SHEET 3, separate
main power supply chassis for one channel.
Shhet 3 shows
what is within the 25Kg separate power supply unit supplied
with each
mono bloc 300W amplifier.
Basically it is a linear type of supply with a large power transformer
with a
GOSS E&I laminated core rating for 1.9kVA, and a winding rating for
600VA,
so that temperature rise is negligible,
and fault conditions can be sustained for
lengthy periods.
There is excellent natural regulation for the B+ anode/screen supply.
A voltage doubler rectifier is used with a pair of 6 Amp silicon diodes
with
PIV = 1,000V.
There is provision to reduce the B+ should anyone wish to operate
different
octal output tubes in different conditions. It is shown with the
200Vrms tap
on the HT winding in use, but if
the 157Vrms tap be used the B+ would be
about +390V and the amp could have higher plate current
for less AB1 power
but more pure class A operation into a lower value RL.
The 135Vrms tap would produce about +340V, and allow the use of 12 x
6V6
as the cheapest option to obtain about 90 Watts of total output power.
There is a
CLC type of filter for the main anode supply, arranged so that
caps C1
to C8 form two caps of 470uF each side of L1, which is a 1.8H choke.
With a dc
draw of 500mA, there is 2.34Vrms of ripple voltage at the bottom
of C7,C8. The C1 to C4 capacitance plus
choke of 1.8H form a second order
filter with an attenuation factor of 0.003, so the
voltage ripple at the top of
C1 and C2 is only 0.007Vrms.
The resonant frequency of the LC filter is 5.5Hz.
The power supply has been designed to work on 50Hz mains, so operation
with 60Hz results in lower heat losses.
I have found that after being on for 4 hours, the transformer has a
temperature
rise of less than 5C above the ambient room temperature, and despite
the size,
operation is
silent.
The choke
filter is arranged so it is placed in the 0V rail line instead of the
B+
line so that there is no constant dc voltage strain on the insulation
within the
choke. The choke
winding resistance is 9 ohms, so choke voltage drop is
low and little
power is wasted in the choke.
There is a
+17Vdc supply to operate relay 1 and 2.
At turn on the HT winding current surge is limited by R3 for 4 seconds.
In that time the B+ will have risen to 2/3 of its final value and then
Rel-1
is closed to shunt R3. There is a second current surge as the B+
reaches
its final value of
over +500V in about a second. This stepped B+ voltage
turn on results in about 1/2 the mains inrush
current than if there was no
stepped turn on, so a lower value mains fuse can be used. The two PN100
bjts and 15k feeding C10 act as the 4 second delay
circuit.
Relay-2 is
used to interrupt the HT winding on the transformer when a
fault occurs
and it is normally closed to allow the ac from the HT winding
to charge the caps via the
two diode rectifiers.
However, should one or more output tubes become faulty and conduct
excessive idle current for longer than about 3 seconds, ie, say 60mAdc
instead of the normal 35mA, the active protection circuit
will detect
the excess tube current and trip an SCR seen on Sheet 7, thus grounding
the Rel-2 coil and it opens to
interrupt the B+ anode supply.
The amp will still have its heaters on, but a red LED will indicate it
is in
fault mode,
and it will stay like this until someone provides service.
The amp can
be "reset" by turning it off, waiting 10 seconds, then turning
back on.
Should the amp protection circuit trip again, there *is* a problem
which
needs fixing.
It should thus be impossible to wreck an output transformer or other
parts
due to prolonged overheating.
No owner
serviceable parts are within the amplifier, and only qualified
and well
experienced technicians should ever service such an amp.
There two
umbilical cables from each power to each mono bloc amp.
The cables are easily flexible cables each with 5 sub-cables and each
is
multi-stranded and unlikely to become unserviceable in the foreseeable
future. The type of cables used are normally meant for many heavy
industrial
applications where 415Vac 3 phase power is applied to mobile
gantry cranes etc.
The heater currents from each of the two 12.6Vac heater supplies for
the tubes is 5.4A per phase. One of the umbilical cables is devoted to
the heaters only and its
current rating 40 amps per phase, and so there
is little possibility of ever experiencing fried cables if
a short circuit ever
occur.
A mains fuse will blow.
The other umbilical cable has 15 amp ratings for each of the 5 wires
within, and has excellent insulation resistance since one of the wires
carries
the +500V B+ supply. It is no more dangerous than a normal flexible
mains
cable.
If you keep a pet lion who likes chewing on cables, then try to wean
him
off the habit; I cannot guarantee that such cables will not sustain
damage
due to your
private household arrangements.
Welded steel stands for the amps could be made to bolt amplifier
chassis
and PSU close togther to reduce the likelihood of stretching cables due
perhaps to an
amplifier falling off a bench.
The amplifier end of each cable is hard wired into each amp chassis.
The power supply ends of cables has an octal tube plug which has a
reinforced location spigot to prevent easy breakage.
The two cable plugs have been painted red and black, and the cable
sockets
are also red and black so you know which cable goes into which socket.
Should the two plugs be reversed by accident, the B+ supply will be
shunted through the heater windings to 0V, and a mains fuse will blow.
The power supply enclosures have a plywood
block which is fixed to the
casing by 1 external and 2 internal screws. The block has neatly sized
holes to suit the
external dia of the cable plugs thus assuring that sideways
force on the cables will not damage the
plugs, an once plugged in,
little fingers cannot reach behind the plug to touch exposed pins if
the
plug isn't fully
inserted.
It is
impossible to remove the case covering the power supply chassis
without
first removing the fuse cap and IEC mains cable.
The umbilical
cables :-
The on-off switch is mounted on the cases of the power supplies,
so they should be mounted within easy reach to turn the system on or
off.
S1A is the
mains entry switch on the above schematic.
Two values of input fuse are used depending on the applied mains
voltages
which can be altered over a wide range from 100Vrms to 240Vrms.
An 8 amp slow blow fuse is needed for 100V to 120V mains,
and a 4 amp slow blow fuse is required for 200V to 250Vmains.
S1B is 1/2
the mains switch which connects a 1K resistance from the B+ to
0V while
the amp is turned off. This discharges the supply capacitances
quickly after turning the amp
switch off. This capacitance discharge feature
stops a high dangerous B+ voltage of
up to +520V remaining present should
the amp be turned on and off quickly by a service person
without heated
tubes which normally quickly drain the B+ supply rail.
It also allows testing the power supply safely without the amps being
plugged in.
There is a second fast discharge resistance and relay within each amp
chassis,
see Sheet 4.
The +17V supply in the above sheet 3 is taken out of the supply on pin
six
of the octal socket containing the B+ and OV and error signals, painted
red on
the amp PS
and cable plug, so that +17V can power the relay in Sheet 4.
So at turn on, the relay opens, allowing the B+ to rise, but at turn
off,
the 820 ohm x 10w R shunts the B+. But should anyone remove the
red power supply cable from the power
supply, the electrolytic caps
will take some time to discharge through the
output tubes and during this
time, the plug pins 4 & 5 will have the B+ voltage on them.
If there was no plate current or tubes installed, this could pose a
danger
for anyone touching the pins of the plug.
So the the use of an 820 ohm resistance will discharge the effective
300uF
caps from a maximum of +500V to +68V within 0.5 seconds, and so a
shock to
anyone is extremely unlikely. Under normal operations with
tubes all fitted, the discharge of the
power rails to less than +68V will occur
even faster should anyone remove the red cable with the amp turned on.
With all cables correctly plugged in, the rails will be discharged and
voltage
reduced from +500V to +68V when the amp is switched off within 1
second,
even without tubes
plugged in and thus danger to service people working is
reduced. With all tubes
conducting, rail discharge at turn off is even faster.
The fast
draining of the B+ rails at turn off also prevents the slow reduction
of signal levels with large amounts of distortion which could damage
tweeters if the
music is left running when the amps are turned off.
However, mains power interuption will not cause the fast discharge to
occur.
In practice, such power interuptions are unlikely to ever damage a
speaker
because the rails will discharge
fairly quickly because of the high current
draw from the B+ supply to the output tubes.
The supplies run cool and do not need any special ventilation.
The power amp chassis run quite warm.
They DO REQUIRE GOOD VENTILATION!!!!!!
SHEET 4,
shunt regulated screen supply.
The Sheet 4
above part of the power supply is located in the amplifier
chassis.
It has a
solid state regulator for the screen supplies to the 12 output tubes.
It consists of an MJE340 and BU108 mounted on a heatsink under the
chassis.
The two easily available discrete BJT devices are connected as a simple
darlington emitter follower pair. BU208 is also usable. Should
excessive
screen currents ever flow, R5 will develop a voltage
across it thus reducing
the voltage across the BU108, and it will stop regulation when
current
reaches 150 mA and allow the screen voltage to fall.
Excessive output currents are also limited by R6 4r7, so that if too
much
current should suddenly flow in R6, then the 4 diodes from the top of
C5
to MJE340 base will conduct and
the base voltage reduced quickly, and
thus reducing emitter output voltage. The circuit is protected against
reverse flow of supply voltages
applications by the diode from the output
to collectors. But under normal operation, the screen voltages held
very
steady
between low and high power when dc screen currents vary
considerably with
signal levels.
The supply to
the input stage is shunt regulated by the string of zeners
bypassed with C1.
The lower RHS
of the schematic shows the entry points for the two umbilical
cables
from the remote power supply.
I cannot
recommend having 12 bias adjustments per chassis, 24 in all in
a stereo
system, with all tending to be slightly interactive while you set
them.
And many owners often fail to adjust their bias correctly which can
lead
to early tube failures and expensive repairs. So a mixture of one fixed
bias
voltage and self-adjusting cathode bias
is used.
The main
power supply capacitors for the B+ applied to the CT are located
in the
remote PS box so that should anyone remove an umbilical cable
from a power supply
without turning off the amplifier at its switch,
then the B+ rail voltage appearing on
pins 4 and 5 of the red cable will
fall rapidly due to plate current in the output
tubes discharging C1, C2,
and C3 within a second or two.
This is assisted by the use of a relay shown on Sheet 4 whose action is
explained above.
R10 and R11 on sheet 4 bias the contacts of the DPDT relay so there
is
only 1/2 the B+ across the pairs of contacts while the amp is running,
so less
likelihood of an arc. The R9 limits the peak discharge current to
0.6 amps, so the relay
contacts will last well.
Do not allow
pet lions or your pet dog near the amps. Do not set up the
amps near
swimming pools, and please do carefully think about where
and how one keeps such
devices to ensure domestic harmony.
Sheet 8.
This is the
layout of the heater supply within the amp chassis.
Two phases of 12.6Vac are brought into the chassis from the remote
PS
to be applied as shown, and to minimize hum.
rectifiers derive +/-17Vdc supplies to be used for the input tube to
reduce hum and to provide a -14V fixed bias for the output tube grids.