I would have sent this to Ed directly but the Email function is disabled for users with low post counts, so I apologize for not being able to contribute to the excellent information in this thread.
I recently bought GFA-585 #1293 to drive vintage AR9 speakers being shipped over from the States. I connected it to my vintage AR-3a speakers and immediately noticed a violent thump when the power was switched off (regardless of source shutdown sequence). More than a few web pages imply that such behavior is relatively common among large amps and is usually harmless, so I regrettably opted to continue use. A woofer driver was fried within a week, noticed immediately after powering the amp on (the volume was down, so it was the smell that alerted me to a problem). Before killing the power (i.e.- while following my nose to the woofer), I noticed slight popping noises from the midrange and tweeter on that channel, but I don't know if it was from the amp or the crossover. When I removed the driver, smoke poured out of the cabinet!
Perhaps it's coincidental and the driver or the crossover finally let go, but the amp was the only change introduced to this system in years (I normally drive the AR-3a speakers with a GFA-535), so I'd like to have it checked out. I'm afraid I lack the skills and equipment to take this on, so I'm hoping a reader in the UK can recommend someone over here who can help me!
I recently bought GFA-585 #1293 to drive vintage AR9 speakers being shipped over from the States. I connected it to my vintage AR-3a speakers and immediately noticed a violent thump when the power was switched off (regardless of source shutdown sequence). More than a few web pages imply that such behavior is relatively common among large amps and is usually harmless, so I regrettably opted to continue use. A woofer driver was fried within a week, noticed immediately after powering the amp on (the volume was down, so it was the smell that alerted me to a problem). Before killing the power (i.e.- while following my nose to the woofer), I noticed slight popping noises from the midrange and tweeter on that channel, but I don't know if it was from the amp or the crossover. When I removed the driver, smoke poured out of the cabinet!
Perhaps it's coincidental and the driver or the crossover finally let go, but the amp was the only change introduced to this system in years (I normally drive the AR-3a speakers with a GFA-535), so I'd like to have it checked out. I'm afraid I lack the skills and equipment to take this on, so I'm hoping a reader in the UK can recommend someone over here who can help me!
Oooh bad luck mate.
Actually I'm now in the USA, but no matter. The GFA 585 amp you have likely has the same problems that beset mine. It is a known problem in this product. Electrolytic capacitors on the input pcb are prone to leakage. The electrolyte that leaks from them is conductive, so it upsets the amp circuitry, usualy resulting in a big DC offset (68 volts in my case). This is most likely what killed your woofer.
The fix can be simple - it involves removing the boards, and desoldering the electrolytic caps (might as well replace them all whilst you have the chance). The boards are then washed thorouhgly with washing up liquid & hot water. New capacitors are soldered in when eveything is dry and the amps reassembled. See earlier in this thread for more details... Even if you don't do this yourself, the info might be useful for a technician if you take it to a repair shop. Also do a search, there is quite a bit on this forum about the GFA 585 (Stereo) and GFA 565 (monoblock) amps and this particular fault.
Cheers and good luck,
Ed
Actually I'm now in the USA, but no matter. The GFA 585 amp you have likely has the same problems that beset mine. It is a known problem in this product. Electrolytic capacitors on the input pcb are prone to leakage. The electrolyte that leaks from them is conductive, so it upsets the amp circuitry, usualy resulting in a big DC offset (68 volts in my case). This is most likely what killed your woofer.
The fix can be simple - it involves removing the boards, and desoldering the electrolytic caps (might as well replace them all whilst you have the chance). The boards are then washed thorouhgly with washing up liquid & hot water. New capacitors are soldered in when eveything is dry and the amps reassembled. See earlier in this thread for more details... Even if you don't do this yourself, the info might be useful for a technician if you take it to a repair shop. Also do a search, there is quite a bit on this forum about the GFA 585 (Stereo) and GFA 565 (monoblock) amps and this particular fault.
Cheers and good luck,
Ed
I removed the cover for a visual inspection of the internals (the amp has been unplugged for days, but I'm still afraid to touch anything in there!). I can't tell whether the capacitors soldered flush with the PCBs are leaking, but the 100uf 100V caps soldered at an angle to the top-rear corner of each input PCB are both oozing a dark gray substance. I see no obvious damage to the PCB itself, so perhaps this is something that does not require the services of a specialist and could be done by a local electronics technician? DigiKey lists an overwhelming array of choices for 100uf 100V and 220uF 25V electrolytic caps; what are appropriate part numbers? Will the replacement caps be better than the originals, or is periodic replacement to be expected? Is it likely that this explains the popping/crackling noises, the power down (capacitor discharge) thump and/or the blown woofer driver?
Incidentally, blue wires coming from the power supply and terminating in a two-pin connector are carefully tied off. Any idea what it's for and whether it should be connected to something? I see no obvious two-pin plugs. Yes, I ordered a service manual!
Incidentally, blue wires coming from the power supply and terminating in a two-pin connector are carefully tied off. Any idea what it's for and whether it should be connected to something? I see no obvious two-pin plugs. Yes, I ordered a service manual!
Aha, yes, the 100uF 100V caps (mounted on their sides) seem to have leaked, just as they did in my amp. Things should be OK once they are replaced (seems like the amps were built with a bad batch).
It's a good move on your part to get the service manual - this will help you to pick parts and get the amp properly set up after repairs. Replace all the electrolytics on each input PCB. The biggest part of the job is to note all the connections to the PCBs - There are multi pin connectors and wires soldered directly.
The Blue and white wires are for a cooling fan that was supplied as an optional extra - Just as you describe these were also tied off in my amp.
If you leave the amp turned off for 10 min, this is more than enough time for the power supply to discharge. However, please take care, as the power rails are +/-90V and have huge current capacity. Do not work on the amp yourself unless you have suitable experience (Sorry, but I have to say that)
These are very good amps, and it is definitely worth saving if you can fiind a way to get the work done.
All the best,
Ed
It's a good move on your part to get the service manual - this will help you to pick parts and get the amp properly set up after repairs. Replace all the electrolytics on each input PCB. The biggest part of the job is to note all the connections to the PCBs - There are multi pin connectors and wires soldered directly.
The Blue and white wires are for a cooling fan that was supplied as an optional extra - Just as you describe these were also tied off in my amp.
If you leave the amp turned off for 10 min, this is more than enough time for the power supply to discharge. However, please take care, as the power rails are +/-90V and have huge current capacity. Do not work on the amp yourself unless you have suitable experience (Sorry, but I have to say that
)These are very good amps, and it is definitely worth saving if you can fiind a way to get the work done.
All the best,
Ed
Hi bjmsam,
Your blue wires are for an optional fan circuit. I've never seen one installed, so ignore it.
When washing the PCB, remove the bias controls and label what channel each is from. Remove driver heatsinks and capacitors. May as well remove the components in the area of leakage as well. They will need washing also.
Now for the important part. You must wash the board using an ultrasonic cleaner, twice or three times with a rinse in between. In the cleaner for 180 sec. cleaning. The conductive film is not normally visable, so complete the cleaning ritual. I use a product called "Simple Green". It is an aqueous degreaser, a concentrate I dilute 10:1 (or it foams). That's 10 parts water to 1 part cleaner.
Blow dry that board and let it sit overnight to completely dry out. You don't want water trapped under anything (that's why you pulled the heatsinks and IC's as well). Reassemble the next day, if you smell something odd you must disassemble and repeat the cleaning. Darn stuff is very hard to remove.
New capacitors, you've washed the op amps and other parts in that area with the board. Reinstall the bias controls in the same locations. Your bias current should be at the same level.
Take digital picutes from different angles no matter how good you think your memory is before you take anything apart.
-Chris
Your blue wires are for an optional fan circuit. I've never seen one installed, so ignore it.
When washing the PCB, remove the bias controls and label what channel each is from. Remove driver heatsinks and capacitors. May as well remove the components in the area of leakage as well. They will need washing also.
Now for the important part. You must wash the board using an ultrasonic cleaner, twice or three times with a rinse in between. In the cleaner for 180 sec. cleaning. The conductive film is not normally visable, so complete the cleaning ritual. I use a product called "Simple Green". It is an aqueous degreaser, a concentrate I dilute 10:1 (or it foams). That's 10 parts water to 1 part cleaner.
Blow dry that board and let it sit overnight to completely dry out. You don't want water trapped under anything (that's why you pulled the heatsinks and IC's as well). Reassemble the next day, if you smell something odd you must disassemble and repeat the cleaning. Darn stuff is very hard to remove.
New capacitors, you've washed the op amps and other parts in that area with the board. Reinstall the bias controls in the same locations. Your bias current should be at the same level.
Take digital picutes from different angles no matter how good you think your memory is before you take anything apart.
-Chris
Thanks, Chris. Adcom provided a written description of the repair and I bought a printed service manual from Vintage Audio Manuals, but as neither I nor the local electronics specialist were comfortable carrying out the work, I opted to ship the amp to In House Service Company. I'm hoping this experience doesn't have to be repeated down the road!
Hi Jim,
Don't worry. In my experience the event will not be repeated. The key is a good cleaning job. The person doing the cleaning must understand that the coating may still be there, even if they can't see it and the board is squeaky clean. I found out the hard way.
Am I ever glad this didn't happen under warranty. I would have gone broke doing this under warranty conditions.
-Chris
Don't worry. In my experience the event will not be repeated. The key is a good cleaning job. The person doing the cleaning must understand that the coating may still be there, even if they can't see it and the board is squeaky clean. I found out the hard way.
Am I ever glad this didn't happen under warranty. I would have gone broke doing this under warranty conditions.
-Chris
Well, several months ago the input PCBs were disassembled, cleaned and reassembled with new caps, but the thump persists.
With speakers and preamp connected but with no source selected and zero volume after running the amp for several hours at moderate levels, I measure .006 VDC across the RIGHT channel leads and .876 VDC across the LEFT channel leads. I then used a digital multimeter with computer interface to record the attached log files with nothing connected to the RCA inputs and only the multimeter leads connected to the speaker outputs. All values are VDC.
When the amp is COLD, initial (<30 seconds) offset for both channels quickly climbs to approximately -1.25. After 5 minutes, offsets are RIGHT -1.32 and LEFT -1.37. A few minutes later, RIGHT offset begins dropping and reaches -.28 at 15 minutes while LEFT offset holds steady at -1.36.
When the amp is WARM (operating with no load for two hours), RIGHT offset is -.19 but LEFT offset remains at -1.36. Powering off the amp looks reasonable on the RIGHT channel but results in a severe LEFT channel offset spike (the multimeter is auto ranging but displays "- 0.L" for minus infinity, so I used -30 for the attached graph), which presumably causes the loud thump and likely caused the woofer damage that prompted the initial repair.
Could the culprit be something other than the input PCBs?
With speakers and preamp connected but with no source selected and zero volume after running the amp for several hours at moderate levels, I measure .006 VDC across the RIGHT channel leads and .876 VDC across the LEFT channel leads. I then used a digital multimeter with computer interface to record the attached log files with nothing connected to the RCA inputs and only the multimeter leads connected to the speaker outputs. All values are VDC.
When the amp is COLD, initial (<30 seconds) offset for both channels quickly climbs to approximately -1.25. After 5 minutes, offsets are RIGHT -1.32 and LEFT -1.37. A few minutes later, RIGHT offset begins dropping and reaches -.28 at 15 minutes while LEFT offset holds steady at -1.36.
When the amp is WARM (operating with no load for two hours), RIGHT offset is -.19 but LEFT offset remains at -1.36. Powering off the amp looks reasonable on the RIGHT channel but results in a severe LEFT channel offset spike (the multimeter is auto ranging but displays "- 0.L" for minus infinity, so I used -30 for the attached graph), which presumably causes the loud thump and likely caused the woofer damage that prompted the initial repair.
Could the culprit be something other than the input PCBs?
Hi Jim,
Well, the input PCB isn't clean enough then. Did you also clean or replace the op amp? The PCB is not the only thing that needs cleaning. Even under some of the resistors. That's where the ultrasonic cleaner comes in extremely handy.
It is also entirely possible to have leaky transistors. I normally dismount the transistors on the heat sink and test them as the heat sinks have to come off anyway to clean. There are many things I do without even thinking about it.
-Chris
Well, the input PCB isn't clean enough then. Did you also clean or replace the op amp? The PCB is not the only thing that needs cleaning. Even under some of the resistors. That's where the ultrasonic cleaner comes in extremely handy.
It is also entirely possible to have leaky transistors. I normally dismount the transistors on the heat sink and test them as the heat sinks have to come off anyway to clean. There are many things I do without even thinking about it.
-Chris
I did not measure the offset after it was repaired but there was still a noticeable thump from the left channel when powering down. It remained in use (connected to a Panamax 5100) until recently when the thump seemed as bad as before the repair. The measurements are identical with the RCA inputs shorted. IHSC's reply: "As almost everything was rebuilt on the driver boards a power transistor may have developed a leakage. Very unusual, will have to take a look at the amp again." Back it goes!
'That is bad news bjmsam, sorry to hear that you still have problems.
Cleaning up my amp took two tries. After the first cleanup there was still a little drift in the channel that had suffered the extreme fault. I noticed that it eventually settled, but could be provoked by local humidity increases (breathing on/near the board). The other channel would not respond in this way.
I removed the board again and scrubbed with a toothbrush and washing up liquid as before. I also noticed that the capacitor electrolyte had infused into, and discoloured the protective laquer over the copper tracks on the circuit board - in the same area, some of the solder joints showed signs of corrosion. I scraped laquer away from between the tracks in these areas, scrubbed them with detergent and rinsed thoroughly. This solved the problem, and both channels now settle in approximately the same time after switch-on.
I know you had the work done by a technicial, rather than tackling it yourself, but this info might be worth passing on. The servo circuit seems very sensitive to imbalance caused by (current) leakage paths.
Good luck, this can be fixed.
Ed
Cleaning up my amp took two tries. After the first cleanup there was still a little drift in the channel that had suffered the extreme fault. I noticed that it eventually settled, but could be provoked by local humidity increases (breathing on/near the board). The other channel would not respond in this way.
I removed the board again and scrubbed with a toothbrush and washing up liquid as before. I also noticed that the capacitor electrolyte had infused into, and discoloured the protective laquer over the copper tracks on the circuit board - in the same area, some of the solder joints showed signs of corrosion. I scraped laquer away from between the tracks in these areas, scrubbed them with detergent and rinsed thoroughly. This solved the problem, and both channels now settle in approximately the same time after switch-on.
I know you had the work done by a technicial, rather than tackling it yourself, but this info might be worth passing on. The servo circuit seems very sensitive to imbalance caused by (current) leakage paths.
Good luck, this can be fixed.
Ed
update
I own the world's most stubborn amp!
Email received Wed 14 Nov 2007 02:21:11 PM BST
> Could you describe the rest of your system, including power
> supply/input/line conditioner, etc? Here is why: the Adcom uses
> four 10 ohm resistors as fail safe links in the input ground. three
> of these were cooked, indicating a ground loop or voltage backfeed.
> of course this unbalanced the imput, causing the offset at the
> outputs. Could happen if a preamp with a two wire line cord
> estabilished ground threogh the power amp Rca's.
Email sent Wed 14 Nov 2007 05:08:30 PM BST:
> I bought the amp used and do not know how it was connected by
> the previous owner, who seemed honest and advertised it as being in
> perfect working order. I have used it only in this configuration:
>
> power cord: Panamax 5100 line conditioner
> RCA inputs: B&K Reference 30
> speaker outputs: AR 9
>
> The B&K, the Panamax and the Adcom all have three-prong power plugs,
> but I do not know the quality of the wall socket ground.
Email received Thu 10 Jan 2008 01:04:41 PM BST
> It is actually OK, but I ordered some ultra-low offset opamps to replace
> the ones Adcom supplies.The unit workd fine, but one channel still has
> .65 offset. not all that big an issue, but since I have it here I would
> like to get it as close to perfect as I can. The eta was 1/5, but a lot
> of parts orders are running slow. Will let you know when I'm sending it
> back.
I don't mind waiting for "close to perfect," as my GFA-2535 is doing an admirable job biamping my AR-9s!
I own the world's most stubborn amp!
Email received Wed 14 Nov 2007 02:21:11 PM BST
> Could you describe the rest of your system, including power
> supply/input/line conditioner, etc? Here is why: the Adcom uses
> four 10 ohm resistors as fail safe links in the input ground. three
> of these were cooked, indicating a ground loop or voltage backfeed.
> of course this unbalanced the imput, causing the offset at the
> outputs. Could happen if a preamp with a two wire line cord
> estabilished ground threogh the power amp Rca's.
Email sent Wed 14 Nov 2007 05:08:30 PM BST:
> I bought the amp used and do not know how it was connected by
> the previous owner, who seemed honest and advertised it as being in
> perfect working order. I have used it only in this configuration:
>
> power cord: Panamax 5100 line conditioner
> RCA inputs: B&K Reference 30
> speaker outputs: AR 9
>
> The B&K, the Panamax and the Adcom all have three-prong power plugs,
> but I do not know the quality of the wall socket ground.
Email received Thu 10 Jan 2008 01:04:41 PM BST
> It is actually OK, but I ordered some ultra-low offset opamps to replace
> the ones Adcom supplies.The unit workd fine, but one channel still has
> .65 offset. not all that big an issue, but since I have it here I would
> like to get it as close to perfect as I can. The eta was 1/5, but a lot
> of parts orders are running slow. Will let you know when I'm sending it
> back.
I don't mind waiting for "close to perfect," as my GFA-2535 is doing an admirable job biamping my AR-9s!
Hi Jim,
The op amps used in that unit are low offset types, but here's the rub. They are also low current types. He should have removed them when cleaning the PCB and cleaned them as well off the board. My bet is that the op amps are actually fine.
.65 what? Volts? In that case it isn't fine. .65 mV? Why are we even talking about this if this is the case.
This really sounds as though an incomplete cleaning was done. Did you get it back yet?
-Chris
The op amps used in that unit are low offset types, but here's the rub. They are also low current types. He should have removed them when cleaning the PCB and cleaned them as well off the board. My bet is that the op amps are actually fine.
.65 what? Volts? In that case it isn't fine. .65 mV? Why are we even talking about this if this is the case.
This really sounds as though an incomplete cleaning was done. Did you get it back yet?
-Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.