sheet 6
The above may seem like a horrible jumble of lines, but in practice is is a
simple bias detection condition monitor and bias fault detector network where
there are 12 output tubes.
The circuit was easily constructed using a board
with hooked solid wires for tracks and components were surface mounted and
soldered between wire tracks. Longer wires from the board to each cathode are
are easily done single strand hard wiring taken from some multi-colour
multi-wire telephone cabling.
All the voltages involved are lower than
50Vpk.
The signals from the 12 output tube cathodes are fed into the 12 input points
at the top of the schematic.
The signal consists of any ac or dc voltages on
at the cathodes.
From each of these input points there is a test point in the form of a
phillips head screw in the circuit board which is accessible from outside the
chassis through a two rows of 6mm holes corresponding to the output tube
positions.
A test probe from a cheap multimeter can connect from the to the
test points to the chassis to obtain the dc voltage at each cathode, and so in a
minute one can rapidly check all the bias voltages at each cathode.
Such dc bias measurements are always made at zero signal voltages.
From each test point, the dc voltage is divided by resistances R1 and R13, R2 and R14 and so on across the board, so that if there is 35mA flowing to 0V from each cathode, there will be about 18V dc at each cathode test point and about 4.4vdc at each junction of the two R which have values of R1-47k and R13-15k.
During normal operation, we want this 4.4Vdc to remain free of signal voltages because the following fault circuit ( on sheet 7) has been designed to react only to Vdc changes. So 220uF has been strapped across each of all the R13 to R24, and so the output from each RC filter is largely free of all signal voltages.
Each of the 12 points where 4.4V is to appear are all connected through 12
diodes D1 to D12 to a common rail shown as point Z.
The voltage at point Z
is fed to the input of the circuit which works an SCR which can only turn on if
the input voltage at point Z is 7.0Vdc which could occur if too much dc flowed
in one or more cathode bias resistors.
If 7.0V were to appear at point Z, it
means that one or more cathodes must be at about +32V,
indicating that there
must be approximately 65mA dc of tube current flow, or nearly twice the
usual idle bias amount, which is a fault condition at idle.
Should one
of the 12 output tubes suffer a fault and conduct too much current, then its
cathode voltage will be above the rest and its voltage will be divided by its
own voltage divider to give a voltage at point Z which is is then applied to the
SCR circuit on sheet 7 below.
Due to the direction of the diodes, one tube's
excessive bias voltage cannot create a current flow to other cathodes at lower
voltages, so point Z voltage is controlled by whichever cathode is at the
highest voltage level.
To ascertain which cathode is the cause of a bias problem, one has to be
quick with the dcV meter to check the tube cathode voltages as they rise and
before the faulty one trips the SCR and turns off the B+ supply.
Alternatively, one can temporarilly disconnect the wire to point Z and allow
all the cathode voltages to rise without a faulty one causing the active
protection to work. The faulty tube will easily be spotted.
The amp may be reset by turning it off at the mains, and back on after 20
secs, and measuring the tubes that could not be measured during the first
attempt to measure cathode voltages.
The 12 diodes D12 to D24 shown uppermost in the schematic are in two groups
of 6 diodes each.
Each of the two lots of 6 diodes are taken from the
cathodes of the 6 tubes on each side of the push pull cathode circuit
which
works with oppositely phased signal voltages.
D13 to 18 each have a separate
100ohm 5 watt R in series ( R25 to R30 ) and then all R connect together
to give point Y which is connected to the collector of Q1 on sheet 5.
Similarly, D19 to D24 have separtae 100 ohm R in series R31 to R36 ) and
then all connect together to give point X
which is connected to the
collector of Q2 on sheet 5.
Current in the 100 ohms plus diodes can only flow when the current shunting
transistors are turned on by heavy
dynamic signal currents flowing in the
output tube cathodes. Since the transistors Q1 and Q2 remain turned off
for
most of their life they do not conduct any current through the D13 to D24 unless
the escessive ac current
in the grouped cathodes causes them to turn on.
sheet 7 
In this sheet 7 schematic, a dc voltage appearing from point Z is applied to
the high impedance input of a darlington pair emitter follower Q4 and Q5 and the
emitter voltage is applied across the 5.6V zener diode.
If VZ = +7V, current
will flow in the ZD, and through R7, and R8 and raise the voltage at the gate of
the SCR and it will *latch* on and its anode voltage will be pulled from 17V
down to near 0V, therefore turning on a relay in the remote PSU, and thus
turning off the B+ supply.
The SCR also turns on Q3, the transistor which also operates the clipping indicator led, and the led indicates that no B+ is present; ie, that a fault condition exists.
The clipping indicator is fed a current through a 1M resistor R2 from the
anode of the input triodes from a point C on sheet 1.
This low current
signal has no effect on the audio signal flowing from V1 to V2 and V3 on sheet
1.
The signal from the 1M R2 is cap coupled via C1 to the high impedance
base input of darlington pair Q1 and Q2, and Q1 and Q2 turn on when the V1
signal is over a certain signal level of 10Vrms which indicates that clipping
must be occurring, since the signal at v1 anode rises above 10V rapidly because
it is the internal "error" signal within the amp after NFB is applied.
So when you see the clipping indicator flashing sometimes, the music is probably too loud, and maybe you need to ask junior to turn it down lest the police make a visit to ask you.
On the far LHS, the blue ON led is fed a signal from the heater circuit.