Even inexpensive solid state amps and preamps seem to last essentially forever.
94 responses Add your response
Hopefully one of the A'gon techies will chime in, but FWIW, here's my thought. SS does age. I base this view on the "care and feeding" I occasionally do with my old Crown DC300A. For those old-heads out there, you know this amp - it's immortal and bomb proof. Nevertheless, the last time I sent it to the Crown factory for a "grease and lube," they replaced some of the old electrlytics, believe or not, the power supply caps and a couple of transistors. The stats are incredible, thank G-d for NF.
I am still (as a backup) using some 25+ year old SS amps and they still work fine other a higher noise floor due to aging caps in the pre/amp sections. While I don't think this is unusual, I would say the most amps of either SS/tube nature are pretty reliable. Other than tubes themselves, the main components that age are the capacitors. Strangely two factors can effect their longevity, one is heat-which would logically take a bigger toll in tube amps or class A biased higher power amps. The other killer is lack of use as the caps will dry out. Many people advocate using a Variac to bring power back to an amp that has been idle for an extended time to bring the caps back up to life slowly as a sudden rush of current could blow them. I've also noted that amps with several smaller output caps (as opposed to fewer but larger caps) are less susceptible to aging -don't know why...
Spinaker01 - Drying of the electrolyte increases ESR of capacitor. When it gets into many ohm range it becomes very audible and after that it might even go to thermal runaway since current going thru cap x ESR = Heat but temperature increases ESR (positive feedback). Smaller caps are connected in parallel making ESR lower to start with - less audible when old.
Each 10degC of temperature increase cuts life of capacitor in half.
Electrolyte inside of a cap eats out aluminum oxide (dielectric) lowering breakdown voltage. Presence of voltage rebuilds this layer.
The one thing that tends to "wear out" with solid-state amps is the capacitors ("caps"). In a decent amp, they tend to last many thousands of hours, therefore providing years of use, but they do fail over time.
Provided that an amp does not run really hot, like a pure Class-A design (e.g., Plinius, Gryphon, and certain Pass amps), the best way to increase longevity is to keep it powered up 24/7. The thermal cycles that a component experiences when it is turned on, whereby it heats up, and then turned off, whereby it cools down, is harmful to the internal componentry, making parts brittle over time and inducing failure. 24/7 operation does wear out caps faster, but caps are relatively easy and cheap to replace - they can be viewed like tires on a car (albeit a set of tires that gives you 300,000+ miles - caps last many years). What you don't want to have happen is output transistor failure, because they have to be matched and occasionally even go out of production - this is the most compelling reason to keep the amp on 24/7.
That said, you do not want to keep a pure Class-A biased amp, or other designs that run really hot, turned on 24/7, because high heat also degrades parts and thus can induce failure (not to mention the large amounts of electricity that Class A-biased amps consume).
Likewise, it's arguably okay to turn on and off a cool-running Class D amp, as such amps really don't experience thermal swings when turned on and off - they're always cool.
To specifically answer your question, a decent quality solid-state amp that has not been turned on and off with each listening session, and that has its caps replaced every ten or twenty years, should never fail. Of course, this assumes that the owner takes necessary precautions in the event of weather events and other hazards - equipment must be unplugged from the wall in the event of electrical storms (lightening strikes can destroy audio equipment), moved to high ground if there is a flood, protected from children, pets, house cleaners, and drunks, etc.
Between my buddy and I we have 3 SS amps still going strong after a combined 75+ years. All of them are 200WPC and up. Adcom, Kenwood, and Carver. The only issue is noisy volume pots on the Kenwood but I leave it at full volume all the time anyway so it is not an issue. Even the meters still work on the Kenwood and the Carver.
Kijanki, got a question. As I mentioned above, Crown rehabed my DC300A about 6 months to a year ago. I keep it around as a back up amp, but haven't turned it on since Crown checked it over. Per your post about caps drying out, can I safely plug the amp in without an input signal or a load just to get some juice running through it?? Will that keep the caps healthy?
You will get opinions from electrolytics last forever to they need to be replaced every 10 years. The truth is they do wear out but the the variables are too great to make an accurate prediction of when. The 10 year people may be overly cautious, the 40+ year people are ignoring the facts, the truth is probably somewhere in the middle.
"Any" gear last longer if used? It may last much longer if used occasionally, especially if it gets hot or has moving parts or has tubes.
Electrolytics form a thin layer of oxide on the aluminum and this layer gets thinner the longer they sit around with no voltage. If they sit too long and this layer gets too thin they can breakdown and be destroyed if you suddenly apply full voltage. That is why it is recommended to bring the voltage up slowly with a variac if you are turning something on that has been sitting around for a long time. This is called reforming the cap.
Too many variables,.... room size, listening level, speaker efficency, speaker efficency, etc. Bryston is at least 20 yrs. Otbers vary. Pass is very good, bullet proof in fact, but with all ss equipment, one transistor failure, replacement is not cheap, sometimes more than used price of the piece. A friend had the krell mda 300 fail...try 2800.00 on for size. That is why I go for tubes, tubes are cheap. Sometimes though, only ss can do the job. Pick your poison....jallen
McIntosh for example seem to last forever. Most people feel an amp that are 10 years old are getting ready to retire but ss McIntosh amps from the 1960's like the MC-2100 still show up on this site regularly. That's always been a curiosity to me, is the McIntosh infinitely better than my Krell KST 100. I've recapped it once but are the transistors and other components that much better that they can survive decades longer? Is it the autoformer that makes them last. I would like an amp that lasts and I'm willing to trade off sound quality to achieve that. Some people are critical of the McIntosh sound but no one says they don't last forever.
Bifwynne, No input signal is OK. Tube amps need load, SS amps don't.
SS amp might fail when input signal is present and output shorted (overcurrent). I've heard that some poorly designed class D amps might go into oscillation without load but likely will survive that.
Tube amp might fail when input is present and output is open (overvoltage). It might even happen without input signal if amp goes into oscillation.
I increase volume slowly, at first, making sure I hear the sound from both channels. It is good habit, IMHO, since things tend to get shorted or disconnected between listening sessions.
My personal experience of owning SS gear for many years is that leaving gear on 24/7 is silly, wasteful, completely unnecessary, tawdry, untoward, and pretty much unjustified by any meaningful data ...the wasted juice far exceeds the cost of any "premature ageing" of a component due to turning it on and off.
Filter capacitors have a half life of about 20 years. What that means is that in a 20 year period, about half of the caps will have failed. This is true of transistor and tube amps.
Sometimes when the cap fails the ESR goes up like Kijanki was talking about, you can also have the experience of the cap slowly loosing its capacitance. This may not result in the hum that is commonly associated with filter cap failure, but it will impose sonic degradation as the power supply is no longer properly bypassed. This can increase distortion and affect bandwidth, things that you may not notice over time but would certainly show up on the bench.
So in general, by the time a piece is 15 years old, there will be some degradation or outright failure, by the time 20 years has passed the chances of that are 50/50. This will be true even if the unit has been in continuous use, although such pieces in general do have the caps last longer, that's not the same as saying they actually still meet spec!
Ralph, can you back that up with some references?
Curious where you got 1/2 fail in 20 years.
Even if true, given that the lifetime of a particular cap is inversely proportional to operating voltage and temperature I fail to see how this is useful for predicting the lifetime in any particular piece of equipment.
BTW that is not the definition of half life.
Marakanetz. capacitance could be OK but ESR and/or leakage too high. It is doubtful a 50 year old electrolytic is still in spec.
Rrog, half life is the time it takes for something to decay or decline by 50%, it is not the time it takes for 50% of a group of objects to fail. For instance; the half life of a discharging capacitor is the length of time it takes for the charge to drop 50%. The half life of an ingested drug is the time it takes the body to eliminate half of it.
I'm dumbfounded you would continue to defend what is clearly just plain wrong. You are evidently just making stuff up as you go along this time.
That is absolutely not the correct use of the term. It means something has half it's life left, not that half of a group are dead. Find one other example where it is used as you used it.
How could you possibly have seen it many times? Give me one example where you have seen a group of caps where 1/2 have failed after 20 years.
You'll have to do better than they teach it in technical school to have any credence. I taught in a technical school for ten years and I've never heard that 1/2 after 20 years statistic before. If it is correct you should be able to easily produce at least one credible source, but of course you can't because there are too many variables to have a statement like that be true.
Plutonium is a good idea though, it has a half life of 80 million years ;>)
In 1981; I bought a Hafler DH500(kit), to power the woofers of my biamped system(I was a Hafler dealer, at the time). Gave it to my son, 15 years ago, when I bought a Hafler Trans*nova 9505, to replace it(same purpose). He recently sold it to another friend of ours, and it still functions perfectly. Until now; it NEVER had an easy life!
Many new electrolytic caps are + or - 10-20% and they are usually further out of spec after 10-15 years. They may still work, but they are cheap, why not just replace them. I like Spragues or F&T. If they will fit, I use v-Caps in coupling or input cap locations. They are soincally just superb. I have had a few Illinois electrolyics leak, the brown goo, and replaced those. But, they were 15 years old or more. Jallen
> Filter capacitors have a half life of about 20 years.
> What that means is that in a 20 year period, about
> half of the caps will have failed. This is true of
> transistor and tube amps.
Ralph, I would call this MTBF, or mean time between failures.
"A vendor-quoted MTBF implies that half the drives in a large population will fail within that time of operation."
If in technical school they used half life, I would suggest they either got the term wrong or the definition has now evolved into its current usage as "the period of time it takes for the amount of a substance undergoing decay to decrease by half."
Here is another concern. My company has been dealing with this for several years now. The caps in the computers that run our equipment are failing at a very high rate.
I agree that MTBF is not applicable when talking about caps. It applies to complex equipment that needs to be repaired and put back into service. The time it takes a cap to fail is simply lifetime. Failing caps may lead to a lower MTBF for whatever they are installed in though.
English is an evolving language.
'Half life' works really well in this case since you have the statistical issues of a single cap and also that of several.
In the case of a single cap its half life is 20 years on average. IOW, you will find that it may well have dropped to half its capacity, with its ESR likely more than doubled. I have never seen caps over 40 years old that are safe to operate; but by that time they should be replaced, without question.
In the case of a bank of filter caps it will be found that in 20 years time half of them have reached the end of their useful service life, by leakage, by shorting, or simply so ineffective that the equipment using it no longer meets spec.
It seems that 'half life' works very well for this. However if sensibilities are offended (we're not in Alice's Wonderland, apparently) then MTBF is fine.
Herman, A few years ago I heard about a Chinese firm doing some corporate espionage. Apparently Panasonic had developed a more effective dielectric for their electrolytic caps. Fearing the discovery of their formula by competitors, they kept it in 2 halves in different locations. How I heard it was that the Chinese company had managed to get one of those halves and used it in their caps, which got used in a lot of computer boards that began failing about 6-8 years ago on this account. Is this anything like what your firm was experiencing?
Again, I'm dumbfounded you continue to defend what is clearly wrong.
I've been dealing with electronics for over 30 years, taught it for 10, and have looked at this topic quite a bit since it often comes up on this and other forums. I've never seen it mentioned that 1/2 fail in 20 years or any other statistic that puts a lifetime like that on them. Any data always factors in temperature and voltage. I really respect most of what you say here but you clearly have no real data to back up any of these 20 year claims.
Yes, English evolves but that doesn't mean it is acceptable to use it in any way you wish. Half life has a very clear, concise, scientific definition and your use of it is well outside that definition. Ir would be nice if someone on these forums would just once admit they are wrong.
MTBF to describe cap failure is also incorrect as stated above. It refers to complex, repairable devices, not components. How can you possibly have a "mean time BETWEEN failures" for a device that is thrown away when it fails? There is never a second failure so there can never be a between. From this article
MTBF is properly used only for components that can be repaired and returned to service.
Yes, I suspect the caps we are seeing fail are part of what you talked about. I posted a link to it in my response above.
Good question. I don't think there is really much required in the way of care and feeding. I am on more than 20 years on my PS Audio 250 Deltas, and they still sound fantastic.
That said, there are problems that crop up with amplifiers, tubed as well as solid-state, lord knows why.
Not sure that there is really a defined lifespan, but for the prices we pay for some of this stuff, it should be forEVER, right?
True story: I had a 1960 Fender Deluxe guitar amp (brown tolex vibrato version...no reverb) that I thought needed attention (I'd owned and used it as a studio amp for 35 years or so) before selling it (mistake). It had old leakey caps...a reliable tech swapped 'em out for nice new ones resulting in my amp sounding exactly the same.
A couple of things. 1) most decent amplifiers have soft/slow start circuitry which prevents the on/off cycling problems that people mention here. Therefore for a decent high end amp, the cycling on/off damage they are talking about really isn't an issue anymore. 2). Electrolytic capacitors would be a problem and must be checked. They are like any other component in cars, etc. they give you advanced warning that they may probably fail. yes, yes, there are instances where something just simply goes out, but in most parts you can see leakage before they fail. So, take a good look inside or have a qualified technician inspect the unit to see if any problems may exist. If nothing is apparant, then leave it alone until some problem or possibility of problem presents itself. 3) apart from capacitors, the thermal damage due to the loss of heat sink compound that the transistors are using when connecting to heat sinks may be the major issue. Transistors fail for a number of reasons. a) people that short the outputs when playing with speaker cables are the number one reason. b) when the heat sink compound dries out, it presents a problem for loss of thermal conductivitity and also allows for the possibility of the transistor to come in contact with the heat sink, because not only is the heat sink compound gone, but the thermal insulator between the transistor and the heat sink may be damaged. Therefore, what I typically do with older amplifiers is to check the capacitors for leakage and replace if necessary (only if necessary) these suckers are expensive, and replace all pre-driver and output driver power transistors with the newest, latest and greatest more linear transistors and readjust the bias, which makes the amp much better sounding. This also allows me to replace the thermal insulators and the heat sink compound, thereby eliminating any possibility of the amp having issues. my "modified" amps are much better sounding then the original and also are much longer lasting because I have inspected and replaced capacitors where necessary and also I have replaced all pre-driver and output power transistors and their associated thermal insulators and heat sink thermal compound.
One last thing. I typically will add a slow start/turn on relay circuit to any amp that does not have it originally and I sometimes will modify the power supply to separate power supplies with regulators.
After all that, I guess what I am saying is for really old amps, definitely change the heat sink compound and thermal insulators at a minimum and if you can change all output transistors also and check and replace if necessary the electrolytic capacitors.
Minor, sounds like you are a very thorough tech.
Question: you say you check the caps for leakage and replace only as needed. What about a loss of capacitance and/or increased ESR? The amp may function with these problems but well below it's potential.
Also curious what method you use to measure the leakage.
Hi Herman, I too have taught in technical schools and my stint in the service industry goes back to 1974. That's how I financed my way through the Minnesota Institute of Technology. The use of 'half life' is my own and simply comes out of experience - as a service tech I have repaired thousands of amps, preamps, receivers, tuners, tape machines, etc. I'm sorry that the use of the term bothers you but it works so well and is so easy to explain that I will continue to use it. I'll try not to use it around you :)