As an example: my Son of Ampzilla (circa 1977) is now going on its third year of daily use! Working perfectly well and sounding mighty fine! I bought it used with an unknown number of hours from the previous owner. Took a chance and plugged it straight into the AC line (would have been safer to use the VARIAC!). Worked fine, no problems!
Capacitors age over time and thermals play the biggest part. Can seal leakage is another big failure point and the next aging aspect is the level of current that is modulated through it, over time, with regard to peak long term transient levels, or sine levels (sine wave...in-out, at high levels, over time)
Leaving an amplifier on is probably the fastest way to age the capacitors, as the thermal levels are generally much higher if the amplifier is on, compared to being off and in cool dry storage. Each small plastic skinned aluminum can has the materials inside and then two leads coming out. What the end cap of the can is, is a shaped/molded rubber bung. The bung must seal the outer rim of the can, and then seal the capacitor metal connection legs.
The capacitors are filled with layers of thin metal sheet, coiled along with dielectric separators. The capacitor is infused with a fluid or slurry that is called the electrolyte. This electrolyte, when heated, can have an out-gassing issue tied to the seal quality, and general overall micro leakage, over time. The thermal and natural long term leakage or drying out of the fluids in the electrolyte..slowly ages the capacitor until it finally reaches a failure point.
The failure point is impossible to avoid and WILL happen like gravity and the sun coming up. This failure point is not guaranteed to happen all at the same times on identical capacitors due to the mentioned point that all capacitors are stressed differently and not all seals on all capacitors are created equal. The biggest metal can capacitors in most audio (power supply) gear are generally, if utilized correctly (not electrically over-stressed or thermally stressed) ..they are generally the least likely to fail, and the longest lived.
The failure mode of an electrolytic capacitor that is drying out..is to have no remaining capacitance value, it’s like running a gasoline engine dry with no oil. The end result, is messy for the engine at best. The next thing they do, is to internally short with internal arcs, and the residual electrolytic fluid is converted to gas and can ignite. The infamous capacitor *bang!* we hear about and sometimes get to witness. Much amplifier circuitry or general circuitry carnage can ensue in this scenario.
In general terms, most gear over 10 years old should have a general capacitor quality and wear inspection.
Anything that is Class D or has a pulse power supply should likely be inspected at least at the 5 year age point and then maybe every two years or earlier, after that first 5 year period. The stressing that their capacitors are under is much higher than a standard fully analog system that is built with transformers and rectifiers.
I had to explain all of that so I could say that the continually running GAS amp mentioned in this thread, is quite seriously in the hardcore red zone of imminent failure modes.
If the ampzilla has never been recapped since 1977, as a technical type, I would never even turn it on without a full inspection and some critical level recapping, before taking it even further in the recapping realm.
Please do not take to heart such advice about leaving the amplifier on, ok?
One does also..not want to thermally over stress any plastic aspects in the given design which might be exposed to heat..., as heat over time is what stresses plastic components that may be in use. These plastic bits which can be in the given gear will discolor, harden and become embrittled, fragile..as time goes by. Generally, this is stuff that is (or can be) a problem in the 20 year age range and greater.
Recapping is most important with high voltage devices like power amps. If they go, they can take out a lot of the amp with them. Especially important if you have expensive tubes, or irreplaceable transistors. 30 years is a good threshold.
Line level devices like your 901 EQ probably won't suffer catastrophic failure if a cap blows, but the electrolytic caps today are much better than their 1980's counterparts. It may be worthwhile, and relatively inexpensive.
Electrolytic caps kept fully charged will function perfectly well and not degrade!Nope. There are multiple mechanism for electrolytic capacitors to degrade. We once replaced 7 million of them.
The only way to properly test caps is to pull them all out and test with a quality tester. A cap test on a VOM is probably not adequate.
ESR increases as electrolytics age and increased ESR is never a good thing.
Electrolytic caps are rated for 1000 to 10000 hours @ maximum conditions. If operated at less than maximums, life can be doubled for every 10°C below max °C. Assuming 105°C caps and a 35°C internal temperature, that's between 125k and 1.2 million hours. Only a small percentage of the caps would be expected to have the upper limit. 1977 was more than a million hours ago, so a 40 year old amp is living on borrowed time.
If the ampzilla has never been recapped since 1977, as a technical type, I would never even turn it on without a full inspection and some critical level recappingDitto! 1977 is 40+ years and that's nearly 3 times as long as the recommended de-rated 15 years due to seal deterioration.
All capacitors dry out after awhile 15-20 years for sure
Electrolytic power caps for sure . I owned a store and worked modding
caps , there are several grades of quality ,coupling caps,of various types like teflon, paper in oil.foils, as well as Electrolytic , including the top 4 pole cap types. You get what you pay for on average .
Capacitor goes with loud bang because of avalanche effect. Capacitor is charged in narrowed current spikes of very high amplitude. That creates losses and heat with ESR of capacitor. When capacitor dries out ESR increases causing bigger losses and bigger temperature increase. Since temperature also increases ESR at one point it will lead to explosion. Because of that all electrolytic capacitors have some form of fuse. I agree with gs5556 - you will hear loss of bass definition and 120Hz buzzing way ahead of that.
or the past 6 years i only use my equipment like 6 hours a week. should i worry about the 'caps'??Not for at least 20 years.
Just relating my experience. I purchased a pair of Polk SDA-SRS speakers from the 1980's just a few years ago (4 tweaters, 4 mid and 4 SDA drivers and a large passive radiator). I have always enjoyed the tremendous sound stage of this technology since I first heard a pair in a hi-fi shop shortly after they were released. There is a large SDA following on the Polk forum. One of the outcomes of this group is that some of the technically talented have put together several items to 'upgrade' these rather old speakers. The most significant upgrade is to recap the cross-over, and there are schematics for each of about a dozen SDA speaker models that were built during this time. When I got them setup I thought they sounded great for a relatively old set of speakers. Good sonic attributes, but a soundstage with more width, depth and vertical than I have ever heard. But I decided to rebuild the crossovers using the high-end caps as suggested by the forum members. When I finished and fired them up for the first time, it was like someone took a very thick audio absorbing blanket off from over the speakers. In spite of what I appreciated as great sounding speakers, and I auditioned some $10-$15k speakers prior to purchasing these, it is hard to describe the improvement of every sonic aspect with the new and upgraded caps. My point is not the SDA speakers, but the sonic improvement due to the upgraded caps. I have not done anything similar with electronics, i.e. amps. But if it is anything like I experienced with speakers, it is well worth the investment.
It depends what do you mean by "much better". Do you mean lower inductance, like slit foil caps, that are very expensive, or lower ESR caps that might cause rectifier damage? When you recap 30 years later original capacitors don't exist anymore, but I would check their specifications and look for something similar and not the "best" available.
EG, a electrolytic with a much higher frequency response, or lower impedance into higher frequencies...might make an amplifier oscillate out of control, due to the feedback and loop gain being set too close to the limit and part of said limits unintentionally relying on the original capacitor’s limits. It can be done. It happens more often than one thinks.
Willy nilly cap replacement by people without the understanding of the circuit, can result in circuit failure. Or over-stressing that can kill the circuit far sooner than the original design and build intended.
Let’s try a hopefully humorous analogy.
Most solid state audio circuits turn on, when that power switch is hit...like they were punched in the head with a ludicrously high speed truck. Hit the power switch and everything in the box goes *BANG!* (Ok, we’re all awake now, pedaling our little bikes!) Somehow they survive it. Robust little things, generally speaking.
If you get in there and make capacitor value and capacitor quality changes, it’s like adding in bigger and faster truck change-outs... to the circuits which were originally being punched in the head by big and fast trucks. This is generally not a good idea.
Somehow this didn't get said, so I'll describe the modes of failure.
Caps fail in a variety of ways. Their capacitance diminishes or rises. Their equivalent series resistance rises, and they leak DC, otherwise called "leaky" or "electrically leaky". Often a failing cap will exhibit some combination of these symptoms. It's not at all uncommon for a cap to measure higher capacitance, but that usually means it's leaking. Leaking interstage coupling caps will slowly destroy your amp so just because you're not hearing some obvious sign of trouble does not mean there isn't a problem. High ESR is just like putting little resistors in series with the cap and the problems with that are hopefully obvious.
Heat destroys electrolytic caps. That's just a fact. The more heat applied and the longer it's applied has a direct relationship with the life of electrolytic caps. Anyone who says otherwise is a liar.
The Truth is available from the manufacturer.
Sometimes the Truth is optimistic and sometimes not.
Life is predicted based on science and math, but can vary widely due to variation in sub contractor supplied materials.
We once replaced 7 million 85°C/35V electrolytic caps with 105°C/50V version from another manufacturer. All was well for a while until we found that the replacement would fail if the bung was over the -18v rail. The potential sucked out the electrolyte and started fires once the electrolyte reached the positive rail.
Electrolyte is simply drying out. Each 10degC temperature increase shortens the life of capacitor by factor of two. Capacitor rated 60,000 hours at 25degC wil only last 30,000 hours at 35degC, 15,000 hours at 45degC etc. This can be found on many datasheets.
Lets be clear. There was lots of good advice presented here. But, that is all it is, advice.
There is such a thing as preventive maintenance in cars, electronics, etc. If one wishes to keep an item indefinitely, then preventive maintenance is important.
If upon visual inspection the capacitors are bulging or obviously leaking material, then it is past time to replace.
If the amp is over 15-20 years old, but the caps appear okay and the amp is operating well, then, it is up to you. I wouldn't replace the caps unless I notice issues or if I had the funds to replace the caps in advance of issues.
Like cars, you are warned most times of issues before they become really serious. Therefore giving you warning. whining bearings, leaks from the water pump. All indications that you need to do something.
If the amp works, the capacitors are in good shape and not leaking, then you don't have to "do" anything but enjoy. But if you have the budget, it can't hurt. However, for solid state electronics, it is also a very good idea to remove all output transistors and test them and replace if they test outside specs, and very importantly, replace the mica insulators with new and also the silicon heat compound. This dries over time and causes output transistors to short against the heat sink and fail the amp.
I do this on general principle whenever I get older equipment.
Some amps are easy to work on and other's (Mark Levinson 23.5) are a serious pain in the bottom. Great amp, but OMG what were they thinking?
I am restoring my highly Modified Old Sumo Nine right now, It took a lightening strike a few years ago and I have missed it. In 1983, we put what at the time was a very nice Panasonic electrolytic in the power supply, they were an 85 degree part. It was never turned off for most of those years. While repairing this amp, I figured why not, while its open, I’ll replace the caps... I found nice caps with double the capacitance and lower esr to replace them.... The point is, To my surprise, we measured the old Panasonic caps. their values across the board were still in spec. You just don’t know. My experience has been, that when caps sit without a charge, that is when the tend to dry out. I have never had a cap failure and I do not EVER turn off my solid state gear. (when I owned tubed gear, it was always turned off after each use). I have always been under the belief that it is the charging and discharging caps that is most damaging to them.
You've clearly never owned a class A solid state amp. Your anecdotal "evidence" makes no sense. Your use of commas makes no sense either. Unless you measured the leakage, capacitance, and ESR, you don't know anything about their condition. I just had a power supply fail a couple days ago because some idiot stuffed a big, cheap cap right next to a hot diode. Made the thing chirp like a bird.
@kosst_amojan Very nice of you to put it that way to me... So, I've clearly never owned the PURE Class A amplifier that I just referenced that I am currently rebuilding? I am changing the 85 degree caps to 105 degree caps, but I haven't shoved any big cheap caps next to hot diodes. I hope that can be repaired for you. I guess since it is "My Experience" you could call it anecdotal, but either way, it is 100 percent true. Is it ok to use to many periods...... or is it just commas that are the problem? And yes, ESR and Capacitance were both measured. I saw no reason to measure leakage current since there was no loss of ESR or Capacitance, all caps looked great.....
I am ok with others having more knowledge about amplifiers than I do. I am a speaker guy, but I still do not belittle people about what they might not know about speakers.
@g_nakamoto... I apologize, I did not intend to have your thread go awry.
@kosst_amojan Leaving an amp on all the time sounds better, period.
Again, I didn't turn it off for years upon years and caps measured at spec. As, I'm sure that you know, caps with lower ESR should sound better, even thought spec'd at 105 degree parts vs 85 degree parts, ESR was significantly lower.
Yes, it gets expensive and HOT in the summer. I have a Class D amp that I made with Abletec modules that actually sounds quite good. I ran on occasion and am using it now until I finish the old Sumo Nine, but even while owning the Class D, I kept the Sumo Plugged in.
The F5 is a nice sounding amp. One of my best friends used it with Tektons and is now using a DIY version of the Aleph J. I wish that it was a bit more power, I'd give it a try. I could use it with one set of my speakers, but it would be just ok with my other 2 sets. I have a 12 inch 2 way with Heil AMT that is 4 ohm and 95db sensitivity, I have an MTM which is 4 ohm 88db sensitivity and I have a 10 inch 3 way that are 8 ohm 86db sensitivity. My Sumo originally had 4, 10,000 mfd caps, it is now getting 8, 22,000 mfd. With the beefed up power supply and a bit of re biasing now does around 80 to 8 and 160 to 4.
So to this end of "knowing the circuit b4 recapping" .. I know the Counterpoint NPS 400 has been talked around the board a bit .. anyone know a tech who is proficient in Counterpoint? Mine will give a mild pop in one channel when powering up, no crackling or any other symptoms that I detect during play .. but am assuming caps are the issue? Built in 1996 btw.
the problem would be with the electrolytic capacitors.
Therefore, I would plan on replacing all the electrolytic capacitors with like value and size capacitors. However, the number 1 priority would be the power supply caps. They are typically the easiest to get to and replace ("typically"). The problem is finding the correct value and physical sizes that will fit within the allocated space and clamps (if used) available.
There are a few manufactures out there and a few retailers that will sell the large capacitors. If you are replacing, then you really want new caps. You can find NOS unused caps, but they would be just as old as the ones you are replacing. However, they will not have been subject to the extreme heat from use. They may (read should) be connected to a variac input first to reform them slowly. But first try for new.
Also, by-the-way, you definitely do need to look at the other electrolytic caps to see if they are demonstrating problems such as leaking also.
If, after inspection of the circuit board caps and the power supply caps there is no obvious issues, then take your time in planning the cap replacement. no hurry here.
@dweller A recent inquiry into the age of an amp, and my concern for the caps, resulted a response "we've tested the caps and they're all good". Is it possible to evaluate the condition of capacitors (other than working or non working)?Because a capacitor must be removed from the circuit board (at least one end must be de-soldered and lifted) I view such claims as highly suspect. Notice that they never specifically state what tests were performed. Also, realize that an old worn capacitor may test the same as a new high-end capacitor if one is testing for capacitance value only, but the newer capacitor may sound much better.
Everything looks good and the amp sounds good but it's a 1987 and I don't know the history on it. I have a Chinese ESR meter and a regular capacitance meter and so far every cap I test has shown good. The amp has 4 36000uf 100volt caps that are too large to test on the meters I have. Is there any way that they can be tested easily?
Absolutely replace your big caps. They are the one's that will do the most damage. You will also see smaller caps, in the range of 100uF to 1,000 uF electrolytics on each audio board. Those are also good candidates. Make sure to pay attention to the uF (same or slightly larger), V(same or larger) and temperature (higher = longer life) as wel.