Hi everyone. I just got this almost pristine looking 555 mk1. DC offset is below 5mv on both channels with inputs shorted.
Problem: Distortion light on Left channel starts to gradually glow as amp warms up, until reaching a fully lit condition. Previous owner says he never had this fault.
Observations:
1. As it warms up, the left channel distortion light starts to gradually glow until fully lit. Takes about 5 minutes to fully lit. Must be related to some component heating up.
2. Amp works and sounds good, both channels working.
3. Bias pot on left channel reacts inversely, that is, it decreases bias current when turned clockwise. Comparatively, bias pot on right channel increases bias current when turned clockwise. both channels CAN be stabilized at 16mV.
4. Reflowed all transistors.
5. Ecaps test good. anyhow, changed the 47 and 4.7 caps on input board.
6. Doing a comparative resistive measurement (in circuit) between L and R channel, all resistors and diodes test good, and all transistors measure about the same, except Q6 (Q609 in newer schematic).
7. Haven’t yet pulled out the mica, film and ceramic caps for testing.
8. There’s a slight hot component smell at the input board when amp is fully warmed. But no visible burnt marks or hot spots.
9. Left channel drains the PS caps reservoir at a much faster rate than the right channel. Almost twice as fast.
Any clues? I’m baffled by the inverse reaction of the bias pot….
Cheers,
Luis
Problem: Distortion light on Left channel starts to gradually glow as amp warms up, until reaching a fully lit condition. Previous owner says he never had this fault.
Observations:
1. As it warms up, the left channel distortion light starts to gradually glow until fully lit. Takes about 5 minutes to fully lit. Must be related to some component heating up.
2. Amp works and sounds good, both channels working.
3. Bias pot on left channel reacts inversely, that is, it decreases bias current when turned clockwise. Comparatively, bias pot on right channel increases bias current when turned clockwise. both channels CAN be stabilized at 16mV.
4. Reflowed all transistors.
5. Ecaps test good. anyhow, changed the 47 and 4.7 caps on input board.
6. Doing a comparative resistive measurement (in circuit) between L and R channel, all resistors and diodes test good, and all transistors measure about the same, except Q6 (Q609 in newer schematic).
7. Haven’t yet pulled out the mica, film and ceramic caps for testing.
8. There’s a slight hot component smell at the input board when amp is fully warmed. But no visible burnt marks or hot spots.
9. Left channel drains the PS caps reservoir at a much faster rate than the right channel. Almost twice as fast.
Any clues? I’m baffled by the inverse reaction of the bias pot….
Cheers,
Luis
Red herring. Just a result of a mirrored layout....
But I'm curious about what triggers the distortion lights on other Adcom amps. You'll draw more speculation if you post a schematic?
Here's the schematic.
Also might add that left ch Q7 gets really really hot when the distortion light is glowing.
A few days after my first post, I reflowed the whole channel. And removed a couple of transistors (Q7, Q8 and Q10) whose soldering didn't look good to me, after te reflow. Sanded the transistors' pins, completely removed old solder from the board pads. cleaned everything super good, reinstalled. and now I get either a solid brightly lit distortion light (and consequently a super hot Q7), or no distortion light at all, but occasional and random blinking. Q7 only gets hot when the light is glowing.
I might have to remove all components and resolder the whole thing with super extra care. It seems like a bad connection. But can also be a flaky component.
Hopefully someone can chime in with ideas on which components to resolder first. Helping me prioritize work.
Andersonix, thanks for your insight about the weird bias, inversely reacting to the trim pot. Both trim pots are located on the same board, and they both should be the same part number (at least per the service manual). So, turning all CCW should take you to the lowest resistance value, and turning all CW should take you to 1 kohm. Why would the left channel's bias current react inversely? I'm still baffled.
Cheers everyone. This amp is super nice and clean, hopefully it can be brought back to life.
Luis
Also might add that left ch Q7 gets really really hot when the distortion light is glowing.
A few days after my first post, I reflowed the whole channel. And removed a couple of transistors (Q7, Q8 and Q10) whose soldering didn't look good to me, after te reflow. Sanded the transistors' pins, completely removed old solder from the board pads. cleaned everything super good, reinstalled. and now I get either a solid brightly lit distortion light (and consequently a super hot Q7), or no distortion light at all, but occasional and random blinking. Q7 only gets hot when the light is glowing.
I might have to remove all components and resolder the whole thing with super extra care. It seems like a bad connection. But can also be a flaky component.
Hopefully someone can chime in with ideas on which components to resolder first. Helping me prioritize work.
Andersonix, thanks for your insight about the weird bias, inversely reacting to the trim pot. Both trim pots are located on the same board, and they both should be the same part number (at least per the service manual). So, turning all CCW should take you to the lowest resistance value, and turning all CW should take you to 1 kohm. Why would the left channel's bias current react inversely? I'm still baffled.
Cheers everyone. This amp is super nice and clean, hopefully it can be brought back to life.
Luis
After doing a “brute force approach”, it was Q6 that got the issue solved after replacement.
My first obvious candidate was Q5, as it controls the distortion led. Next up we’re Q1 and Q2, which I loosely matched using an eBay do-it-all tester-meter board. Next where Q7 and Q4, heatsinked. Now, it took longer for the distortion light to come up, but it did, faintly, light up. Next came Q8 and Q3, because of their close relation with Q7 and Q4 respectively. Still, the darn led kept coming up. Then came Q6, and voilà. After replacement of Q6, bias voltage went to almost zero, readjusted bias, and after 30 minutes on the bench, no more distortion led.
I kept a close look at both bias and dc offset at the output before and after every part replacement. both where pretty much the same after every part replacement. Offset was 30mV inputs not shorted. After replacing Q6, offset went up to 40mV, and bias had to be readjusted. Both values are stable after this.
Replacement transistors were all sourced from Digikey/Mouser, using part numbers suggested by Chris Hoppe (thanks for all the openly shared info!).
Now, just waiting for matched input pairs from Chris Hoppe to finish up this refurb. Will change all transistors on the input PCB and the drivers as well. Will change all diodes as well. Will test with oscilloscope for square wave and sine wave integrity, and hopefully, very soon another nice amp will join my stable.
Also ordered a DC speaker relay protection board from China, 25 amp per channel, it comes with its own 120v/12v PS, so it seems like an easy add-on. Just to keep on the safest side possible.
cheers,
Luis
My first obvious candidate was Q5, as it controls the distortion led. Next up we’re Q1 and Q2, which I loosely matched using an eBay do-it-all tester-meter board. Next where Q7 and Q4, heatsinked. Now, it took longer for the distortion light to come up, but it did, faintly, light up. Next came Q8 and Q3, because of their close relation with Q7 and Q4 respectively. Still, the darn led kept coming up. Then came Q6, and voilà. After replacement of Q6, bias voltage went to almost zero, readjusted bias, and after 30 minutes on the bench, no more distortion led.
I kept a close look at both bias and dc offset at the output before and after every part replacement. both where pretty much the same after every part replacement. Offset was 30mV inputs not shorted. After replacing Q6, offset went up to 40mV, and bias had to be readjusted. Both values are stable after this.
Replacement transistors were all sourced from Digikey/Mouser, using part numbers suggested by Chris Hoppe (thanks for all the openly shared info!).
Now, just waiting for matched input pairs from Chris Hoppe to finish up this refurb. Will change all transistors on the input PCB and the drivers as well. Will change all diodes as well. Will test with oscilloscope for square wave and sine wave integrity, and hopefully, very soon another nice amp will join my stable.
Also ordered a DC speaker relay protection board from China, 25 amp per channel, it comes with its own 120v/12v PS, so it seems like an easy add-on. Just to keep on the safest side possible.
cheers,
Luis
Following up on this thread, today I finally finished replacing all transistors on the amp’s input board. for now, it is true to the original schematic, just now using modern replacement transistors from Onsemi. Output transistors are still the original Toshiba‘s. End result is very nice. Output DC offset is about 35mV with non-shorted inputs, and less than 1mV with shorted inputs. Will burn it in a few days, and maybe next weekend I’ll do some oscilloscope and audio analyzer tests just to make sure all is perfectly OK. Next steps will be an update to the bridge rectifiers (new and higher capacity Onsemi’s + snubber caps), speaker protection board, and new/modern binding posts. Not sure if this amp really needs anything more. I liked what I heard this evening.
cheers,
Luis
cheers,
Luis
And... just 15 minutes ago, poof! it went. Had been working perfectly since late August, on a daily basis. 4 months of excellent sound. A few moments ago, I was listening to some background music, low volume, and....Suddenly, it explodes. One of the fuses on the right power stage is blown. And a burnt smell from the input board (on the right). No visibly damaged components as far as I can tell. Will start diagnosing at the turn of the year. Too many activities this time of the year. Took out my GFA-5500 from storage and is now reporting for duty....
I can't seem to blame any of the new transistors, as they have been working flawlessly for 4 months. Very thorough burn-in. Hopefully none of the output Q's are damaged. But lets see....
Luis
I can't seem to blame any of the new transistors, as they have been working flawlessly for 4 months. Very thorough burn-in. Hopefully none of the output Q's are damaged. But lets see....
Luis