I have a GFA-545 (original, not the II) that has a large DC offset on the left channel. I've followed the service guide and biased it to 6.5mw on the test points, but am currently seeing 700mv at the terminals. I've pulled the left channel board and am powering it via two bench supplies, with no difference in behavior. All of the transistors test out and seem to be in spec. I've tested each of the emitter resistors, and they are also in spec. If I feed the amp a since wave, it will drive an 8ohm load to a bit over 100w per side before clipping, and looks clean. No difference between L and R.
Where should I go from here?
Where should I go from here?
Thanks, I'll check that out. I did clean the board, it was a bit crusty with some residue and corrosion, but all the traces seem to be intact. I also touched each component leg to reflow the joints.
R603 is perfectly in spec. Could this just be user error? If I short the input + to ground, the offset goes away. I'm guessing that since there is no DC blocking on the input this could just be stray current and a false indicator of a problem?
If you have a good channel and a malfunctioning channel then the logical steps would be to compare their voltages in various points.
700mV - if measured correctly - is a bit too much for my taste to just "write off".
It seems to me that the hfe of Q601 has probably deteriorated dramatically and is the likely cause of the offset. Measure voltage at base of Q601 (i.e. R603) with respect to ground with input open. In should be less than -137mVDC (magnitude) if hfe is minimum spec of 160. Given your experiment showing improvement with shorted input, I imagine base will be closer to -700mV.
If the input pair is well matched, there will be near 0mV difference between the bases of Q601 and Q602. There should equal voltage drops across R603 and R623. Ideally, there will be no voltage across R619; any drop indicates leakage through C609 and that current will flow through R623 and push the output offset more negative.
Good luck!
If the input pair is well matched, there will be near 0mV difference between the bases of Q601 and Q602. There should equal voltage drops across R603 and R623. Ideally, there will be no voltage across R619; any drop indicates leakage through C609 and that current will flow through R623 and push the output offset more negative.
Good luck!