When to re-cap?

[rastanearian]

I am completing a minor repair on a ML 23.5 and I'm wondering about the condition of the electrolytics. Some say "they are old, replace them".
Does anyone want to comment on the determining factor of cap replacement?
All the caps measure correctly and an ESR meter says they are fine. What to do?
I ask as the amp is a bugger to disassemble and if I'm going to do it, now is the time. It's not going to be cheap though, so I hesitate.

Thanks in advance for any advice.

[sgrossklass]

You have already determined that the caps are likely to be just fine. Of course age is a potential factor, but so are parts quality and sheer luck. Large caps like these can be 40 years old and in fine health, or less than 20 years old and dead.

Can you test how much voltage they still withstand? One of the major failure patterns is small "holes" in the oxide layer developing that cannot take the full rated voltage (+10% or so, as it's usually a "WV" = working voltage rating). Once large currents start flowing there and a carbon bridge forms, the cap is a loss.

Small electrolytics are usually more promising recapping candidates, especially ones with low voltage ratings (sensitive to the above, showing as excessive leakage) or physically small ones (which dry out more easily). If there are any on the power amp supply rails to ensure stability, you may want to replace these with high-quality new ones in any case, as they are so critical. Caps responsible for differential stage supply filtering or similar may also be critical.

[indianajo]

I find on consumer grade electrolytics, 20 years is about it no matter how the hours, and about 8-10 years if itis run 3 hours a night. I had a Cornell dublier 1982 installed cap with about 3 hours on it run great one night this year (when I got the new tubes) then leak and short the fuse the next night. The rubber sealant deteriorates whether you use the cap or not. On the other hand, pro long life caps in industrial equipment are not sealed with rubber, so they can last much longer. They sure will go 15 years in the motor compartment of an oven (very tough duty, these were nichicons).
If you look at new caps, you can buy 1000 hour life, 8000 hour life, or no rating caps. Which were these when these caps were new? That tells you what sealant they used. Caps also have a shelf life rating now, probably having something to do with when the rubber goes anyway without use. (maybe original "forming" also)
I think room temperature capacitance ESR measurements are funny, when the dried up caps always fail when the room temperature is cold (low capacitance) or the equipment is especially hot (short). However, the experts on here highly recommend it. On the other hand, I am listening daily to 40-50 year old consumer grade electronic equipment. (with new caps).

[rastanearian]

Thanks guys,

While I have it apart I will replace the ones that are the most difficult to reach and address the others at a later date. I guess there is no crystal ball when it comes to these things.

[jez]

ESR is really what matters. If you have a proper ESR meter (I designed and made my own and it's been one of the best things I ever did as far as far as fault tracing goes) and it says they are ok, than that is all that matters... so long as they are not dead short which would show good on an ESR meter!

[rastanearian]

I'm using the Blue ESR meter from AnaTek. I have found it to be indispensable.
I'm about to build the Blue Ring meter.

[Conrad Hoffman]

I have never seen a cap that shows good dissipation factor (Rs or ESR, just different ways of saying the same thing) have a problem with capacitance or anything else. My advice is contrary to most- never replace caps because of age. Only replace them for two reasons. 1) They're bad. 2) The unit is filled with the same cap and two or more are bad. That tends to happen with consumer gear like receivers where they stuff the thing with dozens of cheap crappy little caps. Once they start to go, you might as well just get 'em all out of there. If you're dealing with good commercial parts like Sprague epoxy sealed caps, they can go 30-50 years with no problem at all. You'll probably be hard pressed to find a modern cap that's as good and long lived. OTOH, some of the lower grade consumer caps (often well known brands that start with P and M and N and R) are pretty bad right out of the bag. Also, caps with low voltage ratings seem to have poor performance and short life. Any time I see a 6, 6.3 or 10 volt cap, I check it carefully and more often than not, replace it. There's no downside to using a 16 volt or higher cap at very low voltages, if it fits. Claims to the contrary are old wives tales except in the case of high voltage caps, where they should be used at some reasonable percentage of their rating.

[indianajo]

Originally Posted by Conrad Hoffman

My advice is contrary to most- never replace caps because of age. Only replace them for two reasons. 1) They're bad. 2) The unit is filled with the same cap and two or more are bad. That tends to happen with consumer gear like receivers where they stuff the thing with dozens of cheap crappy little caps. Once they start to go, you might as well just get 'em all out of there. If you're dealing with good commercial parts like Sprague epoxy sealed caps, they can go 30-50 years with no problem at all. You'll probably be hard pressed to find a modern cap that's as good and long lived. OTOH, some of the lower grade consumer caps (often well known brands that start with P and M and N and R) are pretty bad right out of the bag.

The name on the cap doesn't matter. If you look at the datasheets, Nichicon and Panasonic both sell 1000 hour caps, and 8000 hours caps. If you read carefully, some of the same part number panasonic caps are 1000 hour in the small sizes and 3000 hours in the bigger sizes. this is on the same datasheet. The motor drives that lasted 15 years and would start on cold mornings (20 deg F in the factory) were Nichicons in TBWoods brand drives. I won't pretend these are the same caps in your hifi amp, but you can buy them if you want to pay the price. As far as brand names go, I had a quad of Sprague Atomlytics in my ST70 go low power in 8 years. Wasn't the ESR I measured, it was the DC voltage on the cap at maximum input signal. I won't pretend an atomlytic was epoxy sealed. The line and datasheet spec, and ultimately sealant matter, not the name. As I said any consumer gear is full of short life caps. Maybe some of your boutique hi end amps are built with better caps, but the nearest boutique retail store to me is probably 300 miles away in Chicago. That stuff never gets here even used on Craigslist. These organs were $4000-$6000 in 1968, built of fine walnut, had rhodium or pallidium plated contacts, and cost more than a convertible SS Malibu.they were full of 20 year caps, Hammonds and Wurlitzers both.
I have kick started a number of motor drives and weigher power supplies on cold Sunday midnights (typically 20-30 deg F in the factory in January) with a Master Specialties heat gun. That doesn't mean the caps were good, it just means they were good enough at 70 deg F. So again, I am suspicious of 70 deg ESR measurement. I've got a semi-professional 1998 build Peavey CS800s that worked great when I bought it (for 1/5 of list price) but blows the breaker when it is 55 degrees F in here. (heat is too expensive a hobby for retirement). It is not *****y consumer gear, but I'm not going to open it up until I have a 2000 hour or better replacement cap for every electrolytic in it.
Do what you want. I haven't touched my H182 organ since I put the 70 caps in it 2 years ago- not even when it is 55 deg in here. The H182 has about 30 db more volume too, than it had in 2009 on the same tubes.

[Conrad Hoffman]

A good rule is, "test it the way you use it." Fortunately my hif-fi stuff is at a comfortable room temperature most of the time. All manner of bad things happen in the cold, so if cold is important you need to spec and test for cold. That might explain why my drill press blows start caps in the winter. Long ago there was a little Mallory booklet, Birds, Bees and Capacitors, that was used as an intro for factory workers. It describes all the contamination hazards that shorten the life of electrolytics. It's no surprise that cheaper caps use lower grade materials and may not last very long. OTOH, there's no way to know how long a given cap will last, even if it was low grade, thus my feeling that if the dissipation and capacitance are OK, and if similar neighbors are OK, let it be. Once a cap is out of the bag and installed, I don't usually have any way to tell precisely what series it was. Good caps look a lot like crummy caps. Sprague published a lot of reliability data in their app notes, but those are tough to find as well. I think I've got some Atomlytics and they were sort of middle of the road. Environment plays a big factor. I've seen too many caps installed right next to hot power resistors and those never last very long. A good shot of chlorinated contact cleaner risks getting a bit of chloride through the seal and into the cap, and that always leads to rapid failure. Alas, deciding what to do involves predicting the future, and nobody's going to be perfect at that.