[samsquanch]

So I've obtained a stack of Altec-Lansing 9444A/SA power amps, and I'm currently going through inspecting everything, removing prior modifications (power switch was bypassed because they were on a power sequencing circuit, and the such).  I'm debating replacing the electrolytics, but I'm not sure if I need to.
 
The amps were stamped with approval from a QA person in 1989.  In the power supply there are two 10,000 uF/100V caps, screw type connections, I would test them with my meter, but it tops out at around 10 uF.  And the rest of the caps would have to be removed from the boards to be tested, and at which point, why not just replace them?
 
Thoughts?  Suggestions?
 
And as a secondary question, anyone have experience with these amps, any special tweaks worth noting?

 

[tangent]

It depends a fair bit on the design of the caps themselves.
 
You can get caps, even today, that only claim 1000 hours of life.
 
"But, tangent," I hear you say, "that's only 6 weeks! That makes no sense."
 
The reason electrolytics are specified this way is that such a cap is guaranteed to maintain its specs for 1000 hours at maximum temperature, maximum ripple current, etc.  If you were to run the cap at is max, it wouldn't suddenly stop being a cap after 1000 hours.  It would simply degrade in some way, such as dropping below such a terrible cap's likely +20% -50% value tolerance range.
 
There are two things allowing electrolytics to have useful lifetimes of years.
 
First, designers overdesign.  A new 2200 uF cap might degrade to < 1000 uF and still allow some circuits to work.  The limit of such a benefit usually shows up as an increase in audible line hum.  Do these amps hum?
 
Second, and more importantly, there's a rule of thumb that says that lifetime doubles with every 10°C drop.  A cheap 1000 hour cap is probably rated for only 85°C.  If someone's parked such a cap right next to a heat sink in the amp (stupid, but it has been done) you might only get a few doublings out of this rule.  If instead it's a nice 105°C cap, well away from any heat sources, you might get six or seven doublings, which is even better if it's also a 5000 hour cap. At room temp, such a cap gets 8 doublings, or 256 ₓ 5000 ≅ 146 years!
 
So, find out what you've got, and we can answer the question.

 

[samsquanch]

Since I always ask for a picture when I try to help someone, here's a picture!
 
Obviously, the two I'm talking about are the ones just above the transformer.  The fan blows air from the rear across the heat sinks and out the front, if anyone is curious about air flow.

 
They're rated at 85C, they're Mallory's.  As close to the model number I can find on Mouser is:  http://www.mouser.com/ProductDetail/Cornell-Dubilier/CGS103U100W4C/?qs=7%2fmqHVvJdq0Mnm8pSXgCMw%3d%3d
 
Their document says ~2000 hours at 85C.  These amps were installed in, let's say 1990, used in a college stadium, removed in 2010, so 20 years of use, around 4 hours per game, and (I have a minimal knowledge of college basketball, so a internet searchin I went) about 20 games a year, puts these at a very conservative estimate of 1600 hours of use.  I doubt the racks were in a chilled room, but having not removed these myself it's hard to say what the airflow in the racks was like, so a temperature estimation is hard to make on my part.  These were also probably stacked 8 high, one on top of the other in their rack.
 
The one that I have used, which is now powering speakers in a restaurant didn't have any noticeable hum when I tested it.
 
I'm under the impression that they're fine, but my boss came walking through while I was fiddling with this one, lobbing questions at me, one of which was, "Are you sure the caps aren't dried out?"  Which has gotten me paranoid, but I think that might be the extent of the amps issues, paranoia.

 

[liamstrain]

Regarding the "caps dried out" - there are some easy visual cues to look for on a that type of failing cap. Also - those tended to fail pretty early in their lives.
 
The wiki entry on capacitor plague, covers it well.
http://en.wikipedia.org/wiki/Capacitor_plague
 
Based on the picture you posted, I don't think you have anything to worry about. 
 
Also - holy crap those are big caps.

 

[samsquanch]

Quote:

Regarding the "caps dried out" - there are some easy visual cues to look for on a that type of failing cap. Also - those tended to fail pretty early in their lives.
 
The wiki entry on capacitor plague, covers it well.
http://en.wikipedia.org/wiki/Capacitor_plague
 
Based on the picture you posted, I don't think you have anything to worry about. 
 
Also - holy crap those are big caps.

 
I'm pretty sure that "dried out caps" is just old guy speak for old capacitors, old roadies/techs have a language all their own.  Not always accurate, but it gets the point across.
 
Also, replacement caps of that value are around $25 a piece, so nah thanks to changing those "just for fun."
 
The amp is 300W a side, and 20+ years ago if you wanted that kind of power everything had to be huge.  Now I regularly see Crown/QSC/Lab Gruppen amps that are 400 watts by 8 channels in two rack spaces, yet this 2 channel 300 watt amp is three rack spaces, technology man......
 
Thanks for the help guys

 

[tangent]

Most electrolytics have a wet dielectric, which is why heat matters. Cap drying out = exceeding its rated lifetime. It's just a less precise way of saying the same thing.

 

[nikongod]

Quote:

The amp is 300W a side, and 20+ years ago if you wanted that kind of power everything had to be huge.  Now I regularly see Crown/QSC/Lab Gruppen amps that are 400 watts by 8 channels in two rack spaces, yet this 2 channel 300 watt amp is three rack spaces, technology man......
 

Do you know if the amps are rated for peak or continuous duty at that power level? 
Bridged channels for maximum power? 
Just food for thought as to why some amps only get 300w in a 3ru case and others get 3200w in a 2ru case. 1w is 1w, but 300w is sometimes more than 3200w

 

[samsquanch]

Quote:

Do you know if the amps are rated for peak or continuous duty at that power level? 
Bridged channels for maximum power? 
Just food for thought as to why some amps only get 300w in a 3ru case and others get 3200w in a 2ru case. 1w is 1w, but 300w is sometimes more than 3200w

300W (400W max) @ 4ohms continous
600W (800W max) bridged @ 8ohms continous

 

[kevin gilmore]

Take a very close look at the red air vents on the caps.
If they look even a little dry and cracked, then you have
to replace the caps. For safety reasons if nothing else.
 
Likely the caps have dried out a bit.  If you hook the thing
up to speakers, and short the input, if you hear even the
slightest bit of hum then its time for new caps.

 

[Uncle Erik]

If they're 20 years old, I'd just replace them.  They're going to give out sooner or later and I doubt they're still in spec.  Might as well do some preventative maintenance.

 

[chetlanin]

Quote:

It depends a fair bit on the design of the caps themselves.
 
You can get caps, even today, that only claim 1000 hours of life.
 
"But, tangent," I hear you say, "that's only 6 weeks! That makes no sense."
 
The reason electrolytics are specified this way is that such a cap is guaranteed to maintain its specs for 1000 hours at maximum temperature, maximum ripple current, etc.  If you were to run the cap at is max, it wouldn't suddenly stop being a cap after 1000 hours.  It would simply degrade in some way, such as dropping below such a terrible cap's likely +20% -50% value tolerance range.
 
There are two things allowing electrolytics to have useful lifetimes of years.
 
First, designers overdesign.  A new 2200 uF cap might degrade to < 1000 uF and still allow some circuits to work.  The limit of such a benefit usually shows up as an increase in audible line hum.  Do these amps hum?
 
Second, and more importantly, there's a rule of thumb that says that lifetime doubles with every 10°C drop.  A cheap 1000 hour cap is probably rated for only 85°C.  If someone's parked such a cap right next to a heat sink in the amp (stupid, but it has been done) you might only get a few doublings out of this rule.  If instead it's a nice 105°C cap, well away from any heat sources, you might get six or seven doublings, which is even better if it's also a 5000 hour cap. At room temp, such a cap gets 8 doublings, or 256 ₓ 5000 ≅ 146 years!
 
So, find out what you've got, and we can answer the question.

Great information. I had wondered why a 35 year old amp I have with original caps in the power supply worked so well. Doesn't it also make the Naim recommendation to have their amps serviced just after a few years, mainly for "recapping" appear strange? At least the classical Naim power amp were famously running cold must of the time, because of a very low idle current, and i many cases they sold/are selling separate power supply boxes for "upgrading" purpose; these must be running still cooler. And you say 146 years!  Olaf
 

 

[samsquanch]

I think my current plan is to listen to the amp as is for a while, and then when I get some spare cash that isn't slated for another project (amps, pre-amps, 1989 F150.....) I'll go through and replace all the electrolytics (as you can see in the picture there aren't that many), probably around tax return time.  I guess a cap job in a 20 year old amp isn't that outlandish, especially in comparison to ex-member Parrot's obsession with piccolino wire.
 
Thanks again for all the input!