PSU and cap ageing

Associate
Joined
30 Jan 2003
Posts
1,514
Location
Cardiff
I have a bequiet! Dark Power Pro 850w from 2013. £149 well spent, still runs fine.
Except the other day I noticed my electricity bill is twice the average for a terrace house.
And my bedroom is very warm, 20c in winter without a radiator on...always the warmest room in the house. What else is in my bedroom? Why, a Panasonic 46" plasma from 2008.

So I decided to check both on my multimeter. The last time I looked was many years ago, and the TV was around 230 watts, and the PC used to idle at 30-100w, 350w load. The most I had before was with 2 cards mining Bitcoin, pushed it to 500w. I have updated cpu and gpu since I got that PSU, now I have a 8086k not oc'd, and a 1080TI, not oc'd. So imagine my surprise when I saw my PC was drawing over 900watts, when not even gaming, just browsing! And the plasma TV 950watts, together 1.8kw, total cost about £68 a month. I thought it was a faulty multimeter, but the LCD tv downstairs is a mere 80w, and the PC in sleep mode is 1watt. Still shocked by it tbh.

I'm guessing the PSU has dropped to about 40% efficiency or less. Same for the caps in that Panny plasma. Gonna replace the plasma with my other LG LCD that only pulls 80w, as for that PSU, I don't need 850w to drive a 1080ti, so I'll downsize to a tier 3 600w, and replace that after 5 years.

tldr; check your power draw on old PSUs! you may be ******* away money, like I was.
 
Man of Honour
Joined
13 Oct 2006
Posts
91,129
Something not right there - if the caps in a PSU go you'll get poor power regulation/ripple and inability to supply instantaneous current demands causing random restarts but it won't go from drawing 500 watt to 900 watt with age normally even as components are on their way out.

My main system's current PSU is almost that vintage and my setup an inline power monitor on the mains side - there has barely been any change in draw at the wall under the same circumstances since day 1.
 
Soldato
Joined
6 Jun 2008
Posts
11,618
Location
Finland
Only way for capacitors to consume power (very momentarily) is by going boom!
And neither can PSU itself consume that much power without some nice smell of burnt electronics rising from it.
Also that plasma TV can't consume that kilowatt without being hot to touch.

There's propably some kind of interference there messing what ever you use to measure power.
 
Associate
OP
Joined
30 Jan 2003
Posts
1,514
Location
Cardiff
I'll test it directly from the plug. Could be my extension doing it. I hope so cos that will save me ££££
caps do age 10% a year and its 8 years, so its not impossible the efficiency is well down.
 
Man of Honour
Joined
13 Oct 2006
Posts
91,129
Capacitor ageing is complex as it slows with time - they don't age 10% a year, year on year. "10% ageing" is usually from their max temperature rating as well - at nominal operating temperature and/or room temperature that can be much lower. Under normal conditions a capacitor at 10 years will only be twice as "aged" as it was at the end of 1 year give or take.

There are some exceptions - some cheap generic ones and some older notable brand ones, etc. would age poorly and fall off a cliff after 8-10 years - especially when experiencing the temperatures of a PC PSU environment but the effect would be instability under higher loads than they previously could handle and at the more extremes degraded component lifespan.

That won't reflect directly into an efficiency difference 1:1 of a power supply - that efficiency mostly comes from the transformer, voltage regulators* and how warm the power supply is running - if you are seeing a big efficiency hit it may be due to the airflow being blocked but if it was resulting in an additional 100s of watt draw some components would be getting insanely hot and dying in hours.

* capacitors can have a minor effect on the efficiency of transistors and regulators.
 
Back
Top Bottom