3
should have near infinite drain to source resistance when not driven. The voltage across the large 20,000
ufd capacitors should remain for several days after power is shut off, if not, one or more of the transistors
are most likely leaking.
C
onclusion:
What fixed this amplifier is the replacement of a SMT NPN transistor and the change of bias for
the input and driver stages. I ended up replacing all the MMBT5088 transistors with 50C02CH-TL-E
NPN transistors made by ON Semiconductors. The MMBT5088 transistors are only rated at 30 volts from
collector to emitter, these amplifiers use plus and minus 24 volts. Over time they become leaky and don’t
turn off like they should. After replacing them, the thumping during turn on and turn off was gone. To
cool down the 4 driver transistors I added a 1k pot and a 330-ohm resistor across D5 and adjusted it for .8
volts across R44 and R48. This resulted in a bias current of about 16 ma going through Q14, Q15, Q16
and Q17. Even at full power, the current through the driver transistors only changed by a few milliamps.
Now the transistors won’t cook and fail. I also changed R12 and R10 in the input stage to reduce the idle
current though transistors Q4, Q5 and Q6, Q7 to 100 to 120 ma. The bias changes I made should not
affect the performance of the amplifier unless it is driven hard. I changed all the electrolytic capacitors but
not the 20,000 ufd ones. I also changed all the MOSFETs in the output stage only because the thumping
damaged the P channel MOSFETs. This can easily be checked by measuring the voltage across the
20,000 ufd caps. With the amplifier turned off, the caps should take a day or so to discharge, if they
discharge over night or in a few hours, a transistor is leaking, or of course the associated cap may be bad.
Do the same modifications described here to the other channel as well. Good Luck
Engineer Ben
