AudioLinear

Copyright

Any use of material published directly in this site is subject to the laws of copyright. No right to copy or use is granted or implied by the publication on this site or elsewhere.

Any linked information may be the subject to restrictions by the owners of the copyright of the information therein.

What we offer

We have developed a module that operates as a direct replacement for the amplifier modules used in the Quad 405 Power amp. These modules are made to the highest standards of design and manufacturing and take advantage of the latest developments from semiconductor manufacturers to provide the best performance available.

Details:

The module uses no less than six Motorola MJL3281A transistors in the output stage and is conservatively rated to provide more than adequate current drive into the lowest impedance speakers you are likely to come across. All parts have been carefully selected to give the lowest distortion and noise levels, including a Burr-Brown OPA134PA voltage amplifier stage, metal film and wire wound resistors, polystyrene, and non polarised capacitors in all signal processing stages, and high voltage capable devices everywhere. The board is designed to fit in the standard 405 space and includes a new heat sink bracket arranged to take the six output devices and drivers. The heat sink is fully insulated electrically and an extra lead is supplied if requested to provide the chassis ground connection.

Owners of Quad 405 and 405-2 amplifiers are able to use this as a drop in upgrade to their existing systems and enjoy both renewed reliability, superior performance, and greater satisfaction with their equipment in the knowledge that the amplifier has a new lease of life and more than matches the latest power amplifiers available today whilst retaining their investment in this classic piece of British engineering. The module could also be used to build a multi channel system for home theatre, and information or supply of suitable power supplies is available

For more information please send E-mail as above. To purchase please send an order to AudioLinear, 35 Monument Chase, Whitehill, Bordon, Hampshire, England, GU35 9QS. At present we must ask for cheque with order, but hope soon to be able to accept credit cards. We make to order, but hold stock of some bare PCB's and other long lead time parts. Assuming we have bare boards, our lead time from receipt of order is generally less than ten days.

Various angles of the 6100-1 amplifier modules showing connection points, mounting arrangement compatible with Quad 405 module, and 6 transistors for output stage.

6100-1 Amplifier Module £180.00 each

Last revised: February 20, 2003

More information

The amplifier module is the same shape and size as the original 405/405-2 board, but mounts the output transistors and drivers on ‘L’ shaped heat sink brackets fitted to both sides of the board. This way we are able to fit 3xTO264 and 1xTO220 device side by side mounted to both sides of the PCB.  Each output transistor has its’ own 0.1 ohm wire wound emitter resistor to ensure current load sharing between the three devices. The drivers are Motorola MJE15033 devices selected for their high voltage capability, linearity, and gain, as well as having the right current, power, and package. The frequency response of these devices matches well with the Motorola MJL3281A output devices, and the rest of the circuit, although in this arrangement the lower driver is damped by a choke to ensure the overall response is matched.

The other modifications to the original are aimed at increasing the current limit and reducing noise and distortion – real or imagined. We have used 1% metal film resistors and wire wound to avoid question as to whether there is an audible contribution from carbon devices. We have also used the best available OP amp for this purpose, i.e. the Burr Brown OPA134PA, again to remove any blame for distortion from that area, but also to reduce the real measurable component contributed by the LM071 device of the original. The OP amp power supply is additionally regulated by transistors. In the area of capacitors, electrolytic types are used where the purpose is purely supply decoupling, combined with low value ceramics to improve RF rejection and prevent oscillation. In the signal path polystyrene and polypropylene devices are used where their attributes could make a difference.

To make the replacement job as simple as possible the board has been designed to fit within the 405 case, and the heat sink bracket is designed to mount using the same holes as the original. Connector tags are used to allow existing wiring to be used although the position of tags is slightly different (the tags are lined up on the board edge at the end opposite the heat sink). Because the heat sink is insulated electrically from the circuit an additional tag is used to provide chassis ground connection, and a suitable wire can be supplied complete with a screw tag that allows it to be connected either to the heat sink (spare hole for this) or to the power supply reservoir capacitor tags.

At present we have tested the modules extensively, mainly with real music input to real loudspeakers. Our favourite speakers for testing have been Celestion A3’s, which have a 200W rating and present a 4-ohm load. In A/B tests at normal listening levels most of our listeners have preferred the upgraded device. At high output the preference has been unanimous in favour of our upgraded board. Given the aim of renewing the life of the 405, and improving drive to modern low impedance speakers we are pleased with the results so far.

Installation

Before performing any work on the amplifier ensure that it is completely disconnected from the mains power source and has been left with the speakers connected for at least 10 minutes in order to discharge the power supply capacitors. Mains power can be lethal. Under no circumstances should the Amplifier be operated with the lid removed. The internal power supply contains high value capacitors that can hold charge for a significant time after power is removed from the Amplifier. To ensure the capacitors are discharged leave the speaker connected and the input signal feeding the Amplifier for a few minutes to discharge the capacitors after removing mains power.

Before you rip open your 405…

Please, first read all the way through in case we have written anything near the end that you later wish we had put at the beginning.

Opening the Quad 405

The Quad 405 has two screws holding the lid on. These are to be found on the rear underside of lid, and are plain cross head screws. After removing both screws the lid will slide back and then lift off.

Removing an Amplifier Module

Each module is held in place with four cross head M4 screws, together with shake-proof and flat washers. You will need a long bladed screwdriver for these, and they may be rather tight, so take care not to damage the heads! After removing these screws the modules will lift out, but may be a bit sticky due to the use of heat transfer compound.

Disconnecting the Amplifier Module

!    Before disconnecting the wires from the amplifier modules be sure to write down the colours of each wire and which terminal it was connected to.

You can disconnect the leads from the right hand module with the module still in place, but you will need to lift out the left hand module to get at the wiring and disconnect the leads. Left and right refer to the front view of the Amplifier, i.e. when looking at the heat sink fins.

Free up the cable loom wires

On the replacement modules the terminals are positioned at the end of the board. This differs from the original which has terminals at the side, and it is usually necessary to loosen the cord on the cable loom to allow the wires to reach the new position. The Quad loom is very neat and tidy, but there is no sonic benefit to tying together the wires.

Remove screws from module land on heat sink

Depending on model, after removing the old modules you will find two screws in the area under the heat sink bracket. You will need to remove these screws before fitting the replacement. If you have these screws then be careful to not move the Amplifier while the screws are removed as they hold the heat sink to the chassis, and with it the threaded slides into which the amplifier is screwed. If they are moved you can still get the screws in, but it is more fiddly.

Fitting the new Module

You will find it easier to install the new modules if you place the Amplifier heat-sink down on a stable flat surface. The anodised finish is durable and unlikely to be damaged but if you are concerned then place it on something soft.

Hold the new module vertically with the heat sink bracket down, and then place the screws and washers from the old module into the four small holes. Lower the board into the case and line up the screws with the holes.

Turn each screw a few turns to start the thread so that they don’t drop out and fall inside, then tighten each in turn.

Connecting the new Module

AudioLinear 6100 Modules have 6 x 2.8mm Faston terminals along the end edge of the board. With the modules fitted to the 405 chassis it should be easy to get at these and connect the leads. Each terminal is marked on the board with its’ function as follows:

+v        Positive DC supply line (nominal +50V)

In        Input

Gnd     Ground side of input lead, from braid or shield in coax lead

0v        Power supply zero volt line

Out     Speaker output line

-v         Negative DC supply line (nominal -50V)

Connecting the chassis ground lead

In the original 405 chassis ground to the module is via the heat sink bracket. AudioLinear Modules have a terminal for positive connection of chassis ground, and this may be connected via the supplied lead to the heat sink bracket using the spare hole occupied by a screw for this purpose, or may be soldered directly to the earth tag on the mains input connector.

Replace the lid

Although you may be tempted to try it out now please put the lid on first. Mains power is directly connected to the top of the transformer and easily touched with the lid off.

Powering up for the first time

Finishing off and optional items

Depending on the age of your 405, and whether or not the Power Supply capacitors have ever been replaced, you may benefit by replacing them. If you do this you can also upgrade them from the original 10,000uF items to 22,000uF parts. Thanks to advances in manufacturing these are now available in the same size cans and are a drop in upgrade.

If you replace the DIN input connector with RCA/Phono sockets it becomes a lot easier to mate up the 405 with commercial leads. Quad thoughtfully put two holes in the unit that can be used for this, but make sure you use the isolated type because otherwise you will create a short circuit between chassis ground and the input stage ground of the amplifier module. This will not create a safety earth problem but may cause earth loop problems.

If you want to go to town you could rewire the whole thing with higher grade wire internally. You are unlikely (in our view) to hear any difference, but you would feel good about having done it. About the only place that is likely to make any improvement is to use good thick low resistance wire from the power supply capacitors to the amplifier modules, but don’t spend too much on it…

Personal area

Contents

Hot List

Back to Top

Contact Information

E-mail alan@asawyer.demon.co.uk
Web address http://www.asawyer.demon.co.uk
Telephone +44-1420-472100

Back to Top

Comments and Suggestions

Please tell me what you think about my home page and how I might improve it.

Back to Top

Biographical Information

Little Waltham Church of England Primary School, 1962-1969

Sir Anthony Browne's School, Brentwood, 1969-1976

Middlesex University, 1976-1978

IMCA, Diploma in Quality Management, December 1996

IMCA, MBA Completed and Graduated at Ceremony in Edinburgh on 24th August 2002 Papers

Back to Top

Personal Interests

Computers

My career proper started with a job as an office manager for a food wholesaler in London around 1976. I stayed there too long but eventually moved on and took a job as a programmer for the company that serviced our 17bit (yes really 17bit) mini computer. The old beast was programmed in machine code - not assembler - by entering programmes as octal numbers into a key pad !

We developed ways of writing the programmes on early 8088 PC's using text editors and a simple assembler that could manage forward references only, but did at least save us from both remembering the opcodes in octal, and lessened the problem of having to guess the addresses of operands in advance.

After a while we came across a guy called, I think, Dave Green, who had developed a wonderful Meta disassembler / assembler package that was able to assemble for just about anything going and he added our system to its tables. From then on we really were moving !

After a lot of Unix work things took a move towards Microsoft. This seems to be really due to the popularity of Word, WordPerfect, Quattro Pro, and Excel which of course needed Windows to run. Windows for Workgroups 3.11 was relatively light on resources compared to running Unix and much cheaper on hardware. This meant that soon almost everyone has a PC for typing memos and the like. Such was the popularity that Microsoft went and developed Windows 95, which although much prettier and easier to set up, was just as unreliable. The result of this was to set back mainstream adoption of a proper operating system for 3-5 years while Microsoft continued to play catch up, with NT paradoxically the better O/S being perceived by the vast majority of users - even some clever ones - as better. Now that we have Windows 2000, which is solid as a rock, users are finally starting to come on board. This in some cases is because the same users that used to think Win98 was great and good have now started to do more demanding tasks - like receive E-Mail whilst browsing a Web site (!) and have found out just how plain dumb Win9X (nee Win3.xx) really is at multi-anything. Anyway the long and the short of it is that I took to Windows NT3.1, 3.5, 3.51, 4, 2K and now XP with a vengeance and saw just what could be done on the X86. I have no doubt that the architecture of WinNT and its lineage is more than adequate in all the areas that matter to the masses and to almost all heavy duty users. It certainly does the job. Above all, most of the time, it is easy to set up, administer, and maintain. There are notable exceptions like setting permissions in IIS - an area that seems to need a degree in muddled thinking to get it right.

Back to Top

A potted history of my computing interests

This section describes my current home PC and evolution of parts and software that it has seen.

The story may be of interest to anyone who is contemplating building or upgrading their own PC and wants reassurance that the parts they are considering will run successfully with the operating system and software they want to use.

The start

During 1981-2 I picked up a computer magazine that has long since disappeared that described a home build 6809 computer system. It was really advanced (!) and had a 1MHz CPU with 2KB ROM and 1K of RAM. Having taught myself to program the computer we had at work (17 bit mini programmed through a classic binary switch and lights front panel) I had gained an interest in understanding how the things worked and what could be done with them.

I duly bought the PCB and parts for the design and got it up and running using a surplus VDU terminal. This thing could only be programmed using hex input so the reference card and home brewed coding sheets were pretty much a necessity. The first programmes I wrote for it were a noughts and crosses game, followed by a dis-assembler. A cassette tape interface followed as it was too much of a pain having to retype thousands of hex digits in each time I lost power from the box. With "mass storage" available I now branched out and wrote a simple assembler for it, along with some halfway decent drivers for the serial ports. These allowed interrupt driven I/O and gave me a clock display and status line on the terminal, but more important they were the foundation of a set of I/O drivers for floppy disc and printer, all interrupt driven, that allowed for a multitasking executive to operate. Many late nights and wasted brain cells later I decided that this was never going to be the basis for commercial exploitation and so started looking for a C compiler to learn on.

Around 1982 I read "The C Programming Language", and "The UNIX Programming Environment". It seemed to me at the time that there was something elegant and almost naturally logical about the Unix Operating Systems "everything is a file paradigm". C as a language appealed because to me it read almost like an abstract pseudo code version of the assembly language that I was so familiar with. It also allowed the same control of the machinery but with a short and simple way of expressing things that took a lot of trouble in assembler and needed more attention to unnecessary detail to prevent human error. between them these two books probably shaped the period of my life from 1986 to 1996. Spurred on I had formed the view that C was the way to go, and Unix was the operating environment for developing and hosting programmes on computers.

At this time the PC was in its' infancy, meaning even a PC was too expensive as a home machine, let alone a PDP11 or some other mini computer on which Unix would run. C compilers were also a little thin on the ground. The best deal around seemed to be the Amiga 500 for which you could pick up Lattice C for less than £100. I duly purchased an Amiga and the software, but found out that in software, as in so many thing size does count. In this case the size of the storage, both disc, and RAM. The machine was basically useless for serious C compiling - in fact it was useless for any C compiling. To write a program involved swapping floppy discs to load the OS, then the editor, then several times over for the passes of the complier. despite upgrading the following week to an Amiga 1000 which had more ram it was still virtually unusable, so I decided that a PC was the only way to go.

Just about then Amstrad came along with their first really usable, and cheap PC, the 1640. This had an 8086, a 5 1/4 inch floppy drive, 640K RAM, and best of all a 20 MB hard disc. This was ideal for both assembly language and C development with the also very cheap by the standards of the time Borland C compiler. Over the next year I developed a suite of programmes for the PC that effectively did the same job as the package we used to sell customers where I worked. This was a complete business suite for distribution businesses covering sales, purchasing, accounts, stock control, etc. The only real problem was that the PC used DOS and was only really single user.

So moving on and up I upgraded to new Amstrad PC2386. This was another Alan Sugar bargain of the day at something like £2,300 for a VGA 14" colour monitor sat on top of a 386DX 20MHz CPU unit containing 4MB of RAM and a 20MB hard disc. Now, on this machine all the code I had written really flew, and best of all I could now use Concurrent DOS 386, that allowed for multi user working with all that CPU power :-). Of course what I really wanted was to run Unix on it....

So the next logical step was to buy a copy of SCO XENIX 386 and the C development system. Duly acquired a very large box arrived complete with a shelf full of manuals. At last I could get down to some really serious work and develop the system into something that might one day be a commercial wonder. OK, well maybe not, but anyway after about a month of fighting the machine to get it to run XENIX (there were some patches necessary to the kernel - something to do with a tiny but critical difference in the way the Amstrad keyboard controller reported its' self test result) I finally had my suite of programmes ported to - nearly - Unix. Some small code changes, mostly concerned with the change to a full flat 32 bit model on XENIX compared to the Borland C / DOS 16:16 segmented model and I had a system that would run 18 users on the 386 at a comparable performance to the Mini we used at work. 19 users ?, yes this was from the console, the serial port on the PC, and a 16 port card from Chase Research.

Then what ?

Then I ran out of time and had to get back to some real work... No what I mean is I am working on my MBA dissertation, and need to finish it before I get back to this story. No excuse now. I've submitted the papers so I can find some time for this I hope. In between however I want to work out how I can put all this into SQL Server and have it published automatically from a set of forms. The idea (at present) is to use Access Forms as the front end (at the publishing end that is), then this uses a SQL back end to hold the data which is published to html pages and ftp'd up to demon each night. This is really just a learning exercise so I can do it better at work.

1989 SDL closed

1990 Lynwood, as Unix Development Manager

1993 Almost left for an offer at DHL, but stayed to take up the challenge of Quality Management

1996 Earned my Diploma in Quality Management

1997 Almost left for an offer elsewhere -

1999 Started on MBA

2002 Hopefully I'll get it finished real soon now.

Approximate dates and upgrades to PC system

1986 Amiga 500, Amiga 1000, Amstrad 1640HD20

1988 Amstrad PC2386, 4MB RAM, VGA, 14" monitor 640x480x16 colour

1989 16MB RAM,  AHA1540, Seagate 500MB SCSI Disc, 16 port serial card

1990 486-25, 16MB, VGA, AHA1540

1992 486DX2-66, AMI P9000 based VLB Video, IIyama Vision Master Pro 17, 1280x1024x16 colour @ 70Hz

1993 Pentium 90, 32MB, Matrox Millennium 4MB PCI Video, 1280x1024x65k @ 72Hz

1994 Pentium 133, AHA2940UW SCSI card

1995 Dual P133, Tyan motherboard with 8 SIMM slots, 128MB RAM

1996 P200, AWE64

1997 PII233, Supermicro P6SBA AGP Motherboard, 3x128MB DIMM, 384MB, Quantum 18GB 10k SCSI disc, SB 128

1999 PII500, Matrox G400 Max AGP video (thought it was AGP 4X), Rainbow Rubber G series for TV and analogue video capture, SB Live Platinum 5.1

2000 PIII700, Tyan VIA based Tiger 133, 3x512 PC 133 ECC DIMM, 1536MB, 60GB IDE, G400 Max found to only be 2XAGP, Matrox RT2500 "Real Time" video editing card , Adaptec 29160 SCSI,

2001 PIII1G,

And some guesses...

2002 P4-2.6, 4x512MB DDR DIMMs, 180GB SCSI Disc, 21" IIyama Monitor, Matrox G800 video card,

2003 P4-3.5GHz

2004 P5-4G CPU (hey wait a minute, wasn't the Pentium so named because it was a "P5" - perhaps they'll call it a P². No doubt this will need yet another Motherboard format), 4GB RAM 1TB Disc drive, 4 headed video card with 3D automatically generated from 2D images.

2010 P10-10GHzSMP CPU with equivalent of 16 concurrent processors and a very long word instruction set executing over 100 billion instructions per second. RAM is approaching 1 Terra byte for most users, and discs have lost the commercial battle with silicon and battery technology. RAM is self refreshed and self powered, needing an occasional watt or two of power. It gets this courtesy of the operating system turning on the power supply after a few hours if the user doesn't. Voice input and output meaning little need for a keyboard. Also uses lip-reading camera input for a quiet office. No one has yet managed to convince users of the safety of direct to brain connections. Output is via 3D images and usually a wireless earpiece that has an override for when the wife is calling. One good thing is you can sit in your lounge and have it whisper in your ear when a message arrives and you can hear it spoken to you without getting up. We have long since lost the limitation of BT connection costs and speeds. All the memory and a low power processor sit with a battery in a combined remote control / PDA / Mobile phone / In Car Entertainment unit that fits in your pocket, and you just drop it into a slot at the side of your display unit at home or office to add the large screen and high power processor for faster working. The Terra Byte per second data links mean that you can back up your data in "two ticks" and still carry it all around with you. Video and audio at SACD / DVD quality is all built in and you can wirelessly send you holiday videos to the HEC (Home Entertainment Centre) when you visit friends. Just take it out of your pocket, press "play" and away you go. All this in a unit the size of a Journada. Of course the displays are somewhat larger thanks to LCD wallpaper. Buy it at you local DIY store, stick it on the wall, and attach a controller to however many strips you want, depending on width of picture.

Back to Top

Interests

Computing

See elsewhere for the development of my interest in computing.

WEB surfing

I never cease to be amazed by what I can find out there.

Swimming and Cycling

My only form of exercise, well O.K. I do walk to the coffee machine, pub, car ...

Holidays

Frankfurt, Germany 1998

Royan, France 1999

Vias, France 2000

Nowhere, England, 2001

Pubs

Good real ales that have been well kept, this stems from college days and low cost beer I'm sorry to admit, but it is fun.

Restaurants

We all have to eat, but I especially like Indian, Chinese, French, and English food

Books

Albeit slowly I am building up a small library of all sorts of, mainly non-fiction books.

Back to Top

Our children

Christopher

Christopher's WEB site

Lucinda

Joseph

Alex

Mr Waine - Joseph's and Alex's  Teacher 2001-2002

Miss Brewer - The children's Head Teacher

Back to Top

THE END (for now...)

Last revised: February 20, 2003