components would not want to stock variously marked front panels. What we have suggested should meet most requirements, while retaining a large element of flexibility.

In the fully clockwise position of the Mode switch (the position as drawn) the signal frcm the tone control stage is directed straight through to the front left and right power amplifiers via a 2-gang slide volume control. Using this much of the system, the amplifier works as a conventional 2-channel stereo unit, with no redundant circuitry in the line, and no "blend" to deteriorate left/right stereo separation.

The same pair of signals is made available to the back channel amplifiers via a second 2-^ang slide volume control. • You will note that phone jacks are provided so that the front and back amplifiers may optionally feed pairs of loudspeakers or stereo headphones.

Taken together, the foregoing paragraphs add up to a range of useful options in this first position of the Mode switch:

1. Double stereo in the one room, with balance determined by the volume control settings.

--------

2. Two-channel stereo at either end of a long listening room. If this is an important option, the links to S2e and S2f could be interchanged so that the right and left channels are in their proper positions when you face the "back" end of the room.

3. Two channel stereo in separate rooms with the level of each controlled by its own "front" or "back" potentiometer.

4. Mixed loudspeaker and headphone, or double headphone listening, with each channel independently adjustable.

In the second position of the Mode switch, the front signals again go straight through to the front power amplifiers, without passing through redundant circuitry. On this setting, however, the signals are also fed to the input of the EA 2/4 decoder, which can provide a measure of decoding for matrixed signals, and which does an excellent job of synthesizing four channels from 2-channel program material.

Output from the decoder goes to the back channels via the appropriate volume control.

The provision of separate slide type volume controls, mounted side by side, obviates the need for a separate front/back balance control, and also renders unimportant slight variations in the gain of the signal paths to the power amplifiers, in the various modes. It is very simple to operate the sliders together or to nudge either one up or down slightly to get Just the right effect.

The back channels can therefore be set very easily for "ambience", or a matching level, or even a dominant signal.

On the third position of the Mode switch, the signal is routed via a basic SQ decoder, about which we will say more later. This is the position you will normally choose for SQ encoded discs; you may or may not prefer it for other discs or for synthesized effects.

Quite deliberately, we have avoided marking the panel in a specific way. Instead, position 1 is marked "ST 2/4", the others "HAT1" and "MAT2".

The "ST2/4" position is the appropriate one to select for straight 2-channel stereo and also for discrete quadraphonic, when the Mode switch is set to "Ext 4". For this reason, we have wired the switches so that the "ST2/4" and "Ext 4" positions are adjacent and easily, bridged by a line drawn on the panel.

In fact, if you are not interested in the double stereo facility, the 2/4 decoder can be wired to this first position. The labeling is still valid and the straight-through frontal stereo is not prejudiced, since the 2/4 decoder does not blend the front channels.

By doing this, or else eliminating the 2/4 decoder altogether, a switch position can be cleared to select some other type of in-built decoder that might turn up.

Alternatively, we have provided for an octal or other socket of the chassis. on the rear While we have not actually wired up such a socket, it should be possible to connect to it output from the tone controls and Inputs to the power amplifiers, exactly as with the in-bu1lt SQ decoder. This done, it should be possible to plug in some future elaborate logic style decoder or a complete external multiple decode facility.

As far as the main amplifier is concerned,'all you have to remember is the significance of the markings "HAT!" and "MAT2".

Leaving the switching, you will notice that filters are shown between the volume control sliders and the inputs to the four power amplifiers.

To be perfectly frank, we are not enamored with the idea of Hi Lo filters of the simple kind, as they usually are. Kith a slope of 6db/octave, their actual effect is very little different from merely turning down the bass and/or treble response by the equivalent amount. However, Hi-lo filters are

. "in" and we have conformed to the trend. Physically, they involve a pair of push-buttons, each of which activate a 4-pole changeover function, sufficient to serve the four channels.

While we may sound rather negative about the provision, there are positive aspects. The buttons are certainly quick and convenient to use and, since they operate symmetrically in all four channels, they would provide a means of suppressing hiss or rumble in the "Ext 4" mode, where the main tone controls operate only on the front channels.

Again, their action is additional to that of the tone controls, so that use of both facilities makes available very heavy bass and/or treble cut.

A further point is that, if a constructor has a real need for a special filter contour, the controls are there and he can hang whatever circuitry he likes behind the panel.

The Playmaster 140 uses a common - and simple - power supply involving a power transformer, four rectifier diodes and three chassis - mounting electrolytic filter capacitors. This provides the required outputs: common, plus 21V and minus 21V. The supply is unregulated but is adequate, nevertheless, in the prototype amplifier it provided 21.5V under quiescent conditions,

-falling to 18.5V under the provocative test conditions of all four channels driven simultaneously and held at clipping point.

In terms of power output, this means that channels driven individually can be expected to deliver about 16W RMS, just' short of clipping. With all four channels so driven, the figure is likely to be at least 13W per channel. Under ordinary program conditions, it is reasonable to rate the amplifier as 15W RMS per channel, or 60V RMS total.

This is into the recommended 8-ohn loads. We did not run complete tests on the 140 into 16-ohm loads but, the output would probably be in the 8 to 10W region. Operation into 4-ohm loads is not recommended.

So much then for the broad concepts of the amplifier. What about the problems of constructing it? Basically, we have-tried to keep in mind the needs of the home builder and it may be relevant to say that the prototype was actually developed and constructed in the home situation during evenings and weekends. Host copies will typically be produced in the same situation, with the difference that individual constructors will be able to follow pictures and diagrams, rather than have to work out the tedious detail for themselves.

We have done our best to assist, in this regard, in the space available, but we must warn that construction of the complete amplifier is a fairly formidable task, and one that will absorb a lot of hours. If you have successfully constructed a normal stereo amplifier, you should be able to cope with the 140. But don't tackle it as your first substantial electronics project; it's too big a job for that! If you do decide to "have a go", you'll need to pay plenty of attention to the main wiring diagram, into which we have tried to cram a lot of essential information.

First point of interest is the wiring to the OIN input connectors. To the best of our knowledge they conform to current "standards". The connectors are drawn looking on the socket tags but, if you are in anyway confused, work to the numbers which are usually molded into plugs and sockets alike.

In the case of phono socket, pins 1 and 5 are bridged, allowing either a 3-pin or a 5-pin plug to be used. The chokes, intended -- to combat radar or other RF interference, involve a couple of ferrite RF beads, typically 3.5mm diameter and 5mm long. Loop about 5 turns of thin enameled wire through the beads; anchor the chokes between pins 3 and 5 of the socket and a tagstrip secured to the socket mounting screw.

All signal wiring from the input sockets to the rest of the circuitry should be carried out in figure 8 shielded wire with outer PVC covering, to prevent random contact between the shield and chassis.

It is absolutely essential to follow the earthing procedure suggested in the diagram. Failure to do so will almost certainly result in hum problems and possibly Instability as well.

Looking again at the phono input, the socket shell and pin 2 are shown earthed to the chassis at this point. The intention is that the incoming phono leads be earthed where they enter the amplifier chassis and that the internal circuitry be earthed back to - and only to - this same point.

As a first step to this objective, the braids of the figure-8 leads running back to the preamplifier input should be earthed to pin 2 of the phono socket and also to the earthy pattern of the preamplifier board. When you ultimately mount this board on its pillars, you will have to make sure that the pillars and/or screws are clear of the copper pattern.

For the sake of simplicity, all other input sockets can be wired with the braid joining to pin 2 and thence to the earthed shell.

However, there is one vital difference; whereas the braids from the phono socket join to the preamp earth pattern, the braids from the other sockets go nowhere at the far end.

When it is time to terminate these leads, split the figure-8 as necessary and snip them to the required length. Now, with a razorblade, cut through and remove about Jin of the outer sleeve.

Pull the remaining outer sleeve back, snip off the exposed braid and let the PVC covering slide back into place. Only the inner conductor should raw be visible, which can be stripped and soldered to the appropriate circuit point.

This procedure is necessary to avoid creating earth "loops" or parallel earth switches much easier because you are concerned only with the inner conductors. The shield ends somewhere back inside the PVC covering.

In short, no shields whatever are terminated or interconnected at the switch banks. The shields have to be earthed, of course, but this is done at the remote ends, where it is usually more convenient.

The wiring to the switches is probably the most tedious part of the whole job and the one where you are most likely to make mistakes. The switches are drawn in the same "ST 2/4" position as shown in the preceding diagram, and for the same options.

If you choose to vary from these, you will have to work out the revised connections for yourself.

The figure-8 shielded wire we used had red and white inner conductors, we adopted the convention throughout all the signal channels of "red equals right".

Also, because we weren't convinced about our own infallibility, we made a point of approaching the switch contacts with a loop in the leads, so that there was generally a small length of lead in reserve. It also allowed us to thread the leads past one another a little more easily and to push leads out of the way of the hot iron tip.

From the phono input socket, through the phono shield braids to the input end of the preamp board. the earth path is through the copper pattern to the output end of this same preamp/control board.

In fact, we secured a solder tag to the corner of the board adjacent to the volume controls and this can become the earth reference point for all the circuitry just ahead of the main power modules.

The decoders earth back to this point, as also do the volume controls and the input circuits to the power modules. As the diagram indicates, we used shielded wires for the signal leads to the 2/4 decoder, because it is mounted well away from the switches; non-shielded leads will suffice for the SQ decoder, which sits in a socket very close to them.

From the volume controls, the active and earthy signal leads run to the filter block, which will be detailed later, and thence to the respective power modules.

i The output circuit from the power modules involves a lot of wiring, mainly because we have again followed commercial trend and provided stereo output phone sockets for both front and back channels. The sockets are wired so that plugging in the phones automatically disconnects the particular pair of loudspeakers.

------------

This Involves the provision of stereo phone sockets In which the loudspeaker switching function is electrically Isolated from any part of the phone circuit. The phones themselves are fed . through series resistors, which should limit the available power to a convenient level, almost irrespective of the impedance of the phones. However, if you need more level in the phones, the resistors can be reduced in value.

As indicated in the wiring diagram, the phone series resistors and the parasitic suppression components for each channel can conveniently be mounted on a tagboard attached to the inside rear of the chassis, alongside the fuses.

While one side of the loudspeaker output circuit is nominally at chassis potential, it is essential that there be no DC path to the chassis, except via the copper pattern on the power modules and thence through the earth wiring already described. The phone socket does earth direct to chassis, for the sake of convenience but this is permissible because of the high impedance of the phone output circuit.

The same remark about earthing paths applies to the power supply, the only direct earthy path being through the RF bypass capacitors, which are there to help shunt any mains-borne RF direct to chassis.

The actual DC output (common, plus and minus) runs through separate leads to each power module. Thus, while the power supply common is at nominal chassis potential, it is so only by virtue of the path through the amplifier chain back to the phono input socket.

It is Important to note that the earthy side of the loudspeaker wiring returns directly to the filter capacitors. This minimizes the flow of output stage current through the earth link from the capacitor to module, and the resulting IR drop which would be imposed across the input circuit braid. It prevents amplification of any hum component and minimizes crosstalk between modules.

As already mentioned, the power supply itself is a very simple arrangement. The power transformer is rated at 15V per side at 2A. This feed into two pairs of rectifier diodes which we mounted, for convenience, on a small tagboard. The tagboard also serves as a convenient distribution point for the OC wiring. Note that there is provision for a link in the transformer CT return. A resistor of 15 to 20 ohms at 5W between the link points can limit the current in the case of an accident during the initial testing. It must, of course, be replaced by wire link once the amplifier is provisionally operational.

The filtering relies on three electrolytics which provide 3000uF across the -21V supply and 5000uF across the + 21V supply.

They reduce the ripple level to a faint hum which is just audible if you hold your ear directly in front of the loudspeakers); this with compact systems of average sensitivity.

If your loudspeakers are more than usually sensitive, or if you want to reduce the hum to an academic level, more capacitance can be added to the +21V line in particular. Whether this takes the form of an additional capacitor or high capacitance units which may turn up in somebody's catalog is immaterial.

Once final point: Note that the indicator lamp is wired back to the common line. Even though it is nominally a DC circuit, you will find yourself with an hum loop problem if one side of the lamp returns to chassis.

In the interest of economy and accessibility, we used a simple dish chassis, as pictured. Dimensions of the prototype are 38.3cm wide, 28.1cm deep and 9cm high. Such a chassis can be slid into a wooden case, if your preference or your skill lies in that direction. Alternatively, a simple metal fold can be slipped over the top and secured to the side flanges by self-tapping screws.

Such a cover would normally be dressed with a wood-grain adhesive cloth.

Cooling is not a critical problem with the particular output stages, but it certainly should not be ignored, particularly if you are likely to push the amplifier towards its power limits.

There is a pattern of holes under the chassis and, for these to be effective, the chassis must be stood up on rubber feet.

If a wooden case is used, make sure it has a cut-out to expose the holes, with feet on the cabinet instead.

A wooden case or metal top cover should also have some kind of venting at the top to allow free air circulation. The styling is a matter for individual suppliers.

In approaching the construction of the amplifier, we strongly advise that you install the components along the rear and front edges of the chassis, interconnecting them by wires laced together into cables. An alternative, which you might prefer is to push the wires through a series of rings cut from nylon tubing.

By doing this when the amplifier is little more than a shell, you will have room to work. If you do things the other way, installing the modules and then simply running the wires individually, you will almost certainly end up with a disorderly tangle.

Install the DIN input sockets first, all with pin 2 towards the bottom of the chassis dish. The phono socket is at the end with a 3-lug tag secured under the mounting rivet adjacent to pins 3 and 5. The other sockets follow in the same order as in the main diagram. We used a hand riveter to attach the sockets but screws and nuts can be used instead.

To fill the central socket hole, as yet unused, we mounted an ordinary octal socket.

Next come the four polarised loudspeaker sockets. If you plan to use rivets, make sure to get the sockets with metal, not molded, flanges. The larger "earthy" pins go towards the center of the chassis.

Alongside the loudspeaker sockets are the fuseholders, which need to be fitted with 1.5A fuses.

Then comes the power cord, which will involve an access grommet, a 3-hole junction block and a cord clamp.

Also attached to the inside of the rear chassis face by lin stand-off spacers is a 20-lug section of tagboard which accommodates the output shunt and headphone feed components. This can be made up and mounted temporarily in place, ready for the associated wiring looms.

The remaining item on the back panel, a phase reversing amplifier for 4 -channel external input, can be ignored for the time being and will be covered later. Just make sure that you leave space for it.

Turning now to the front of the chassis, we elected to use slide potentiometers for a variety of reasons: styling, convenience of control and electrical tracking. Those shown have a travel of 4.5cm. The volume controls are 50k log, the tone controls 500k linear, and the balance control a 2meg linear. This latter value was chosen by the way, to maintain a high input impedance for the high level inputs. A lower value of balance pot would give smoother control, with some loss of input impedance and overall gain. We didn't make the change in the prototype amplifier because, in practice, the balance control is not used a great deal anyway.

Mounting the slide potentiometers does present something of a problem, in that slots are more difficult to produce and align than simple mounting holes.

More than that, the slide potentiometers which we used were supplied with Special semi-roundhead screws, not intended to be exposed on the front panel, yet not suitable for counter-sinking.

In any case, if the potentiometers were mounted directly to the front panel, too much of the shafts would be exposed.

What we did was to bend up two brackets from scrap 16g aluminum, shaped to accommodate the five slide potentiometers. Bolted by a flange to the bottom of the chassis, they support the potentiometers about 1cm back from the panel. Most likely, similar, brackets, will be supplied as part of the chassis hardware but they aren't hard to contrive with ordinary bench facilities.' Provision of the bracket obviates problems with mounting screws, allows fore and aft adjustments of the control arms and also - permits ready replacement of a potentiometer, should it ever be necessary.

A variety of push-on knobs, is available to suit slide potentiometers but, in our view, the common large black rectangular knob best suits the styling. It measures about 2-cm wide and 1 cm high.

One point we should stress: Make sure that you mount the volume controls with the low resistance segment of the element towards the bottom, otherwise the logarithmic taper will be working against, rather than for, smooth control. It may be wise to check this with your multimeter before mounting the controls finally in position.

The Select and Mode switches mount directly to the front panel, although you may have to pack them back with at least one washer to avoid having too much of the thread protruding. It would be wise at this stage also to select suitable knobs, preferably with a skirt to hide the nuts, and to cut the switch shafts to the proper length. Mount the switches so that, viewed from the back, they occupy the positions shown in the main diagram.

The filter block at the other end of the front panel uses two pushbutton switches, spaced 35mm between centers and each providing a 4-pole changeover function. These have to be set back on spacers about 1.5cm, so that only part of the button is exposed through the front panel.

We suggest that you obtain a version of the switch assembly with a backplate fitted. It stabilizes the switches and also provides a convenient insulated common point for the "earthy" circuitry.

Details of the filter wiring are given in the accompanying drawings. Note that the contact pins pass straight through the body of the switch and it is purely a matter of convenience whether connection is made to the pins on top or underneath. With larger components it may be different.

In fact, if you want to save time and money, the amplifier can be put into service without the filter switches. Simply route the signal lead direct from the volume control tapping to the power amplifiers. However, we do suggest that you make the lead a little longer than necessary so that it can be re-routed via the switch assembly at a later date.

Alongside the filter switches is a filament type indicator lamp fed from the DC Supply. Such lamps are typically rated at 6V, 40 or 60mA and can be supplied through a 5W resistor of about 470 ohms. You may have to experiment with the resistor to get adequate brightness, while conserving current - and lamp life.

Below the lamp is an ordinary mains type off-on toggle switch.

Indicator lights and switches vary from miniature affairs to , quite bulky ones. We suggest that chassis makers provide only small holes in the panel and escutcheon, sufficient for the smallest components. If you fancy something larger, or have no real choice, the holes can easily be enlarged as necessary.

The remaining items on the front panel are two stereo headphone sockets. These are rather special types. They provide connection to both headphone actives, remembering that the convention is to wire the left channel to the plug tip.

But the sockets also provide two switch functions which MUST be electrically insulated from the headphone circuitry. When the headphones are plugged in, two pairs of contacts open which are wired to break the circuits to the respective loudspeakers.

Provide a suitably marked escutcheon to fit in front of the amplifier chassis panel. It will be locked in place by some of the controls but, before it is finally mounted in place, it lay be wise to anchor the corners and center with blobs of suitable cement to prevent it from being kinked.

Next logical step is to provide mounting for the four power modules. Many options were open to us when planning the prototype but we settled for the one shown. It will involve you in a little bench-work but should be inexpensive and with the advantage of good accessibility. , You will need two scraps of particle board or wood 17.5 x 4.5 x 1.5cm and a couple of scraps of composition board about 17.5 x 2.5 x 5cm. Run saw cuts part way through the larger pieces wide enough to receive the edges of the modules, drill holes ¦ for the incoming wiring and fix the thinner pieces to the bottom as footings to keep the modules clear of the chassis. Details are given in the accompanying drawing.

For the sake of appearance, we sprayed the prototype supports flat black.

They are held in place by two countersunk screws in each, driven up through the bottom of the chassis. The rear one in our case was mounted 3cm from the rear of the chassis and about 2cm from the end. Exact positioning will, of course, depend on the boards and your saw cuts but it is well to attend to the matter at this early stage.

The next item to assemble is the 9+9 tagboard, which carries the four silicon rectifiers, and provides convenient distribution for some of the power wiring. Rectifiers could be used with a PIV rating of 100 or more and a current rating of 3A or more.

The tagboard should be stood off the chassis on 1cm or 0.5 in pillars.

At this stage, it is probably appropriate to plan, make up and install, the looms which account for most of the wiring in the amplifier. The rest of the wiring can be added as the modules are inserted.

To assist in the task, we have prepared a wiring diagram which shows the nature and position of the looms.

First off, seven pairs of figure-8 shielded and insulated leads are necessary from the input sockets to the Select switch, with a couple branching out halfway to the preamp board, leave a fair amount of slack at each end when you make up the loom so that the leads can loop naturally to the points where they are supposed to terminate. Be careful to observe the requirements in regard to termination or otherwise of set out the shield braid.

Depending on how economically you work, it is likely that you will need approximately 7m or 8 yds of twin shielded cable for the amplifier as a whole.

Another simple loom runs from the switches to the volume controls and to the 2/4 decoder. This can tuck conveniently into the space between the pot bracket and the front panel, just above the travel of the pot arms.

Incidentally, you will need a multimeter when wiring in these various looms, since it is difficult to trace leads once they are laced securely and laid into position. The color coding will enable you to identify left and right channels but the meter will be necessary to sort out the cables.

Turning now to the non-shielded circuits, a 16-lead harness needs to be made up to cope with the mains switch, indicator lamp, and the headphone sockets. We suggest you make an effort to get a few feet of as many colors as possible for this one, otherwise you will have real fun sorting them out. Either that, or the wires will have to be soldered in place as neatly as possible and then laced.

---------

At top Is the information you need to make the supports for the power module boards while below are the circuit end wiring diagram for the filter switch assembly.

---------------

----- - SHIELDED PAIRS TO SWITCHES

By the time you have made up these loons, installed them and completed as much of the associated wiring as is possible you will, in fact, have completed most of the wiring in the amplifier. What is more, it should look tidy and disarmingly simple! It remains Only to add the power Supply components and the main chassis dish should be just about complete, ready for the modules, which have yet to be described.

In the prototype we used two 2500uF 35VW electrolytics, connected in parallel for the positive supply, and a single 3000uF unit for the negative supply. These are simply bolted in position, using countersunk screws. As mentioned earlier, electrolytic capacitors seem frequently to change in shape and rating.

Variations should not cause concern, however, provided alternative capacitors fit, have adequate voltage rating, and not less than the required capacitance value.

The power transformer must deliver an output of 15-0-15 volts under quiescent conditions, but it must exhibit only modest voltage drop as the load peaks up to about-100 watts with all channels driven simultaneously.

In the prototype transformer, the mains lugs were exposed at the top and we have suggested that the assembly be modified to keep them on the underside. If they don't do this, it would be wise to cover the lugs in some way to minimize risk of shock.

One final point: A 0.01uF ceramic capacitor (100V or higher) should be connected from each end of the 15V secondary winding to chassis to bypass mains-borne RF energy.

With these remaining components installed and wired up, attention can be turned to the various modules.

While the power amplifier modules have been derived from the original Playmaster 136 design, there are important differences.

The most obvious has already been mentioned, in that the new board does not carry any power supply components. Power input is by three wires from the loom mentioned earlier; power supply earth or common, plus + 21V and minus -21V. A fourth lead from the same end of the board feeds the active side of the loudspeaker circuit. The remaining connection is the shielded signal input lead which also provides the earth link back to the preamp board.

Adjacent to the plus-21V lead is a link into which a milliammeter can be inserted to measure the quiescent current of the output stage. For simplicity, we used a loop of hook-up wire with a soldered joint in the middle. In the finished amplifier it rests on the top of the support bracket, our of harm's way.

The passive components, resistors and capacitors on the wiring diagram require no special comment.

Note, however, that we have added one resistor to the amplifier, a 6.8k bridging one side of the quiescent current adjustment potentiometer. This is to protect the output transistors in the event that potentiometer wiper or element becomes open-circuit.

What happens in this circumstance is that the quiescent current control transistor, 2N3565, is turned off and the output transistors are turned hard on, drawing heavy current which can cause them to burn out.

While the possibility of an open circuit potentiometer is fairly remote, the 6.8k resistor provides cheap insurance. Now, in the event of an open circuit pot, the control transistor is turned on and the output transistors draw zero quiescent current. In this condition, cross-over distortion occurs but no damage eventuates and the situation can be rectified.

With the 2N5088 and the 2N4250, it is just a matter of bending the leads sufficiently to fit into the triangular pattern of holes, the transistors sitting about a centimeter above the surface of the board.

The real difficulty has to do with the other three which need to make physical contact with the output transistor heat sink to provide thermal feedback.

The TO-9 style transistors don't lend themselves to this approach. They don't sit down snugly on the board and the small, flat top doesn't mate naturally with a dimple. We are therefore suggesting that holes be drilled in the heatsink, which will be a clearance fit tor the TO-92 bodies, allowing the transistor to sit part way through the heatsink. A blob of silicone compound can blend the two thermally.

The mounting method has a possible bonus in that the transistors are no longer trapped under the heatsink. They could, if necessary, be extracted and replaced through the holes.

Turning now to the actual construction of the power modules, the first step is to inspect the heatsinks, which should be of aluminum, not less than 16 gauge. Make sure that the two power transistors Sit flat against the surface, with mounting holes aligned and with adequate clearance around the base and emitter pins. If there is any inaccuracy, lead the holes as necessary with a small round file.

In fact, we rubbed the inside surface of our own heatsinks with a large flat fine-gauge file to remove any high spots and then buffed the surface all over with steel wool.

Now check the heatsinks against the wiring board. Make sure that the bolts securing the power transistors can pass straight through, and the pins likewise. If there is any fit problem, the mounting holes in the board can be elongated as necessary.

If the small-signal transistors happen to be the older glob tops, the heatsink will need dimples or countersunk holes in the underside. If, as likely, the transistors are of the TO-92 configuration, the heatsink will need snug clearance holes instead. In fact, dimple type heatsinks can be adapted by drilling appropriate holes. They may not be concentric with the dimples, however, since the TO-92 transistors sit most naturally between the collector and emitter pins, with the base lead kinked outwards to fit the triangular pattern in the board.

With all this sorted out for each of the boards, the heatsink assemblies can be completed. Smear the underside of the power transistors with silicone compound and secure the transistors firmly to the heatsinks with l-inch long bolts and nuts, either 0.25 Whitworth or 5BA. The transistors do not need to be Insulated from the heatsink. In fact, the heatsink and mounting bolts form part of the collector circuitry.

You will need four spacers per board, either Aim or 5/32, such that when the heatsinks are mounted, the base and emitter pins of the power transistors just come through the copper pattern. We found some brass nuts of a larger size which we turned into spacers by running a Jin clearance drill through them.

This done, the power transistors assemblies can be put aside for later installation.

On the wiring board itself, it is wise to smear the copper around the heatsink mounting holes with a thin layer of solder.

If this is done, the nuts will bite into the solder when they are finally tightened and make good contact between the transistor collectors, heatsink and mounting bolts and the copper pattern on the board.

At this stage, the remaining components can be installed as per the diagram. Use an iron with a clean, slender tip and flow the solder around each component lead as quickly as possible, to avoid overheating either the pattern or the component. Hake sure that you install the electrolytics with the correct polarity. This done, the three "2N" transistors contacting the heatsink can be slipped into position without, however, soldering them to the pattern. Now slip the heatsink assembly into place, passing the power transistor leads through the board, and the three small transistors up into the clearance holes. Secure the heatsink.

Raise the TO-92 transistors by a small amount and smear silicone compound around them. Pull them back down into the holes so that they are about 1mm proud of the surface and solder them into position.

Flow a generous collar of silicone around the head of each.

Solder the pins of the power transistors, check everything over carefully, and the module(s) should be ready to drop into position in the chassis.

Before doing this, however, the basic chassis assembly should be activated to make sure that the power supply is delivering the correct voltages, etc. We show a position on the power distribution board where a resistor can be inserted temporarily in the transformer CT lead. Something around 15-22 ohms, 3W could be used; it will not offer complete protection against mishaps but it will limit the current if there is something wrong between either side of the supply and common.

Having made sure that the supply wiring is complete and that no trailing wires are resting against the chassis, plug in and switch on.

If all is in order, the indicator light should come on and voltages, plus and minus 21.5 approx., should appear across the respective filter capacitors.

You are ready to install and check the power modules, one at a time.

But first a word of warning:

While the dial lamp will discharge the negative supply capacitor after switch-off, the positive line filter may retain its charge for many hours. If, while working on a module, you discharge the capacitor through the wrong path, one or more transistors can be ruined. Remember to discharge the capacitor before you dive in with a soldering iron.

Better still, wire an oddment resistor (say 580 ohms 5W) temporarily across the positive filter, so that it will discharge automatically.

Install the first module, connect it up as necessary, and insert a milliammeter in the link, with plus to the supply, and set to the 250mA range. Rotate the current set potentiometer, as viewed in the chassis fully anti-clockwise, and set the volume control pot at full off. Since the earthy side of the pot may not at this stage have an earth return, run a temporary link to a chassis earth.

Now watch the current meter and switch on. If the meter slams over, there is something radically wrong. Switch off instantly and check.

You may have the power transistors interchanged, or one of the other transistors the wrong way round. Whatever you do, don't tempt fate by switching on again and repeating the overload condition -- whatever is the cause.

In fact, the current flow, with the preset pot retarded, should be 0. If it is, reset the milliammeter to 50mA and carefully zero rotate the potentiometer clockwise. Bring the current up to 12mA and leave the module run for a few minutes. If all is well, switch off, remove the milliammeter and close the link.

Note that the current should be set without a loudspeaker or other load connected. When the loudspeaker is plugged in current distribution in the output stage will change due to the small offset voltage (0.2V approx) across the loudspeaker terminals.

By now connecting a loudspeaker and feeding a signal to the volume control from any source capable of producing a reasonable signal across 50kohms, it would be possible at this stage to check the module for sound, or yet again to run instrument tests.

Other modules can be added progressively, and similarly checked.

It is far better to do this than to wire all the modules in and switch them on simultaneously. One with an inadvertent fault could be "cooking" for several minutes while you adjusted the others! Just before proceeding, a point could be clarified to advantage:' The first transistor in the power module is shown as type PN5088. It should read 2N5088. Despite the "2N" prefix, the transistor is in the general T092 class and uses the connections as shown on the circuit diagram.

Now to continue:

Having installed and checked the power modules, the next obvious step is to build and install the preamplifier and tone control module; this is secured to the chassis floor on four pillars in the space between the power modules and the panel controls.

For your guidance, we show the schematic circuit of one channel of the preamp-tone control module.

The compensated preamplifier uses two selected gain silicon transistors - a high gain low noise type (BC109C or BC549C, &c) and a medium gain BC108B or BC 548B, Sc. The types specified are the ones most likely to be supplied but the "Sc" indicates the existence of possible substitutes.

However, in addition to the electrical characteristics of possible alternative types it is important to take note of the connections.

The board is drilled for transistors having the traditional triangular CBE lead configuration. Transistors of this type can simply be dropped straight in, with little risk of confusion.

If you are supplied with T092 style transistors of another brand, check the base connections carefully. For example, in the NS range, the BC548 does not have a cranked base pin and the lead sequence is reversed in relation to the flat on the body.

Electrically, the preamplifier proves enough gain (75 times) to ensure full drive to the main amplifier from an input of 2mv RMS, with good signal/noise ratio and adequate tolerance to peak level input signals. It provides a nominal loading of 50k ohms for a magnetic cartridge and compensation which conforms closely with the required RIAA characteristic.

Output from the compensated preamplifier goes to "Select" switch, where it is made available, along with signals from other sources: radio tuner,- tape player, auxiliary input or external 4-channel input. Since the signal levels at this point are normally 150mV or higher, shielding is not a critical requirement. It has been specified for long lead runs, but the switch banks, shorter leads and other associated circuitry are open, relying only on sensible layout to avoid instability problems.

From the Select switch, the signal passes to an emitter follower, which presents a fairly high impedance to the various signal sources, and a low source impedance for the tone control circuitry. At the same time, it provides a convenient point from which to derive signal for an external tape deck; this signal is independent of the amplifier's own volume and tone controls.

The tone control circuit is of the feedback type, favored because of its inherently low distortion and its tolerance to higher level input signals.

The point should be made, however, that signals derived from the Select switch are applied directly to a transistor base and can overload the circuitry between this and the volume control, if the level is excessive. Signals from radio tuners, tape players *c, must be limited in some way if there is any suggestion of overload, or if you find that the amplifier is being fully driven with the volume controls only fractionally on.

In constructing the preamp-tone control module, a useful first ,\ step is to check the fit of the board and chassis mounting and make any adjustments necessary by filing holes, &c. We used threaded pillars, with 0.25" long 0.25" diameter countersunk screws securing them to the chassis and J" long 1" diameter roundhead screws securing the printed wiring board.

However you arrange it, remember that there must be no direct circuit between the copper pattern and chassis. The only earth must be via the phone input braids back to the phone input socket.

To provide a common earth point for other circuitry, we suggest you add a solder lug to the corner of the board adjacent to the filter switches, bolting and soldering it to the earthy copper pattern. For the rest, the components drop Into place, as per the accompanying drawing. Polarity is important only in respect to the electrolytic capacitors, and this calls for some care. Note that we have shown an alternative position for the main decoupling capacitor, in case it happens to be larger than the one originally planned. An extra hole may be required and, for this, you will need a fine twist bit and preferably, a small "egg-beater" hand-drill.

As indicated, an extra hole will also be required for the shield braid of the leads running back to the "Mode" switch. He have suggested that constructors cope Individually with these minor matters to save us originating a new board pattern for the sake of two or three extra holes.

The most tedious job with the module has to do with the leads which must ultimately connect to the potentiometers and switches. The leads range from about 3 to 7 inches long and are unshielded, except where otherwise indicated. It is a good idea to use as many different colors as possible to facilitate lead tracing and to use thin rather than thick hook-up to retain maximum flexibility.

The leads should be anchored to the wiring board during initial assembly and left trailing. Please yourself whether you make them all generously long, or save wire by cutting each one discretely to suit the requirement. But, whatever you do, plan for each wire to follow a gently curved path so that the board can be unbolted and lifted up for testing or service. For the same reason, leave some slack in the phono input and tape outputs leads, which run back to the DIN sockets.

With the module in position, wired and checked, and connected to the plus-21V supply line, the amplifier could operate in normal stereo mode from phono or other inputs. The signal and earth paths should have been established from the phono input socket, through the copper pattern on the preamp-tone control module, thence to the Hi-Lo filters and volume controls, and to the input connections to the four power modules.

Just make sure that any other temporary or accidental earth path has been obviated.

One other point we should mention: If you do operate the amplifier 1n this semi-finished state, remember that you have two volume controls and four power modules. While operating one volume control, two modules and two loudspeakers, it is all too easy to have the other volume control advanced, and be overdriving the other power modules without knowing it. It wouldn't necessarily cause trouble, but care 1s usually preferable to lament! If this much of the amplifier Is working normally, 1t will be driven to full listening volume by an average magnetic cartridge, with the volume control(s) about halfway up, in terms of physical movement.

Frequency response will be level with the bass and treble controls at about half setting though, in our prototype, they needed to be about 1' above center for a nominally flat response.

In these circumstances, lifting the pickup off the record and leaving the controls untouched, the amplifier should sound dead quiet from the listening position. Only if you put your head right up to one of the loudspeakers should any hum or hiss be audible, and then only faintly, with loudspeakers of the usual modern compact design.

The Stereo-24 Adaptor is Intended to process signals already being fed to the front channels and from them to synthesize signals for the rear channels.

With this in view, the left and right signals are made available respectively to two NPN transistors, operating with equal loads in the collector and emitter circuits. It is, in fact, the old time phase-splitter configuration which provides a pair of out-of-phase signals, each about 0.9 the amplitude or the original input.

A network of resistors merges the four outputs in such a way that a "mono" or " center" signal that is common to both channels is partially cancelled; as a result "difference" or ambience signals become relatively more prominent. These "processed" signals pass through the upper transistors in the schematic circuit to their respective rear channels.

For example, a signal in the left front channel will be applied to the base of the lower left transistor and will appear at slightly reduced amplitude at the emitter. If it is unique to the left channel, it will be fed through the 47k resistor, with a further loss of amplitude to the base of the upper transistor feeding the left rear channel.

However, if a similar signal is present in the right channel, an out-of-phase replica will appear at the collector of the lower right transistor and this, via the 82k resistor, will tend to cancel and diminish the "left" signal which actually reaches the upper-left transistor.

The same will, of course, apply for what reaches the upper right transistor.

This partial cancellation forms the basis of most of the currently popular systems which synthesize 4-channel sound by loudspeaker matrixing. The Stereo 24 Adaptor incorporates a further subtlety, however.

The rear channel feed transistors which are also connected as phase splitters and the signal is taken from an R/C network strung between collector and emitter.

At low signal frequencies, where the reactance of the capacitor is high, the output comes effectively from the emitter and is in phase with the original front input signal. This means that low frequency energy from all loudspeakers will tend to be in phase, ' avoiding the thin" sound that might result if it were otherwise.

At high frequencies, the signals come from the respective collectors while, in between, they are derived from both collector and emitter in a phase which varies with frequency between 0 and 180 degrees.

The networks feeding the respective rear channels are deliberately different, such that there is an approximate 90-degree phase difference between the rear channels over the middle audio range. The partial cancellation and the progressive phase rotation in the rear channels, relative to the front and relative to one another, minimizes any tendency for the system to produce a firm image of "nono" material at the sides or rear. Thus center front signals tend to "stay put", while the side and ambience signals are dispersed, subject of course, to the level in the rear loudspeakers.

---- Above is the circuit diagram of the power module. May be constructed on Vero board.

--------- ------- The circuit of the SQ decode module, complete except lot the decoupling components, which were included in the main cheats wiring. The components shown with en asterisk should preferably be 5 pc tolerance types, since they determine the phase characteristics on which the decode function depends.

---------- STEREO-24 DECODER. May be constructed on Vero board. Circuit diagram of the phase change module.

----

Experience has shown that the system works out very well indeed in synthesizing quadraphonic sound from existing 2-channel recordings, it can also produce some very compelling effects from matrixed discs, even though the energy distribution may not be in line with the producer's original idea.

It was our original intention to derive the left and right signals for the Stereo-24 Adaptor from the output circuits of the front channel power modules - as was assumed when the unit was first devised. It was for this reason that preset input pots were provided on the board.

However, on this basis, the front volume control would have affected all four channels, making operation in this mode different from any other. To avoid this, the Stereo-24 Adaptor now picks up its signal from the tone control output. The preset pots are simply turned full on; in fact, they could be replaced by resistors and the signals fed directly to the coupling capacitors.

The effective gain through the Adaptor is less than unity but the imbalance between front and back channels is of no consequence in practice. If you want full "surround" sound, it is simply a matter of nudging the rear volume control a fraction of an inch above the front control; there will always be gain to spare.

Assembly of the unit should be a simple matter, using Vero-board.

Two figure-8 shielded pairs need to be attached, which run back to the Mode switch in the space between the pot mounting bracket and the front panel. One pair conveys signal to the unit, the other from the unit back ultimately to the rear channel amplifiers. The remaining leads are the positive 21V lead to the main filter bank and the earth lead back to the lug on the preamp-tone control module.

Placement of the Stereo-24 adaptor is fortunately not critical and , we were able to locate it in the space behind the Hi-lo filter switches.

To mount the unit, we made up a small angle bracket which was secured to the adaptor, on the copper pattern side, but stood off by 0.25 in spacers. This was secured to the floor of the chassis with countersunk screws, with the edge of the board just clear of the metal. The potentiometers face outwards although, as we have already mentioned, they are not strictly necessary in this application.

One point about the pots: they are mounted in a mirror-image configuration and rotate in opposite directions for the same function. For full on, both wipers should be turned towards the center of the wiring board.

And there it is: with the Stereo-24 Adaptor built and installed, you should be able to synthesize four channel sound to your heart's content fro* existing recordings, whether disc or stereo tape. And you will also be able to cope reasonably well with matrixed quadraphonic discs.

Two other modules remain to be discussed, one to change the phase of rear channel signals being fed into the amplifier via the Ext-4 input, the other a decoder expressly for SQ matrix quadraphonic discs. While Its immediate role relates to the Playmaster 140, it could in fact be used to advantage with any quadraphonic system that you might be putting together.

As explained, rear channel signals fed into the Playmaster 140 through the "Ext-4" input do not pass through the tone control stage, being reproduced "flat" unless deliberately modified by the Hi and to filters. As a result, the signals are not subject to the 180 degree phase change that the tone control introduces and this could lead to difficulties. We have therefore provided a phase change module for the rear channels: strictly speaking, it will be necessary only when you get around to using the Ext-4 input.

Fortunately, the phase change module can be a relatively simple device, devoid of adjustments. As long as the gain approximates unity, any slight difference between the front and back channels will be taken up by the volume control settings, without the user even being aware of it! As indicated by the schematic circuit, the respective rear signals are fed to two small-signal NPN transistors with equal collector and emitter loads. Since a phase change is required, output is taken from the collectors and made available to the Selector switch.

With the split load configuration, the gain through each stage will be slightly less than unity but balance between stages will depend almost entirely on the 5.6k resistors.

If you want to be fussy, the gain of either stage can be nudged upwards by using a slightly higher value of collector load and/or shunting downwards the emitter load. Increasing the emitter load and/or decreasing the collector load will reduce the gain.

Whether all this effort would be worth-while is quite another matter! We built up the phase change module on a scrap of Veroboard 9 tracks wide, measuring 37mm, wide by 67nm long. The circuitry was concentrated on 7 tracks, leaving the Outer two tracks for mounting. We bent up an angle bracket from a scrap of light aluminum and mounted the unit in the space alongside the loudspeaker sockets.

While shielded lead was used to convey signals to and from the module, the level and impedance of the signal circuitry is such that no other shielding proved to be necessary. A positive lead is necessary for the plus 21V supply but the negative return can be via the shield braid to chassis.

The decode module is very compact being assembled on a Vero circuit board measuring 76 x 60mm. It operates normally from a supply rail of about 20V and can be cut into a typical system either just before or just after the volume controls.

Fed with left channel and right channel signal, it will deliver outputs for all four channels of a quadraphonic system. While its prime purpose is to decode SQ matrix recordings, it will produce a multi-dimensional output from other matrix recordings, and doubles as a 4 channel simulator for two-channel stereo recordings.

Distortion is low, noise level is low and the gain from either input to any output is nominally one. It dropped into the Playmaster 140 system without the slightest hint of trouble.

It may be helpful at this point to explain a little of the background.

Although foreshadowed by Blumlein In 1931, the basic idea behind matrix style quadraphonic recording received little or no publicity until it was brought forward by Peter Scheiber in 1969.

Subsequently, there was much argument about the validity of the scheme but it came to be seen as a compromise which had definite commercial possibilities. Research then went "underground", to emerge later as a whole array of different and competitive matrix systems.

Such as the resulting confusion that the whole idea of matrix quadraphonic was at hazard in the market place and, under this kind of pressure, rival companies managed to compact them into two main systems:

SQ from the Sony/CBS group and RM (Regular Matrix) embracing the remainder.

It was a step in the right direction but, in practical terms, it wasn't big enough. The circuitry to cope with even two systems was still too clumsy to fit neatly into consumer equipment.

A short while later we had on our own test bench the working prototype of a switchable decoder, using discrete components. However, because of the uncertain status of quadraphonic sound, the complexity of the decoder and the difficulty of merging it with other gear, we discarded the whole thing and settled for our much simpler, though limited Stereo-24 Adaptor.

Meantime, however, SBS Laboratories in America, in collaboration with Motorola, had come up with a single integrated circuit and some passive components - necessary for a basic SQ decoder. With the addition of a few peripheral components, it made it possible to produce a decode module compact enough and inexpensive enough to slip into a unit amplifier or receiver.

There was just one problem: The IC in question was Intended primarily for equipment manufacturers; it could be sold only under license, and upon payment of a royalty to CBS So much for the commercial and legal constraints. The IC itself transforms a basic SQ decoder from what would otherwise be a large array of discrete components into a very compact module.

But why the specialized treatment for the SQ system? Mainly because the CBS/Sony group have had the initiative to originate and propagate such an IC, along with the appropriate application data.

The other point is that, for the time being, most of the currently available quadraphonic recordings are encoded SQ.

The module provides basic decoding of such recordings and the channel, separation available is that inherent in the system. Separation between i front left and front right is preserved but separation between front and back signals is only 3dB.

CBS engineers point out that subjective improvement results from some deliberate blending of the front output signals (10 pc) and the output signals (40pc). This imposes a limited of 20dB on the left/right front separation but it Increases the front jack separation to 7dB.

To achieve greater orders of separation than this, Sony in particular have devised a number of "logic" circuits, so-called. The function of these circuits is to continuously sample the signals at each of the outputs from the basic matrix and by combining the samples in suitable phase, produce a resultant which relates to what is apparently the dominant signal at any instant. The resultant is used, in the manner of an AGC circuit, to increase the gain of the relevant amplifier channel(s).

By this means, a signal which may have only a marginal emphasis from the matrix, is augmented in the amplifier chain, thus increasing the apparent separation and decreasing the apparent cross-talk between channels.

While there is room for argument about the merits or otherwise of logic circuitry, it is mainly of academic interest, at present, to the average home constructor. Effective logic circuitry is far more complex than that necessary for decoding and at this stage, we are happy to settle for the basic system.

The Playmaster 140 provides for external units and for an external full logic decoder, if and when you are tempted to buy one! The basic circuitry surrounding the MC1312P IC is taken directly from the Motorola application data. No details are given of the internal circuit but the data suggests that IC contains several amplifier stages, with the appropriate interconnections, and some resistive elements for the Wien bridge type phase shifting networks.

The remainder of each network is provided by the peripheral components.

Since the design is the result of collaboration between CBS and Motorola, we saw no reason to question the values suggested.

Most of the capacitor values would be regarded as standard by a , supplier, although he may not have them in appropriate miniature form or to the required tolerance.

The resistors include some which are not in the usual "preferred" range but stockists can obtain them from manufacturers or importers.

Note that capacitors and resistors forming part of the phase networks are specified as 51 tolerance. Observe the requirement if you can, although it may not make all that much difference to the sound if circumstances force the use of some 10 pc components.

We have shown the blend resistors as permanently wired into circuit, rather than optional or brought into circuit with a double pole switch. If you need to use the Playmaster 140 under conditions of maximum left/right frontal separation, this is available in either of the other mode switch positions.

A further point is that the Motorola data did not extend to complete input and output coupling arrangements. Our circuit and the related module provides for input resistors as well as coupling capacitors, and also output coupling capacitors and resistors. These additions obviate any clicks which might otherwise be produced when the SQ module is switched in or out of circuit.

Because we had a specific application, the module was simply installed in the Playmaster 140 and operated as part of it.

However, for the guidance of those who may wish to use the module in another context, the following information is relevant, taken from the Motorola data:

Supply volts: 20 typ; 25 max , Supply current: 16mA typ; 21 max.

Input imp: 1.8 meg min; 3 meg typ.

Output imp: 5k ohms.

Input volts: 0.5 typ; 2 max.

Distortion: 0.1pc typ.

Sig/noise ratio: 74dB typ; 70dB min.

Gain: within -1 and +ld8 L/R balance: within -1 and +1 dB

Op. temp: 0 to +75c.

It will be evident from this table that the module is being operated in the Playmaster 140 at well below the maximum rated voltage, being effectively at about 12V for the two prototypes which we made up. If you have reason to do so, the supply could be increased in the Playmaster 140 by simply reducing the value of the decoupling resistor. However, be prepared to increase also the value of the decoupling capacitor, otherwise power supply ripple could penetrate the signal chain.

In point of fact, we increased the value of the decoupling capacitor to 400uf to ensure the lowest possible ripple at this point.

The circuitry can most logically be put together on a Vero board.

For our purposes, however, we completely redesigned the pattern, providing for the additional input and output components, as mentioned earlier. There is room on the board for everything - provided you use modern miniature components. If you have to use some older more bulky components, it will be necessary to resort to such measures as standing resistors on end or mounting capacitors well above the surface of the board. Electrically this will be of no consequence and, fortunately, there is room in the chassis to accommodate a more bulky assembly.

. As with any compact wiring board, and especially one involving an IC, it is essential to use a small iron with a clean, sharp tip and apply the solder sparingly but quickly to each point. If the solder doesn't flow immediately, don't fiddle with the Iron tip on the board in the hope that brute force - or brute heat - will do the trick. Try to scrape the offending surface with something small and sharp and try again.

Be particularly careful with the connection to the IC itself, first to get it round the right way, as indicated by the indentation on one end, and secondly to avoid overheating or bridging the connections while soldering. ' If you feel at all apprehensive, it is possible to buy a suitable socket (McMurdo, 4cJ which can be soldered to the board. The IC then plugs straight in with little risk of incident or accident. We installed such a socket in our prototype, not just for this reason, but to have a test-bed into which we could plug other speciments of the IC. Incidentally, the writer is not too proud to admit having an old-fashioned reading glass handy on the bench. It makes it d lot easier, upon completion of such a unit, to inspect it for solder runs, or "whiskers'1, dry joints and so on. , For the rest, it is simply a matter of identifying the components and placing them correctly on the board. If you do this, the nodule can hardly do anything else but work! be apparent, the module Is designed to plug into a PC edge connector of pitch 0.15in and of sufficient length to accommodate N the 6nm or 2-3/8-in width of the board. The connector lugs must be sufficiently stubby to clear the bottom of the chassis when the module has clearance height at the top.

As will We used Vero board connectors which are available In packs and can be bought quite easily. Care must be taken to allow enough clearance at the top of the module.

With the module on hand, the connections can be identified readily.

Oust make sure that the board mates correctly with the contacts and if the connector is longer than necessary, take some step to ensure that it will always be plugged in correctly. The connections to be made to the module 'via the edge connector are shown clearly.

Note what we said earlier about the 220uF decoupling capacitor.

That just about completes the electrical description of the Playmaster 140. He may have occasion later to suggest additions or extensions, but, for the time being, we propose to sit back for a while and enjoy the music.' 8ut perhaps we should say a few words about the cabinet work: The up

turned dish chassis lends itself to the fitting of a simple metal hood finished in any way the supplier believes to be appropriate. How simple or ornate, or how cheap or costly such metalwork will be governed largely by supply and demand.

Although we have not had occasion to investigate the matter, a metal cover with not-too-open ventilation would have a possible advantage in areas close to powerful radio and television transmitters. Curiously the problem of RF penetration is less likely to occur in the input stages, than directly into the power modules, with their relatively high gain and extremely wide inherent frequency response. When it does occur, such interference is independent of the volume control setting.

But, leaving aside RF penetration, many constructors may prefer to house the amplifier in a wooden case, finished to match other equipment or other furnishings. The case can likewise be a simple item, although not everyone finds it easy to achieve a really pleasing finish.

If you do plan to build your own cabinet, we suggest that you regard our diagram as provisional only and wait until your own chassis is to hand. At this point, the metal work can be measured and the cabinet made to fit snugly - not 2nm too small or with a 2” gap all around! Unless you have more facilities than average, and a great deal more skill, you won't attempt 45 degree mitre corners; butt joints are manageable.

--------------

If using natural timber or veneered particle board, it will be necessary to make some provision to hide the end grain at the corners of the cabinet.

One method is to leave the top very slightly short and to glue strips of veneer against the end grain, flush with the side surfaces. Another method is to make the end cheeks deliberately oversize, butting the top and possibly the bottom against them. The exposed edges of the cheeks can then be dressed with veneer, while the front edges can also be veneered as necessary.

If the cabinet is built of natural particle board, the surface can be covered with adhesive wood-grain fabric or, again, with grained laminate or veneer.

You may have your own preferred method of wood finishing but one we have used ourselves is ready-made for the not-too-skilled handyman:

Thoroughly sand the surface, being careful not to introduce linseed putty or a high lacquer content filler, which tends to soak into the adjacent fibers, making them react differently to the rest of the surface.

Now go over the whole surface with one even coat 'of thinned clear lacquer, allowing it to dry thoroughly. This tends to soak in and partially seal-without producing a surface gloss.

If you want a natural timber finish, it may be necessary only now to apply a coat of "teak" oil, rubbing it in with ordinary steel wool (not the soap pads, or course). The result will be an oiled, matte finish,, probably about honey color.

If you want a darker finish, apply an appropriate oil stain before the finishing oil, rubbing it in with steel wool. In fact (let's just whisper this) some professionals use selected shades of shoe polish as a handy and easily controlled form of stain! Finish with clear oil, wipe over with a soft rag and you will end up with a pleasing matte finish, not easily marked but easily touched - up if it is.

PARTS YOU NEED FOR:

MAIN SECTION

1 chassis, nominally 38 x 28 x 9cm or 15 x 11 x 3.5 ins.

1 Escutcheon plate.

1 Metal cover or wooden case to suit.

1 Mounting bracket for slide potentiometers.

4 rubber feet.

2 3-pin DIN sockets, with plugs as required.

3 5-pin DIN sockets (180 degree) with plugs as required.

2 tagstrips, 3-lug and 4-lug.

1 Octal socket (optional).

4 Polarised loudspeaker sockets and plugs. ' 4 Fuse holders.

4 1.5A fuses.

1 Power cord, with grommet, clamp, 3-way terminal block.

1 2-bank, 4 pole, 5 position rotary switch.

1 2-bank, 6 pole, 3-position rotary switch.

2 2-gang 50k log potentiometers (volume controls) with mounting screws, 4.5ou travel..

2 2-gang 500k linear potentiometers (tone controls) with mounting screws.

1 2-meg single linear potentiometer with mounting screws.

5 Push-on knobs for slide potentiometers.

2 skirted knobs for Select, Mode switches.

1 Mains off-on switch (if required).

1 6V 5ftnA indicator light assembly.

2 Stereo headphone sockets with

Isolated stereo speaker switching (see text).

2 pieces of particle board and composition board to support power modules.

(See text and diagram.)

1 Tagboard, 9 lugs per side.

2 Mounting pillars, 1cm or Jin.

4 Silicon diodes (100P1V, 3A min).

7 meters (8 yds) twin shielded and braided figure-8 wire, plus hook-up wire-as necessary in a variety of colors.

1 Resistor 470 ohm 5W.

2 Electrolytic capacitors, 2500uf 35VW upright can type.

Electrolytic capacitor, 3000 or 3300uF, 35VW upright can type.

1 Power transformer 15-0-15V 2A nominal.

2 0.01uF ceramic, 100V or higher.

POWER MODULES (4)

4 Vero boards.

TRANSISTORS FOR POWER MODULE AY9171=RS276-2025. RS276-2026. 2N3740. BD202. BD236. TIP32B.

AY817I-RS276-2017. RS276-2018. 2N4232. B0201. B0235. TIP3IB. 2N5088=RS276-2013. BC548/9. BC183L/184L.

2N4250=RS2?6-2024. BC558/9. BC213L/214L.

2N3565=RS2?6-2009. BC548/9. BC183L/184L. , 2N3643=RS2?6-2009. BC548/9. BC183L/184L.

2N3638A.RS276-2021. BC558/9. BC213L/2141.

Aluminum heat skinks 16 gauge 7.6 x 4 x 2.7cm.

Resistors (JW 10X) 4 100k; 4 82k; 4 10k; 8 6.8k; 4 4.7k; 4 2.7k; 4 1k; 8 560 ohm; 8 390 ohm; 12 150 ohm; 12 27 ohm.

4 220 ohm preset tab pots.

Capacitors 4 470uF 16V vertical electrolytic.

4 110 uF 25V vertical electrolytic.

4 0.47uF 160V polyester.

4 390pF polystyrene.

4 270pF polystyrene.

SHUNT, PHONE BOARD

1 tagboard, 10 lugs per side.

2 Jin mounting spacers.

4 0.47uF polyester (small, typically 50V).

4 15-ohm JW resistors.

4 330-ohm JW resistors.

FILTER ASSEMBLY

1 lsostat 2-push-button switch assembly or 3-push-button assembly including mains switch.

4 0.015 uF polyester capacitors, 50V or higher.

4 0.0015 uF polyester capacitors 50V or higher.

4 33k resistors 0.5 or 0.25w

2 Mounting spacers, 1.5cm or 3/8 in.

MODULES

Preamp and tone control board.

Stereo-24 adaptor.

SQ Decoder.

a small Vero wiring board, with plug-in provision and a McMurdo or other chassis mounting socket. One Motorola HC1312P IC. Plus 11 0.5w resistors, mostly unusual values; 10 small polystyrene capacitors.

1 220uF 25 VW electro. 4 4.uF 5 YW electros.

Ext. 4 phase change: Built on a piece of Veroboard 6.7 x 3.7cm, this module contains 2 small signal transistors, 13 JW resistors, 4 0.47uF polyester capacitors and 1 220uF 25VW electrolytic.

PREAMP TONE CONTROL

1 Vero board.

2 RS 276-2009, BC109C, BC549C or similar transistors.

2 RS 276-2009, BC108B, BC548B or similar transistors.

4 RS 276-2009, BC108, BC548 or similar transistors.

2 Ferrite RF beads 3.5nm dia., 5nn long.

4 0.5 in or 1cm spacers with screws.

Resistors (JW or JW, preferably 5pc) 2 3.3N; 2 2.7M; 2 2.2M 2 1,5M;2560k; 2 470k; 2 330k; 6 100k; 2 82k; 2 56k; 2 47k; 4 27k; 2 22k; 2 15k; 2 10k 7 1k; 3 680 ohm

Capacitors 1 220uf 25VW vert electrolytic.

1 100 uF 25vw vert electrolytic.

4 4.7uF 12vw vert electrolytic.

2 1.5uF 20VW tantalum electrolytic.

2 0.47uF 20VW tantalum electrolytic.

2 0.27uF 100V polyester.

4 0.1uF 100Y polyester

4 .022uF 100V polyester

2 .0047uF 100V polystyrene.

2 .0015uF 100V polystyrene.

2 680pF 100V polystyrene.

4 100 pF disc ceramic.

STEREO-24 ADAPTOR

1 Vero board.

4 transistors BC108, BC548.

1 Angle bracket approx 1 x 1 x 7.5 cm.

2 0.25 in or 0.5 m long spacers with Jin dia or similar screws to suit

Resistors: (JW or JW, preferably 5pc) 4 820k; 4 330k; 282k; 247k; 210k; 4 4.7k; 4 3.3k; 1 1.2k ohm JW 2 1M preset pots (optional, see text).

Capacitors:

1 220uF 25v velectrolytic.

9 0.1uF 100 v min polyester.

1 .01uF 100V min polyester.

PHASE CHANGE MODULE

1 piece of Veroboard, 9 tracks, 37nra x 67um.

1 Mounting bracket, scrap aluminum, 2.5 in long.

2 Transistors, BC108, BC548, RS276-2009, SK3020.

BC148-168-208-318.

AM252, ZXT108, MPS6520.

Resistors ( 0.25 w or0.5 w , 5pc):

2 1.5H; 2 820k; 2 330k; 4 5.6k; 1 2.2k.

Capacitors:

1 22ouF 25VW electrolytic.

4 0.4 uF 100V polyester.

PARTS LIST SQ DECODER

1 Vero board.

1 Integrated circuit, Motorola HC1312P

1 edge connector to suit, see text.

1 IC socket, optional, see text.

Resistors (0.5 W or 0.25 W as available):

4 220k; 2 100k; 1 47k; 1 7.5k; 4 4.3k* 4 3.6k* 1 1k )W (if not already in amplifier chassis).

Capacitors (miniature types with DC voltage rating 25V or higher):

4 0.47uF tantalum electrolytic.

2 0.47uF polyester.

2 0.22uF polyester.

4 .039uF polyester

2.0068uF polyester

1 220uF or 400uF decoupling electrolytic (if not already in amplifier chassis).

FOOTNOTE: * asterisk indicates 5pc tolerance desirable.

ALTERNATIVE TRANSISTORS FOR THE PAYMASTER 140

THE CHANGED MANUFACTURING AND MARKETING SITUATION, WHICH FLOWED FROM THE LOWERING OF TARIFFS, HAS AFFECTED THE SUPPLY OF TRANSISTORS USED IN THE PLAYMASTER 140 AMPLIFIER POWER MODULES. WE PRESENT HERE DETAILS OF ALTERNATIVE TYPES WHICH CAN BE USED IN PLACE OF THOSE ORIGINALLY SPECIFIED.

Of particular interest were the two output transistors, types 2H4232 and 2H3740. Employing what was referred to as bimesar dual-epitaxial construction, the two transistors were notable for their ruggedness and for their ability to withstand short-term overloads. This made it practical to operated them from a non-regulated power supply, protected only by fuses against an output short circuit.

The same transistors were used again in the Playmaster 136 stereo amplifier, and in the quadraphonic Playmaster 140.

In the meantime, other manufacturers have come up with power transistors which are similarly rugged and competitive in price. And thus we have the continuity ensured of what has proved to be a very economical and very simply power module.

It simply means that the modules can now be built without electrical modification from a range of transistors, according to what is available and suitably priced.

Host of the alternative transistors which warrant consideration, come in an SOT-32 plastic encapsulation. They can be regarded as a direct electrical replacement for the earlier TO-66 style transistors, but are mechanically quite different. Fortunately, it has proved possible to retain the same circuit board and heatsink, varying only the method of assembly. ' It is still possible to obtain the small-signal transistors, but we have specified possible replacements for these types as well. The complete set of possible replacements is listed in the accompanying table. Where there are two or more types of the same brand listed, the first listing is the preferred choice, although the remaining types will work satisfactorily.

It is possible to mix brands without any loss in performance.

The reader is referred to the accompanying sketch for full details of the method used to mount the plastic encapsulated transistors. Care should be taken that the following points are observed while fitting them.

The power transistors are fitted underneath the heatsink, and bolted to it through the holes provided to mount the TO-66 type transistors.

They must be mounted with the metal part next to the heatsink, as this forms the collector connection. As a result of this, the collector lead provided can be carefully snipped off close to the body of the transistor. (The collector lead is the center one).

To ensure that the base and emitter leads are not transposed, it is vital that the mounting hole furthest from the original center clearance holes be used in each case to mount the plastic transistors. On the heatsink, these happen to be the holes nearest to the dimples for the driver transistors. The transistors are secured to the heatsink using suitable nuts and bolts. A small washer should be used under the head of the bolt, as shown in the diagram, and silicon grease applied to ensure good thermal contact between the transistors and heatsink.

Do not use excessive force when tightening the nuts, as this may damage the transistors. Take particular care that the actual transistors are not interchanged. The heads of the bolts must be on the underside of the heatsink to ensure that the transistor leads are long enough to reach through to the copper pattern on the board.

Once the transistors have been fixed to the heatsink, their base and emitter leads may be bent as shown in the diagram. Use the holes in the heatsink as a guide, and hold the leads next to the body with a pair of small long nosed pliers to prevent them from breaking off.

The heatsink is held in position, and the collector connection made through the remaining holes. We used two nuts as spacers by bolts to position the heatsink a suitable distance from the board. Do not forget to tin the copper pattern underneath the nuts to provide good electrical contact.

No difficulties should be experienced with the mounting of the small- signal transistors, as all the recommended types have TO-92 cases. These TO-92 versions mount partly through the heatsink; drill directly above where the transistors will mount with a suitable sized drill so that they are a snug fit. Do not solder them to the board until the heatsink has been fixed in position, as otherwise their height may be wrong. The final step is to thermally bond their cases to the heatsink with a generous blob of silicon grease.

NOTES

Index | Prev. | Next