Jump to content

Do PSUs have a life span

Kaii_Kuro
By Kaii_Kuro
in Power Supplies
 Share
Go to solution Solved by mariushm,

There are some components in a power supply that will degrade at a very slow pace over time. Some degrade at a faster rate the higher the operating temperature is.

 

Fans can go bad over time. In high end power supplies like AX1200 they tend to use high quality fans, rated for 50k+ hours of operation, well lubricated, and so on.

In cheaper power supplies, depending on how they're cycled on and off, fans can lose lubrication, due to repeated heat/cold cycles the plastic of the fan blades could crack and break, blades could tear off.

 

Electrolytic capacitors use a liquid or semi-liquid substance called electrolyte, which can be affected by temperature, and can also be damaged by a tiny amount by various events (over voltage, sudden very high current spikes etc)... and they can repair themselves to some degree.

 

Modern power supplies use a mix of solid (polymer) capacitors which are much more resilient because they don't have electrolyte, and the electrolytic capacitors are used in a secondary role, so they're less "stressed" compared to how electrolytic capacitors were used in designs from years ago.

 

Usually electrolytic capacitors are rated for some amount of hours at 105c or even higher, in some high end power supplies you may see 125c rating.

So for example, let's say the electrolytic capacitor is rated for 10,000 hours at 105c 

The rule is that for every 10c reduction in ambient temperature (temperature the capacitor is subjected to), the lifetime rating of that capacitor is doubled.

So for an AX1200 power supply that may be around 60-70c (at very high load) around the area where such capacitors are installed, you can do the math :

10,000 at 105c , 20,000 hours at 95c , 40,000 hours at 85c , 80,000 hours at 75c and so on ...  there's around 9000 hours in a year, but you wouldn't be gaming or doing something super heavy so that your psu would average 75c all the time ...

 

Solid/polymer capacitors have a much higher life, which is determined with another formula, but basically the average 5000h@105c solid/polymer should last 20+ years at at 60-75c ambient temperature.

Posted
2 hours ago, 8tg said:

I am intensely jealous as a collector of cool old computer parts 

If you're into retro-tech, I have something you might be interested in. 

 

https://www.guru3d.com/articles-pages/ocz-fatal1ty-1000w-psu-review,1.html

 

Brand new in the box.  Just need to make sure the bulk caps don't blow up after not being plugged in for over 10 years.

 

Too bad I threw away my Ultra UV Green PSU a couple weeks ago.  😞

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
1 hour ago, jonnyGURU said:

If you're into retro-tech, I have something you might be interested in. 

 

https://www.guru3d.com/articles-pages/ocz-fatal1ty-1000w-psu-review,1.html

 

Brand new in the box.  Just need to make sure the bulk caps don't blow up after not being plugged in for over 10 years.

 

Too bad I threw away my Ultra UV Green PSU a couple weeks ago.  😞

 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
On 7/30/2022 at 2:33 AM, jonnyGURU said:

Then the next thing I see go is because the thermal compound between the MOSFETs and the bridge diodes dry up and those parts start to run hotter than they should.  They don't use "Arctic Silver" or "Grizzly" level stuff at the factory and it's not like people replace the TIM on their PSU components like they do on their CPUs and GPUs even though these parts are running upwards of 100°C.

Oof, don't I know it xD, I'm having PTSD from trying to repair some quality speakers a long time ago, and spending weeks hunting down similar diode replacements, all to be undone by my horrible soldering skills (I ended burning the PCB and killing it). Just because Microlab screwed up the TIMs/cooling around the mosfets, ... Far cry from soldering new fan wire for a PSU or two in highs-chool in 1 min 😛

 

Slightly less PTSD for fans, but I got a few fans where I failed to solder the wires, change connectors (around the same time as speaker above) xD So now I got a simple methodology for PSU maintenance:

 

1. If fan goes bad, stick a replacement, and run the wire outside the PSU and connect it to the motherboard or a cheap-ass fan controller and run at constant speed [some high quality matshushita-grade rifled/maglev + ideally LCP top of the line silent {for me} at 1000-1200RPM at least] (with some isolation tape to make sure the wire doesn't get damaged, and making sure I didn't put too much tape/flex the metal housing so the faraday cage effect is not compromised).

 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
9 hours ago, jonnyGURU said:

I had one of these.  Great conversation starter at LAN parties.

 

image.jpeg.5db15cc1dfdf08a3be39573c9ee196d0.jpeg

image.png.b5bb2938c70d10a45c2df4ead8d840a2.png

Question :

Why is it that we can make power bricks (zero internal airflow) but can't make fully watercooled PSUs?

 

What is it that requires direct contact with the coolant (air/non electrictricly conductive liquid/epoxy) and can't be cooled with any sort of a heatsink? 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
5 hours ago, Cifer said:

Question :

Why is it that we can make power bricks (zero internal airflow) but can't make fully watercooled PSUs?

 

What is it that requires direct contact with the coolant (air/non electrictricly conductive liquid/epoxy) and can't be cooled with any sort of a heatsink? 

I think it's the wattages involved, most power bricks top out at like...150W? I bet a 150W PSU would run OK passive if it was designed for it.

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author
6 hours ago, Cifer said:

Question :

Why is it that we can make power bricks (zero internal airflow) but can't make fully watercooled PSUs?

 

What is it that requires direct contact with the coolant (air/non electrictricly conductive liquid/epoxy) and can't be cooled with any sort of a heatsink? 

Totally around the wattage and amperage. Those bricks just don't have as high demand as a psu. Also honestly it would be a really bad idea to get it accidentally wet. You could also argue do you need any cooling at all on a psu 80% of the time anyway. Hybrid psu fans barely ramp when doing the basic tasks.

 

Honestly I just wish they made crazy computer parts. The designs are so boring and functional now days. I miss mad parts.

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
4 hours ago, Bitter said:

... 

 

2 hours ago, Kaii_Kuro said:

... 

 

My question is : " What is it that is impossible to watercool in a PSU? " 

 

If we can passively cool components with thermal epoxy in low power scenarios, why can't we do the same thing in high power scenarios by cooling them actively instead (thermal epoxy + some waterblock of some sort that keeps the thermal epoxy nice and cool).

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
1 hour ago, Cifer said:

 

 

My question is : " What is it that is impossible to watercool in a PSU? " 

 

If we can passively cool components with thermal epoxy in low power scenarios, why can't we do the same thing in high power scenarios by cooling them actively instead (thermal epoxy + some waterblock of some sort that keeps the thermal epoxy nice and cool).

Cost. 

 

Link to comment
Share on other sites
Link to post
Share on other sites
Posted
  • Author
3 hours ago, Cifer said:

 

 

My question is : " What is it that is impossible to watercool in a PSU? " 

 

If we can passively cool components with thermal epoxy in low power scenarios, why can't we do the same thing in high power scenarios by cooling them actively instead (thermal epoxy + some waterblock of some sort that keeps the thermal epoxy nice and cool).

Risk to reward

Link to comment
Share on other sites
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share
Go to topic listing Power Supplies

×