Audio Engineering

Power Supply Design 157

on, will connect Q1 base to the 0-V rail, which, in turn, causes current to be drawn from
Q2 base, which causes Q2 to remain in conduction even if the original input voltage is
removed. The trip voltage will arise if an excess DC signal (e.g.,  10 V) appears across
the LS output for a suffi cient length of time for Q1 to charge to  5 V. Returning to Figure
5.6 , when the circuit trips, the forward bias voltage present on Q19 base is removed and
Q17 is cut off and remains cut off until the trip circuit is reset by shorting Q2 base to the
0-V rail. If the fault persists, the supply will cut out again as soon as the reset button is
released. An electronic cut-out system like this avoids the need for relay contacts or fuses
in the amplifi er output lines. Relays can be satisfactory if they are sealed, inert gas-fi lled
types, but fuse holders are, inevitably, crude, low-cost components, of poor construction
quality, and with a variable and uncertain contact resistance. These are best eliminated
from any signal line.

5.12 Typical Contemporary Commercial Practice ........................................................

The power supply circuit used in the Rotel RHB10 330-W power amplifi er is shown
in Figure 5.8 as an example of typical modern commercial practice. In this design two
separate mains power transformers are used, one for each channel (the drawing only
shows the LH channel—the RH one is identical) and two separate bridge rectifi ers are
used to provide separate 70-V DC outputs for the power output transistors and the

Trip voltage Clamp out

C1

100 uF

Q1

Q2

0V 0V

Reset

R2

47 k

R3

1k

R4

4k7

47 k

R1

0V#1

Figure 5.7 : Trip circuit.