I have gfa555ii with full rail voltage 82 volts on left output, nothing on other channel.
Where should I start?
Where should I start?
With full rail showing up on the speaker terminals, you more than likely have one or more shorted outputs. Partially disassemble the bad channel (for access) and check them with a DMM in diode mode. Then, you will need to work your way back through the circuit checking for damaged driver transistors and maybe even VAS transistors and related components. The general rule is to replace one stage further back than the damage (for DIY, not generally in a repair shop setting). Once the damage is repaired, you will have to check to see if the offset servo circuit is still operating properly. You may even contemplate a complete overhaul with new/upgraded parts, like the Hoppe's Brain boards.
Lots of work, but these ole' Adcom amplifiers are worth it IMO.
Dan
Lots of work, but these ole' Adcom amplifiers are worth it IMO.
Dan
Just because a transistor or diode reads 550 to 750 on the diode scale at 1.5 v, and 9999 or ----- backwards, does not mean it is good. I have had transistors that passed that test blow their tops under rail voltage. Iceo test at 12 v seems to predict the ones that will fail. Base open, <255 ua c to e through ma scale of DVM, okay. More, the junction is blown.
What is probably easier than taking parts out, once you have checked fuses and power supplies, is to put a dim bulb tester series the AC input. 60 w tungsten bulb series the AC input. I have mine in a grounded metal box with a circuit breaker so the breaker blows if a wire falls of the edison socket, instead of the wire flying around and hitting you or something valuable. House wiring device screws are made for solid wire, not stranded. If dim bulb goes out under power, you have probably removed most shorted parts.
Then any amplification transistor reading <1 v c-e or any diode with no voltage across it, is bad. Take data with one probe and one hand, then subtract. Black probe of DVM connected with alligator clip lead to analog ground, center of 2 big caps connected to each other + to -. >25 v from one hand to the other can stop your heart. Ne jewelry or metal on hands wrists or neck. 1 v at high current through metal can burn your flesh to charcoal. Wear safety glasses, parts explode, and solder splashes especially desoldering.
Note switching transistors that actuate relays or lamps, can go <1v when they are on. Other models of Adcom have had none of these.
Packs of alligator clip leads rated ~60v can be bought at part-express.com . I find their solder sucker useful too. Also their vary temp soldering iron. Pro alligator clips rated at 600 vac are so big as to be useless for amps IMHO.
Last edited:
Yup, you have to test all three junctions BE/BC/CE and understand how to test PNP/NPN, and as indianajo states above, the outputs may only be failing under higher voltage conditions. But, since this is D.I.Y Audio, I like to start with quick, simple tests. If all of the outputs on the bad channel "buzz" out as good with a DMM, then one has to move on to higher level testing.
I have been an automotive repair technician for forty years. I specialize in driveability, electrical, and hybrids. We get customers in that want this and that tested and are ignoring the fact that the engine is out of oil and has a rod knock. Simple first.
Dan
I have been an automotive repair technician for forty years. I specialize in driveability, electrical, and hybrids. We get customers in that want this and that tested and are ignoring the fact that the engine is out of oil and has a rod knock. Simple first.
Dan
Transistors that fail under high voltage generally don’t come back and work properly when the voltage comes back down. If they degraded and lost voltage handling ability somehow, they would end up dead dead dead and not test “ok” at low voltage anymore.
For small signal types maybe. You can degrade an input or VAS stage so it sounds like lightning and thunder or zeners and then the amp won’t center anymore. But those stages have current limiting mechanisms. Transistor breaks down and something limits you to a couple, maybe 10, mA. Try that with a degraded output or driver transistor. If you don’t have a dim bulb in there it will go BANG. And no one is ever using a dim bulb with their amp when it fails in the field. If the output or driver transistors go bad they burn up. Could the opposite transistor get damaged? Sure. That’s why you’re supposed to replace it too - even if it still tests good.
false. When there are multiple output transistors, some short completely blowing the fuse. Some others are just damaged, and if only the shorted ones are replaced, they eventually fail too. Not immediately. The dodgy ones did pass the dvm diode scale test. They failed at +-12 v Vce with the base open. Use a 47k resistor series the ma scale of the meter to avoid blowing the fuse in the meter on the ones that passed the diode scale test but are damaged. Or design an ohmmeter that has 12 v source voltage. Use this ohmmeter only on power transistors, not on sensitive components.
Last edited:
Say upper transistor fails short. Lower is subjected to 2Vcc at a lot of current. A transistor that is leaky and can barely hold 12 volts is going to either fail open or short when subjected to that. If the opposing transistor didn’t die then it CAN take full voltage. It may not be reliable, scrubbing off a million hours of lifetime in 15 seconds, but is NOT going to test leaky! If it had gotten leaky it simply would have died during the original failure along with the upper.
Ha Ha Ha Ha Ha! I had output transistors pass the DVM diode scale test, work fine at 55 v while amp was on the AC resistance box. Then at 85 volts off the resistance box, the survivors would fail at random times, blowing their tops in the process. MJ15024/25. The one of 20 outputs that passed the 12 v Iceo test, worked okay as a driver. Many pros on here have said, change one output transistor, change them all and the drivers too. Saves time, also exploding TO3's. This amp was blown back through 3 of 4 predrivers, one of the input op amps too, and the 16 v zeners producing +-v for the op amps.
I never said that survivors would be reliable. I said they would block 160 volts with Rbe of some realistic value. At least during initial testing, Without trying to drive a load. 500 mA ICBo or 12V CE breakdown isn’t a possibility - or it wouldn’t have been a survivor in the first place. It would have been one of the completely dead ones.