How long can the "right" PSU last?

[TRIPTEX_CAN]

I recently rebuilt my system and this makes the 5th system I've built around the same Corsair TX750w PSU. This is probably the first generation of TX750. It predates the Gold, Silver, and Bronze rating system that would come in to play years later.

I originally bough this PSU sometime in 2007/8 (?) and its powered and overclocked a Core2 system, two different Core2Q, one i5 2500k, and my current R7 2700x. During that time it's ran an HD 3870, 2x 4850, hd 5870, hd 5970, hd 7950, gtx 1060, and now my gtx 1080.

From all I can tell its still very stable and consistent. I was trying to use HWinfo to report the voltages under load but It doesn't support my mobo. I guess I'll load up the Asus Ai suite 3 later today and see if that works.

I never have system shitdowns under load unless I'm testing a bad OC setting. Maybe my OC efforts would go further under a new PSU?

Just curious if anyone else has had this sort of lifespan from a PSU

 

[Vayra86]

I think a reasonable time to expect out of a quality (high tier) PSU that is pushed over the years is 10 years, but caps do degrade over time, so if you push it hard later in its life, it may not do too well at peak loads. That is where headroom comes in.

 

[sneekypeet]

I recently rebuilt my system and this makes the 5th system I've built around the same Corsair TX750w PSU. This is probably the first generation of TX750. It predates the Gold, Silver, and Bronze rating system that would come in to play years later.

I originally bough this PSU sometime in 2007/8 (?) and its powered and overclocked a Core2 system, two different Core2Q, one i5 2500k, and my current R7 2700x. During that time it's ran an HD 3870, 2x 4850, hd 5870, hd 5970, hd 7950, gtx 1060, and now my gtx 1080.

From all I can tell its still very stable and consistent. I was trying to use HWinfo to report the voltages under load but It doesn't support my mobo. I guess I'll load up the Asus Ai suite 3 later today and see if that works.

I never have system shitdowns under load unless I'm testing a bad OC setting. Maybe my OC efforts would go further under a new PSU?

Just curious if anyone else has had this sort of lifespan from a PSU

A good indicator of what to expect is the warranty period. While many products can outlast the warranty, the manufacturers will set that mark short of known failure rates to cover their butts when it comes to usage scenarios like this.

It could be fine for while, but without any voltages described, we have no idea if it is in spec other than it is able to run the hardware you listed.

For reference, I just looked, and if I had the right PSU, you have a 5-year warranty.

 

[TRIPTEX_CAN]

A good indicator of what to expect is the warranty period. While many products can outlast the warranty, the manufacturers will set that mark short of known failure rates to cover their butts when it comes to usage scenarios like this.

It could be fine for while, but without any voltages described, we have no idea if it is in spec other than it is able to run the hardware you listed.

Totally makes sense. It shipped with a 5year warranty.

After work Ill do what I can to get the loaded 12v/5v measured and report back.

 

[SoNic67]

In average the electronics can last twice the warranty period (probabilities). If that PS was made in 2008, it's about time to worry about the actual performance (because issues with dried capacitors mainly).

 

[TRIPTEX_CAN]

In average the electronics can last twice the warranty period (probabilities). If that PS was made in 2008, it's about time to worry about the actual performance (because issues with dried capacitors mainly).

Other than instability and monitoring the output under load what else is there to lookout for?

 

[sneekypeet]

I do agree a product can well outlast its warranty, but I will just talk about the obvious.

There has to be doubt in your mind about the PSU, for whatever reason.... hence the thread. Otherwise you would have just carried on to the next build.

Being the one thing that has to be right for all other things to work properly, and some doubt in my mind, I would replace it and demote it for testing other products if the need arises.

I feel like you are asking someone to tell you it is fine, to save some build costs, but again, I would replace it.

 

[27MaD]

Idk , i'll not trust a PSU that has been running on different systems for about 10 years , It might finger you at any moment.

 

[jboydgolfer]

i had an older Corsair unit, it was a Cx600. it lasted nearly 10 years. the caps finally bulged & blew when it was in its 5th or more PC. it did isolate the problem, & didnt take any other components with it. FWIW, that Corsair TX750w from 2008-2009, that is years after the 80+ rating system came into play, and right around the time the 80+ gold/etc came into play.

if yours does similar to my old corsair, id suggest you keep an eye on the Caps.

the thing below , SamXon

 

[27MaD]

i had an older Corsair unit, it was a Cx600. it lasted nearly 10 years. the caps finally bulged & blew when it was in its 5th or more PC. it did isolate the problem, & didnt take any other components with it. FWIW, that Corsair TX750w from 2008-2009, that is years after the 80+ rating system came into pla, and right around the 80+ gold/etc came into play.

Tbh If it was me i wouldn't power up an expensive components with it.

 

[silentbogo]

Just curious if anyone else has had this sort of lifespan from a PSU

Many. Had several Chieftec PSUs from 2003-2004 which are still chugging along. Some were so old, that they only had a 20-pin ATX connector with 4-pin CPU_PWR and not a single SATA.
Those are only good for some random electronic projects, like tesla coils or big-ass LED curtains... and only if you are willing to take the risk of having a fire hazard nearby.

My stepdads PC is powered with my old TT Toughpower (bought it in 2008 or so).
Also had an Athenapower 400W SFX PSU which I gave to my co-worker a few years back. It served well for over 5 years just in my household, and still held the spec better than some newer 80+ Bronze units.

Other than instability and monitoring the output under load what else is there to lookout for?

Temperature and noise. If your PSU was quiet before, and all of the sudden you start hearing high-pitched noise during startup or at high loads (except maybe high-FPS GPU testing), then it's probably time to take it to a doctor or replace it with a new unit.
Personally I wouldn't use a PSU for more than the warranty period. Usually I switch my PSU every time I'm going for a major platform upgrade every 4-5 years, which is due pretty soon BTW. Gonna get me some Ryzen action and replace my 450W Seasonic with something slightly beefier.

 

[TRIPTEX_CAN]

I do agree a product can well outlast its warranty, but I will just talk about the obvious.

There has to be doubt in your mind about the PSU, for whatever reason.... hence the thread. Otherwise you would have just carried on to the next build.

Being the one thing that has to be right for all other things to work properly, and some doubt in my mind, I would replace it and demote it for testing other products if the need arises.

I feel like you are asking someone to tell you it is fine, to save some build costs, but again, I would replace it.

No real doubts other than not being able to get a read on the output yet and that's just to satisfy the need to know.

 

[SoNic67]

Capacitor plague:
https://en.wikipedia.org/wiki/Capacitor_plague

 

[jboydgolfer]

Tbh If it was me i wouldn't power up an expensive components with it.

everybody has their own ideas, or levels of what they're willing to chance. Before mine died ,it showed absolutely no signs of trouble, and when it did die ,all it did was just that ,it died & the computer power down ( essentially the absolute best a person can hope for during a power supply critical failure). I removed it ,put in a new one , & all was well (Aside from the smell).

Of course there is the possibility something different could've happened, but if we're planning for what could happen, in an attempt to avoid it ,that's a fine line between paranoia & carefullness. I say, the best we can do is buy the best components our budget allows, take good care of them, and replace them when needed. Ive had 3 Power supplies die on me personally, 2 were not new when i got them, none caused any damage to other component (apparently it can happen), or so ive been told, but its just not happened to me. & oddly enough, the longest lasting PSU ive owned , was a Corsair CX600 (what some call a VERY bad quality PSU)

 

[TRIPTEX_CAN]

everybody has their own ideas, or levels of what they're willing to chance. Before mine died ,it showed absolutely no signs of trouble, and when it did die ,all it did was just that ,it died & the computer power down ( essentially the absolute best a person can hope for during a power supply critical failure). I removed it ,put in a new one , & all was well (Aside from the smell).

Of course there is the possibility something different could've happened, but if we're planning for what could happen, in an attempt to avoid it ,that's a fine line between paranoia & carefullness.

Once I monitor the output....assuming its within spec, I'll pull the PSU and check the caps for peace of mind.

 

[HTC]

Once I monitor the output....assuming its within spec, I'll pull the PSU and check the caps for peace of mind.

Perhaps i should do the same since i too have that PSU model for roughly 10 years.

 

[TRIPTEX_CAN]

Perhaps i should do the same since i too have that PSU model for roughly 10 years.

This is the reply I was waiting for.

Let me know if you do pull it out have a look.

 

[Bill_Bright]

I agree the warranty period is a good indicator of the confidence the maker has in their product. But the warranty period is no guarantee of longevity. I have seen budget, no-name generic PSUs last 10 years and longer in systems that have never been opened up for cleaning or anything else. They just keep chugging along. And then I have seen top tier units from the most reputable brands be DOA or die shortly thereafter. That's actually typical for electronics (at least that which is not abused). That is, it dies within a few hours of first power up, or it lasts for years and years.

Much depends on the power being fed to the power supply. If you are feeding your supply, good, clean stable power with a minimum number of surges and spikes, you have a good chance for a long life. But note one excessive surge or spike can ruin your day with even the best of supplies.

Other than instability and monitoring the output under load what else is there to lookout for?

Not just under load - under all expected loads. Computer PSUs are routinely running with minimal loads at idle up to full 100 % loads when taxed. A supply should remain stable and provide a clean output under all expected loads.

The allowed voltage variance is ±5%. So acceptable tolerance maximums:

12VDC ±5% = 11.4 to 12.6VDC
5VDC ±5% = 4.75 to 5.25VDC
3.3VDC ±5% = 3.14 to 3.47VDC​

But another important fact is ripple - unwanted AC riding the DC. But most users don't have the necessary test equipment or know-how to measure it. The maximum allowed is 120mV p-p (millivolt peak to peak) on the 12V rails, and 50mV p-p on the 5 and 3.3V rails. But frankly, any PSU that shows over 30mV would be down rated for poor ripple suppression.

As far as the expressed concerns about using an older supply, IMO, those concerns are misplaced. While it is not unheard of, it is extremely rare for a failed PSU to cause collateral damage to connected components. Typically, when a PSU fails, it just stops outputting voltage. It does not suddenly push out 10,000V at 1000A! Nor do they suddenly explode in a ball of fire.

If your PSU is sized with ample head room in the first place, is outputting within allowed tolerances, and if your computer is stable (no unexplained reboots, shutdowns, or freezes), then I see no problems continuing to use the PSU.

Do you replace the PSU in your $2000 home theater receiver just because it is 5 years old or because the warrant ran out 2 years ago? Do you replace the PSU in your $2000 big screen TV because the TV is out of warranty? Both of those devices are very computerized these days. In fact today's TVs are computers. And a home theater receiver might be connected to $5000 speakers. And all those supplies have caps in them too. Yes, capacitor age but electrical engineers know this. And they compensate for it!

And you can pull your PSU and "check" the caps. But do you know how? Yeah, you can look for leaks and bulging caps, but aging and failing caps don't always leak or bulge. In fact, most caps don't! If there is no apparent visible damage, the only way to check them is with a good capacitance checker - and that assumes you know what to expect.

It seems many seem to forget that computers are home electronics too, and follow the same Laws of Physics and characteristics as other consumer electronics.

I have carried PSUs over to upgraded computers many times with no regrets. As long as it has the necessary horsepower and connectors, the fan bearings are not whining, coils are not whining and transformer plates are not buzzing, I say "use it".

 

[dorsetknob]

A good indicator of what to expect is the warranty period. While many products can outlast the warranty, the manufacturers will set that mark short of known failure rates to cover their butts when it comes to usage scenarios like this.

See above
the minimum is what the Manufacture Specify s per the Warranty. your's might last for much longer (with luck it will)

 

[John Naylor]

The original TX certainly was not what the V2 was but it wasn't garbage either. Jonnyguru gave it an 8.5 perfomanmce ratings but back then he wasn't including quality ratings. Tho, for some reason this referred more to labor quality in the "aesthetics" rating (9.5) than "looks". This comments now appear in Build Quality section. Warranty can be an indicator of quality but we have also seen PSU manufacturers take a PSU that nmaybe didn't do so well in reviews and slap long warranty period on it (EVGA B1 for example)

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story4&reid=73

The V2 version picked up the performance to 9.5 but the Build Quality rating was only 8.5.

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story5&reid=230

Of course there is no reason to buy more PSU than you need and this of course will depend on how hard you drive your system. Depending on that answer, these are my yardsticks. Ripple and Noise are of course very important considerations but you won't be getting high JG performance numbers if ya haven't got that well covered. Voltage regulation is important for OC'ing as all the component VRMs will be chasing the supply voltage and have to adapt o any changes. So if the PSU can do better, that makes their job easier.... stable voltages => stable OC.

- For stock systems with no more than "factory" or light overclocking .... JG Performance and Build Quality Ratings in the range of 8.5 - 9.0 and voltage regulation within 5% (ATX standard).

- For moderate overclocking .... JG Performance and Build Quality Ratings of 9.0 - 9.5 and voltage regulation within 2.5%.

- For serious overclocking .... JG Performance and Build Quality Ratings of 9.5 - 1.0 and voltage regulation within 1.0%.

It seems many seem to forget that computers are home electronics too, and follow the same Laws of Physics and characteristics as other consumer electronics.

Good to have another brethern who applies physics to his electronic components. I'm usually content when they follow the Laws of Electricty and Laws of Thermodynamics ... but if if not I'll subject them to the "Laws of Physics" (the science that studies matter and its motion / behavior through space and time) ad throw them against the wall . .... sorry, but being a Noo Yawker, hard to resist the opportunity to do the "elbow tap to the rib" thing.

 

[Bill_Bright]

I'm usually content when they follow the Laws of Electricty and Laws of Thermodynamics ... but if if not I'll subject them to the "Laws of Physics"

The Laws of Physics encompass all the laws of the physical sciences, including those that dictate how electrons flow through conductors and how heat relates to other forms of energy. So I just say the Laws of Physics unless I am referring to some specific law, like Ohm's.

There is absolutely nothing wrong with being self-taught. But I have found that many who have not be formally trained don't realize that computers are NOT very unique. In fact, the vast majority of the components that make up computers are used in many different types of electronics, just arranged a little differently. But an electron is still an electron and behaves the same whether flowing through a computer power supply, a TV, or a toaster oven. As a matter of fact, computer power supplies are just switching power supplies using in many electronics. And they don't get replaced just because they are x number of years old, or out of warranty. No need to do that for a computer either.

My supplies generally get retired because the computer itself becomes obsolete and gets retired.

 

[jboydgolfer]

As mentioned earlier, the warrantee a retail manufacturer is willing to cover is a decent method of determining anticipated lifespan.

For me, I also look at what OEM components power/make up a given item. Since even the fanciest brand can use components that others do & charge much less. Essentially it's best to inform yourself as much as possible

 

[Bill_Bright]

the warrantee a retail manufacturer is willing to cover is a decent method of determining anticipated lifespan.

That suggests planned obsolescence. I don't abide by that. The warranty period can be used as a rough guide to determine the minimum life expectancy - not how long it will last.

 

[Mr.Scott]

I agree the warranty period is a good indicator of the confidence the maker has in their product. But the warranty period is no guarantee of longevity. I have seen budget, no-name generic PSUs last 10 years and longer in systems that have never been opened up for cleaning or anything else. They just keep chugging along. And then I have seen top tier units from the most reputable brands be DOA or die shortly thereafter. That's actually typical for electronics (at least that which is not abused). That is, it dies within a few hours of first power up, or it lasts for years and years.

Much depends on the power being fed to the power supply. If you are feeding your supply, good, clean stable power with a minimum number of surges and spikes, you have a good chance for a long life. But note one excessive surge or spike can ruin your day with even the best of supplies.

Not just under load - under all expected loads. Computer PSUs are routinely running with minimal loads at idle up to full 100 % loads when taxed. A supply should remain stable and provide a clean output under all expected loads.

The allowed voltage variance is ±5%. So acceptable tolerance maximums:

12VDC ±5% = 11.4 to 12.6VDC​

5VDC ±5% = 4.75 to 5.25VDC​

3.3VDC ±5% = 3.14 to 3.47VDC​

But another important fact is ripple - unwanted AC riding the DC. But most users don't have the necessary test equipment or know-how to measure it. The maximum allowed is 120mV p-p (millivolt peak to peak) on the 12V rails, and 50mV p-p on the 5 and 3.3V rails. But frankly, any PSU that shows over 30mV would be down rated for poor ripple suppression.

As far as the expressed concerns about using an older supply, IMO, those concerns are misplaced. While it is not unheard of, it is extremely rare for a failed PSU to cause collateral damage to connected components. Typically, when a PSU fails, it just stops outputting voltage. It does not suddenly push out 10,000V at 1000A! Nor do they suddenly explode in a ball of fire.

If your PSU is sized with ample head room in the first place, is outputting within allowed tolerances, and if your computer is stable (no unexplained reboots, shutdowns, or freezes), then I see no problems continuing to use the PSU.

Do you replace the PSU in your $2000 home theater receiver just because it is 5 years old or because the warrant ran out 2 years ago? Do you replace the PSU in your $2000 big screen TV because the TV is out of warranty? Both of those devices are very computerized these days. In fact today's TVs are computers. And a home theater receiver might be connected to $5000 speakers. And all those supplies have caps in them too. Yes, capacitor age but electrical engineers know this. And they compensate for it!

And you can pull your PSU and "check" the caps. But do you know how? Yeah, you can look for leaks and bulging caps, but aging and failing caps don't always leak or bulge. In fact, most caps don't! If there is no apparent visible damage, the only way to check them is with a good capacitance checker - and that assumes you know what to expect.

It seems many seem to forget that computers are home electronics too, and follow the same Laws of Physics and characteristics as other consumer electronics.

I have carried PSUs over to upgraded computers many times with no regrets. As long as it has the necessary horsepower and connectors, the fan bearings are not whining, coils are not whining and transformer plates are not buzzing, I say "use it".

Right on.

 

[TRIPTEX_CAN]

Ai suite installed. All voltages appear nominal at load and idle.

Like I said before I just wanted to post this thread to get some opinions. This PSU isn't whining, popping, squealing, fizzing, getting hot, or anything else of that nature.