In your initial post you only mention replacing one MJ15003 (output) transistor, and seem to refer to the 132-081 as an output transistor, which it isn't. The very commonest cause for parts getting fried on these boards are one or both output transistors (132-070) shorting out. If one shorts out, always replace both, and it is nearly always a good idea to replace other semis ahead of them in the circuit-- drivers, predrivers, differential pairs, etc. Why? Because when an output shorts, a very high current is then drawn through much the circuit, and that causes stresses in the parts that may cause them to weaken, if not fail outright.
Did you change both output (MJ15003) transistors, or only one? Also, check to be sure that the power supply rail fuses have not been replaced with more than the normal values, 5 amperes in this case. For testing purposes, you can reduce this value temporarily to provide some safety on the initial power-up, say perhaps 2 amps.
Speaking of fuses, why didn't they keep the damage from occurring? One, a fuse a a last-ditch protector in most cases when you are dealing with semiconductors-- it's a race against time whether the fuse opens, or-- the device fails. The fault current has to flow through both, after all. Also, your model likely has what is commonly called a "crowbar" speaker protector, which, when it detects the presence of a DC voltage on the output circuit, literally shorts the output to ground, with the idea of blowing the fuse(s) before the amplified DC can damage your speakers. (The idea is to possibly sacrifice the power transistors to save the speakers.)
So now I have 2 new mj15003s on the left (problem) channel, and replaced the burned resistor(s) and re-oriented the A970BLs which I replaced the BC?416s with. Is that a kosher substitution? I can't find where I heard that when I ordered the A970BLs. The specs seem...like it should work, now that I snapped out of the dyslexia that had B, C, and E....confused. (E is opposite of a 416, with the package rotated 180deg, B and C need to switch places as well)
I then started checking everything with the digital multimeter I borrowed (had only analog needle gage before, so no checking the 22 ohm and 33 ohm resistors etc. very closely with that)
The digital multimeter was skittery as ____. Very hard to get any kind of stable, consistent reading. By comparing the good and bad output board....they at least seemed to skitter in parallel . Eventually I began checking the far right lead...trace or whatever it's called, where the board plugs into the slot, red lead of the multimeter on #1 according to the service manual, black on #2, then #3, then #4 etc. All of these were close to identical until #6, which leads directly to the "metal can" or TO-72 transistor, which is ID'd in the service manual as a diode (D414), but spec'd as an epitaxial planar "N channel junction FET".
D413 on the good output board seemed to have 80ohms resistance across it (two leads are twisted/connected...gate and source?)
D414 on the problem board is shorted/0.3-1.1ohms.
I can find no other deviation between the two boards. Some post here on AK spec's that transistor as a 2N5245
, but there does not seem to be any with a similar "metal can" package new or NOS, just TO-92. Does anyone know if is this a good replacement? Should I replace both of the original, metal package transistors (L D414, R D413) at the same time, or does it matter?
More importantly, does anyone have an inclination as to whether that shorted transistor could have been causing the resistor in the attached picture to flame out less than a second after power on? One lead of 132-079, a driver transistor, goes directly to that resistor. Does D414 control that (enough volts coming through 132-079 to fry a 10ohm resistor) somehow? The ...driver transistors?....including 132-079 seem unharmed.
Just trying to get up the nerve to turn it on again
Hello folks,
