Circuit
modifications and upgrades
There are quite a few circuit modifications floating
around for the 405. As usual, I stay away from anything
which drastically alters the circuit topology of the
amplifier (I try not to second-guess the designers
unless there is a very good reason to) and instead
concentrate on restoring the amplifier and enhancing its
performance while keeping it as near to the original
design as possible.
Back to contents
Op Amps
I normally change the original op-amps (depending on
year, Quad fitted either LM301 or TL072; the 072 is
still fairly competent even by modern standards, the 301
is archaic but both can be improved by quite a large
margin.
Back to contents
Decoupling &
bypassing
This is something which I improve also. I learned a lot
in my early days by studying vintage Japanese designs
such as the ones used by Pioneer - those guys threw
capacitors at everything, just to ensure everything kept
functioning as expected in the darkest of worst case
scenarios they could imagine. Quad did the same with the
606 family onwards, and I just translate some of these
same methods to the 405 (the 405 still shares a lot in
common with Quad's later amplifiers). Essentially this
means adding extra capacitance to help stabilize power
supply lines in any area where wiring impedances might
cause voltage fluctuations; within reason, adding extra
capacitance like this is always a good thing and can
only help improve performance.
All larger capacitors are fitted with film bypasses;
this ensures that they exhibit a low impedance at high
frequencies which has a positive effect on the overall
high frequency response and performance of the
amplifier. In addition I also fit snubbers in places
such as across the transformer secondaries (again, these
were recommended by Quad in their later designs).
Back to contents
The power supply
The standard Quad 405 uses a pair of 10,000uF
capacitors. Compare this to a similar vintage Pioneer
with single power supply would have at least 12,000uF)
and you can probably see that these are undersized for
what is essentially a 200W mono amplifier as far as the
power supply is concerned.
Taking a more mathematical approach, the voltage across
each rail is 50V @ idle. This 50V into the standard
10,000uF capacitor gives 12.5 Joules of energy, or 25
Joules total (across both rails). Voltage swing at
the 405's specified 100W RMS output will be 28.3V RMS,
which means that we will need more than the maximum of 39.9V peak
output voltage available on each rail for an undistorted waveform. Add this
to any voltage overheads caused by emitter resistance,
transistor voltage drop and transformer sag at peak
output levels, and we can see that we can't afford any
more than 5V ripple (and ideally much less) caused by
the capacitors.
If we say that by a not unreasonable approximation that
to keep within this 5% of ripple that we require 1 Joule
per 10 watts of peak output power then for the
original 10,000uF capacitors we have enough energy for
250W/pk, which equates to 88.4W RMS /ch.
Linear power supplies for audio signals are very
difficult to quantify so this may or may not be just
about enough in practice depending on the signal
present, but in any case if we can add some extra
capacitance to this it would be a good idea.
Say that we instead opt to fit 12,000uF (the next
available size up), we now have a total of 30 Joules
available and using the same reasoning as above this
gives enough for 106W/ch on each channel, which rises to
132.6W/ch if we use 15,000uF. In my experience, going
beyond this point will lead to quickly diminishing
returns and 15,000uF is as far as I would recommend
going.
 |
|
Set of 6,800uF capacitors
in a 405 |
I normally fit 15,000uF, or else fit a smaller pair of
6,800uF in parallel; fitting the smaller parts gives a
total of 13,600uF (enough for 125W RMS/ch by the same
reasoning if we add some extra decoupling), still more
than enough for the 405 and also has the advantage that
smaller components tend to exhibit slightly better high
frequency performance and when combined will also offer
slightly better ESR and current performance over a
single, larger component. Once we get to this point,
improvements in instantaneous current delivery and
impedance will yield far more noticeable improvements
than just adding more capacitance will.
As far as component choice goes, the voltage across each
rail is around 51v which means 63v components will do
fine (the originals are 63v); if space is not an issue
then choosing 100v components usually gives a small but
further improvement still (slightly lower ESR and higher
ripple current), but can be a struggle to fit as higher
voltage components are often taller.
One more thing to note is that to ensure continued
reliability, the original rectifier bridge should be
similarly upgraded also. This part dates from the 303
era and is on the edge of its capability in the standard
405 and I wouldn't trust it with an upgraded amplifier.
Back to contents
Component choice
I often get asked about the components I use in my
restorations; my answer is simple, that I use exactly
the same components that I would specify if I had to
design a high performance, high bandwidth and low noise
power system for any other use.
The components which I use are sourced from established
premium brands which I have used for a long time and
trust; I do not use so called 'Boutique' brands
of 'audio grade' component as I have found over time
that these are often not of satisfactory quality or
performance; in many respects, any supposed advantages
of 'Audio Grade' means pretty much nothing as far as
electronic performance goes. Using the example of a
reservoir capacitor in an amplifier power supply, the
properties which this component needs are low impedance
across the frequency range and high current capability
(i.e. it can store and supply high peaks of energy in a
short space of time). A high grade power supply
component will fulfil this purpose with ease and will
also last a lot longer than the average niche brand.
As far as component types go, I always use low ESR, high
current capable BHC/Kemet for large power supply
capacitors, Panasonic for smaller electrolytics, Vishay
or Wima for small capacitors and Bourns, Vishay or
Wellwyn for components such as resistors. My view as an
experienced engineer is that these components give the
best performance of any component, regardless of
price, and the many tests I've performed and also the
virtually zero number of component failures I have ever
experienced with these brands back this up.
Back to contents
Connectivity
 |
|
RCA sockets being
installed in a 405 |
Unless you own a 405-2, your amplifier will only have a
Quad standard 4 pin DIN audio input and tiny pushbutton
loudspeaker terminals which won't accept any more than
18AWG cable.
I always fit a set of binding post type loudspeaker
sockets (the spacing is a bit close but with a bit of
modification they can be made to work well), I also fit
a pair of RCA phono sockets in parallel with the
original DIN input socket, so that either input type can
be used.
The RCAs are connected to the DIN using twisted pairs of
cable, but all other line level audio connections are
run with screened microphone cable.
Back to contents
Wiring looms
When I test a 405 on the bench, noise performance is
always slightly worse on the left channel than on the
right; the reason for this is simple, that on the
original wiring loom the audio wiring for the left
channel is bound into the same wiring bundle as the
primary mains supply. This noise is picked up mostly
through the signal input wiring but to a small extent
the loudspeaker output wiring as well. We're not talking
huge amounts of noise here (in the worst case I don't usually see any more
than 10mV pk-pk, which isn't the worst you'll ever hear), but if
we're trying to reduce the noise floor we need all the
help we can get.
The simple solution to this problem is to replace part
of the original wiring loom; on the 303 I tend to stay
away from doing this as it's virtually impossible to get
a neat result that I feel happy with (i.e. separating
audio and mains etc.), but the layout of the 405 is
simpler and it's much easier to separate the delicate
wiring runs and do it neatly.
Another reason to change the looms is to fit thicker
wiring which is more suited to the task of carrying
several amps of peak currents; the original feeble
wiring in addition to the relatively long wiring runs
places a bottle neck on the ultimate high power
performance of the amplifier. This equates to an added
impedance in series with the loudspeakers which will
worsen damping factor and overall negatively affect
power delivery and dynamics throughout the frequency
range.
 |
|
Original 405 wiring
layout; elegant but not optimal |
One thing to mention here; if you look at the original
wiring routing on the 405, wiring lengths for both
channels end up being pretty much the same length. My thinking is that this was
probably by design as well as convenience, that by
making both channels have equal wiring impedance and
inductance, it helped match the response and
characteristics of both channels (but at the expense of
noise performance), especially when using the original
thin wiring.
Assuming that this was a consideration, if we make a new
loom using much heavier gauge cable then we can run it more
'point to point' without adding substantially more
capacitance, inductance or resistance to either channel.
Making a new loom: I keep the mains primary
portion of the original loom (the currents here are
quite low and the original wiring is routed well) and
instead replace everything after the transformer
secondaries. An added advantage to installing a new loom
is that there is not a lot of spare wire on the original if you want
to fit a new rectifier bridge (as I always do).
 |
|
New loom in a restored 405 |
I also run the supply lines and audio wiring through
anchors attached to the rear of the mains transformer,
set apart in two rows - this ensures that mains and
audio on the left channel only cross in one place, at
90º to each other. This also ensures that the upgraded
amplifier retains a near-factory finish (a professional
finish is important to me) but also that
noise levels are kept as low as possible.
The power supply wires are made from heavy current
capable wire, these being kept as short as
practically possible. I usually make the signal wiring
runs from audio grade cable; I realise as an engineer
that there's very little advantage using audio grade
wire in a place like this, but what's the problem when
it costs so little? It looks nice and it's guaranteed to
work as well as, if not better than normal wire.
The results are always good on the bench; noise floor
for both channels is bought down by at least 10dB and
noise on both channels is about as equal as you can hope
for. A small amount of thermal noise remains, but on the
whole the amplifier is now very quiet. Impedances in the
loudspeaker and power supply circuits is now much lower
(in the region of 20-30% in places) and this in
conjunction with new power supply capacitors will make
quite a noticeable improvement to audio quality,
especially in lower bass dynamics.
Back to contents
Cosmetic restoration
Painted parts:
As with other Quads of the era, the 405 was painted what
I term 'Quad Beige Gold'. Strangely, this is the same
colour as used on parts of the 303 yet seems to
deteriorate more than it does on the older model (it
fades and rusts more readily). As with the 303, the
casework is a mix of steel (the top panel, main chassis
and base) and aluminium (end cheeks and heatsink); the
chassis is powder coated and quite rugged, as are the
anodised black heatsinks on the front panel, however
even on an otherwise mint amplifier the metallic gold
case will have undoubtedly faded over time. If you're
going to attempt something like this, it really helps if
you have a well equipped workshop and some skills in
something like car restoration (luckily I do); the
techniques and materials used are similar.
 |
|
Case parts of a
Quad 33, 303 and 405, stripped to bare metal |
The only way to re-finish everything to a factory finish
is to remove the case parts
and strip all of them down to bare metal. For this I use
a combination of acid dip, sand blasting and a wire wheel (and plain
old elbow grease); there is probably some damage lurking
under the old paint
(including rust, dents or chips), and this needs to be
rectified before the surfaces are primed again. There's
no easy way around this, the only way to get a
professional end result is to spend time on the
preparation and as a result repairing any damage often
takes longer than the paint process does.
Dents on steel parts can be removed leaving very little
(if any) trace using a panel beater's hammer and dolly
with any remaining ripples and scars equalised using filler primer
and a skim of fine filler (I use the word skim here; the
amount of filler used is as small as possible, normally
just to cover any remaining pits in the metal). Rust is treated in a similar way; all
trace is removed with wire wheel and chemical rust
converter, after which any scars are smoothed out with
fine filler and lots of sanding. Chips and nicks in
aluminium are handled in a similar way; unless the
damage is severe they can usually be rectified given
enough time.
Once the damage is repaired the parts are primed (as
quickly as possible with steel parts as they pick up
surface rust in no time at all), then sanded
and base coated using a combination of airbrush and
miniature spray gun. The paint I use is mixed to order
to match the original Quad colours, several coats of
this are followed by several more of matt/satin lacquer
to help make the parts durable in use and seal in the
colour against UV.
Depending on the condition of the amplifier, this
process may take substantially longer than the rest of
the restoration.
Back to contents
Anodised surfaces: As mentioned earlier,
the black sections on the front of the amplifier are
anodised aluminium. This is a chemical plating process
and thus completely different to painting. The
advantages of anodising is that although it's a paper
thin coating compared to paint, it gives aluminium a
harder wearing finish than any paint or powder coating
can do and the reason it gets widely used in industry is
that it's just a more efficient and convenient way of
coating parts during mass production.
 |
|
Anodised aluminium when
cleaned & polished |
Anodising is a tough one for me; I could paint these
parts and they would function and look fine, however the
issue (apart from paint's lower durability) is that it's
very difficult to match an anodised surface finish with
a paint finish - and if like me you're a stickler for
detail, this is a problem.
If you click on the image to the left, although the
surface is barely satin it has a definite pearlescent
quality which still shows the brushed finish of the
metal underneath; this is a quality of anodising which
is difficult to replicate with paint. Sure, I could use
a metallic black, but it still wouldn't look right to a
critical eye and
also the extra thickness of the required primer and
clearcoat would hide the brushed finish.
At the moment I take these sections apart and then soak
them and polish them by hand and fill any small digs
with black dye; this isn't a perfect solution, but most
of the time it gets a near mint result. If I get an amp
in such bad condition then I'll paint the anodised parts
(and the end result will still look very good);
otherwise I leave it as near original as possible.
All this said, I am currently experimenting a lot with
plating and anodising (I started off with Nickel plating
as it's handy for bolts and things when I'm working on
my vintage cars), being able to plate and anodise is
something which will be very useful on audio
restorations. Black anodising is a bit of a pain (the
parts need to be impeccably prepared, degreased and
pickled, then anodised, then dyed and sealed) and
getting anything resembling Quad's factory finish will
take a lot of work, but I'm hoping that I'll be able to
develop this and offer some plating and anodising
services on my restorations sometime in the near future.
Back to contents
The end result
The images below show the difference in finish on a 405;
the image to the left is how it arrived (including dents
and scuffs), the image to the right is of the same
amplifier taken a few days later after being restored:
Before and after photos of a restored
Quad 405
Back to contents
Back to Quad 405 main page
Last updated May 2014
All photographs and text on this website
unless otherwise stated are © 2008-2014
Briarsfield Hi-Fi -
www.briarsfieldhifi.co.uk
