Mackie M-2600, Service Manual

Page: 12 / 32

12
SOA protection
There are two SOA detectors in the M-2600: One that monitors steady state output stage
dissipation and one that monitors transient high current events. Both detectors need to monitor
the output stage current.
A voltage, representing output stage current, is derived by first sampling the voltage drop
across the output transistor emitter resistors. These voltages are rectified via U9A R87, R88, R59
and R89. The rectified voltage is converted to a current via U9B, Q23, and associated. This current
is referenced to the output line and must first be re-referenced to the +115V supply via Q14, D47
and associated. The current available at the connector of Q14 is converted to a voltage that is
ground referenced via R138 and R162. This is buffered via U8A and adjusted via R162 such that
1 amp of output stage current is equal to 100mV of detected output.
This current is compared to what represents a safe operating current at U4A (part of the
transient SOA-Fault detector). If the current is greater than allowed for more than 10mS (Timing
determined by R77 and C36), U4B sends out a fault indication which mutes the amplifier via U3C
and associated, and fires the short led for 5.5 seconds (U3B and associated).
While the output stage is quite tolerant of very high current demands for a very short time, it
would destroy itself if it had to provide these currents on a long-term basis. The second detector
actually looks at long-term dissipation in the output stage, and if excessive, will mute the
amplifier. We already have a signal that is proportional to output stage current as described
above. This signal is fed to one input of analog multiplier U7 via R80 and R84.
U6A and U6B and their associated components look at the -100V supply and the amplifier
output line. These voltages are converted to currents via R101, R129 and R130 and are rectified
and summed such that for every volt between the output line and the supply rail, 100mV is
presented as output. This second signal is presented to the other input of multiplier U7 via R133
and R82. Output from U7 is in the form of a current and is converted to a voltage via R136 and
U8B. The gain of this multiplier is such that for 200 watts of output stage dissipation, 1 volt of
output is obtained. The output of U8B is instantaneous power and is long-term averaged via
R135 and C40. The cutoff of the filter is quite low in frequency, such that even a 20Hz signal on
the amplifier output results in very little ripple voltage at C40. If the voltage on C40 is greater
than 5V (equivalent to 1000W of output stage dissipation), U4C will toggle, muting the amplifier
and firing the SHORT LED as above. Approximately 200mS of dissipation above 1000W is required
to cause a fault.
DC Fault Detection
The amplifier output drives the subsonic lowpass filter of the DC fault detector (R154, C65
and C64). If the amplifier were to fail and short to the +100V rail, C64/C65 will charge to more
that 5V in a few hundred milliseconds. Local +/-5.1V references on U2C and U2D set the threshold
where: when the amplifier output voltage is “stuck” to greater than +/-5.1V, either comparator
output will go low. This will open Q27, and in turn open the speaker relay on the respective
channel board. When the amplifier is initialized, “MUTE-VA” is high and this biases C64/C65
above 5.1V via D14 and R36. This offset results in a delay of about 2 seconds from when the
amplifier turns on (“MUTE-VA” low) and when the speaker relay closes.
Low Voltage Detector
D15, D2, and R7 supply a local 5V reference and are supplied current from the unregulated
+30V supply via R32. 22VAC from the +/-30V windings is half-wave rectified via D26 and stored
by C18. R53 limits the absolute peak current to protect D26. The voltage on C18 is divided down