2
the bias level is such that the transistors won’t wonder into a thermo runaway condition and the driver
stage will not fall out of class A operation.
The change in the bias of the driver stage affected the current through the output transistors as
expected. P2 controls the current and should be set to 400 milliamps as explained in the owner’s manual.
The output transistors operate in Class B as a push pull amplifier. It is important that current through all 8
transistors never drops to zero in order to prevent crossover distortion. Another test showed that at full
output power the current only nears zero if the bias current is set at or below 250 milliamps. Since the
owner of this amplifier will not be using it anywhere near its limits, the bias was set at 350 milliamps.
This will provide plenty of current at crossover and has the added benefit of cooler transistors and
heatsink. I consider the overheating transistor problem solved, but if my friend encounters a problem, I
can always increase the bias level if needed.
The source of the turn on and turn off thumping problem turned out to be far more difficult to
locate. The use of an oscilloscope showed a 50-volt spike at the amplifier’s output right after turn-off. The
source of the spike appeared to be everywhere and nowhere specific. All DC measurements were normal.
The amplifier’s output is feedback to both sides of a current mirror consisting of Q8, Q11 and U9. This
made it difficult to pin down the exact source. The first area I looked at was of course the turn-on turn-off
control circuit which is part of the power supply and consist of 3 transistors, Q201, Q202, Q203, a few
diodes, capacitors, and resistors. This circuit controls Q10, a transistor biased to sink around 11 ma. It
delays the turn on of the current source at power up and shuts it down very quickly at power off. This
circuit was found to be operating correctly. By using a multi-channel oscilloscope, it was found that Q12
was still conducting for a short period of time after the current source had been shut off. It was this
current that generated the pulse that was applied to the base of the P channel MOSFETs. The replacement
of Q12 ended the thumping problem but during the process of finding Q12, it was discovered that several
of the P-channel MOSFETs were leaking. Probably damaged by the erroneous spike. It is easy to tell if
there are leaky transistors in the output stage. The transistors are all enhancement mode MOSFETS and
