I am attempting a repair of my 1980s NAD 7020 receiver, the first version with one large board containing most of the circuitry. For those that don't know, the 7020 is essentially the 3020 integrated amplifier with a tuner added, and like the 3020, has a preamplifier section that can be connected to an external power amplifier while bypassing the internal power amplifier. I have determined that the preamp and all the inputs and controls work well, but the power amplifier has a dead left output. Inspection of the circuit board showed that there is a burned transistor and associated resistors (please see photos). The burned transistor is Q609, a BD139 unit.
I have obtained replacement parts, including BD139G transistors, resistors to replace the damaged units, and all new electrolytic capacitors. I removed the burned transistor from the board for further inspection but since I have never performed this kind of repair before I think I need more information before proceeding further.
The BD139 units are attached to the heat sink with sheet metal screws that go through the body of the transistors and have white thermal compound (which has hardened with age) between them and the heat sink. There is also a thin wafer (Mica?) between the transistors and heat sink which has thermal compound on both sides. The wafer on the removed transistor appears to be undamaged. Both the original and the replacement transistors have a metal plate in the center of the surrounding plastic on the side that attaches to the heat sink. This metal plate has full continuity to the center leg of the transistor (no resistance). The heat sink has full continuity to chassis and signal ground. I am confused as to the type of thermal compound I should use on the replacement unit (electrically conductive or not?) and don't know whether I should replace the wafer given that it is undamaged and I don't know what material to use.
I would appreciate some guidance before I proceed with my repairs. What type of thermal compound should I use (electrically conductive or not), should I replace the wafer, and should I replace the compound and wafer on the transistor in the other channel, given that the compound has hardened? I have performed circuit repairs before this, have experience with soldering and unsoldering, and possess the schematic and service manual for the 7020, so I am confident to move forward if I can get my questions answered. Thanks to all who reply.
I have obtained replacement parts, including BD139G transistors, resistors to replace the damaged units, and all new electrolytic capacitors. I removed the burned transistor from the board for further inspection but since I have never performed this kind of repair before I think I need more information before proceeding further.
The BD139 units are attached to the heat sink with sheet metal screws that go through the body of the transistors and have white thermal compound (which has hardened with age) between them and the heat sink. There is also a thin wafer (Mica?) between the transistors and heat sink which has thermal compound on both sides. The wafer on the removed transistor appears to be undamaged. Both the original and the replacement transistors have a metal plate in the center of the surrounding plastic on the side that attaches to the heat sink. This metal plate has full continuity to the center leg of the transistor (no resistance). The heat sink has full continuity to chassis and signal ground. I am confused as to the type of thermal compound I should use on the replacement unit (electrically conductive or not?) and don't know whether I should replace the wafer given that it is undamaged and I don't know what material to use.
I would appreciate some guidance before I proceed with my repairs. What type of thermal compound should I use (electrically conductive or not), should I replace the wafer, and should I replace the compound and wafer on the transistor in the other channel, given that the compound has hardened? I have performed circuit repairs before this, have experience with soldering and unsoldering, and possess the schematic and service manual for the 7020, so I am confident to move forward if I can get my questions answered. Thanks to all who reply.
Search carefully, usually the BD135 is not broken alone. In fact I do not find until now only thermal control transistor broken and the rest ok. If you short C-E this transistor the amplifier sound work relatively OK but with distorsions at low level. So not this is your big problem.
The thermal pasta is ceramic non conductive. The mica foil if is not damaged can be reused. You do not need to replace pasta on all transistors, only if you unscrew the transistors.
The thermal pasta is ceramic non conductive. The mica foil if is not damaged can be reused. You do not need to replace pasta on all transistors, only if you unscrew the transistors.
Thank you for your reply. I know very little about power amplifier design, so I would appreciate any more information anyone can offer.
Amplifiers from that era were often a bit optimistic about reliability issues, so this kind of failure was very common. In this case I see that there are no emitter resistors and the power transistors and driver transistors used are being pushed to their limits, and there is no output protection against speaker shorts. In those days, I would have upgraded the output transistors to 2N3773 and 2N6609, and ~4Amp drivers like mje243, mje253 being careful to get the leads correct and adding a small heat sink to the drivers. These may be a bit overkill but unless you add current limit circuit, you need the extra robustness. And I would add emitter resistors. 0.22 Ohms is typical but maybe 0.1 would be enough. In some cases, the best plan is to replace the whole power-amp, reusing the power supply etc. There are a zillion such modules available from Amazon etc. I would probably design a new power amp PCB to fit the same space, but with a better circuit.
It's amusing that the burnt resistors are not those mounted on ceramic spacers. Probably because they were thinking of normal heat and not failures. You realize that burn VBE multiplier means the drivers shorted?
It's amusing that the burnt resistors are not those mounted on ceramic spacers. Probably because they were thinking of normal heat and not failures. You realize that burn VBE multiplier means the drivers shorted?
Last edited:
Thanks for the suggestions. In this receiver all the audio circuitry is on one large board that occupies almost all the space of the case, so major modifications are not possible. This board does not have one square centimeter of open space---the components are packed very tightly. I have power amplifiers to which I can connect the output of the preamp of the 7020, but I hope to effect a repair of its own amplifier if possible.
I would appreciate any information that someone with experience repairing the 7020 (or the 3020 upon which it is based) can offer.
I would appreciate any information that someone with experience repairing the 7020 (or the 3020 upon which it is based) can offer.