Is it ok to keep the PC plugged in on wall or switch off the PSU when it is shut down?

[Newblesse Obblige]

I always kept it plugged in from the wall. I heard it is bad but I want some confirmation from people who has knowledge when it comes to this.

[Dujith]

Having your computer plugged in isnt normally an issue. That how most of the computers in offices/school/home are used anyway.

It can be a problem if you live in an area that has recurring outages/spikes. Then you would normally use a protection or just unplug it.

 

I would say just let it plugged in, unplug it if you go away for a while or if there is a thunderstorm incoming.

 

[mariushm]

Unless there's some heavy weather outside (thunder storms, risk of lightning strikes) leave the PC plugged in with the power supply on  (don't unplug or turn off the power supply from the on/off switch)

 

When the PC is off, it's consuming very little power, usually less than 0.5 watts ... which is basically a few cents a month in electricity costs.

 

If you completely unplug the pc, or turn off the power supply, you'll force the motherboard to actually use that CR2032 battery to retain the bios settings, instead of using power from the standby power supply to keep those settings. So, the battery will discharge much faster and you'll have to replace it more often. 

 

Otherwise there's practically no downsides to leaving pc powered with the stand-by voltage of the power supply. 

[Moonzy]

38 minutes ago, mariushm said:

If you completely unplug the pc, or turn off the power supply, you'll force the motherboard to actually use that CR2032 battery to retain the bios settings, instead of using power from the standby power supply to keep those settings. So, the battery will discharge much faster and you'll have to replace it more often. 

there's also the charging and discharging of the PSU's capacitor, which isnt great

 

39 minutes ago, mariushm said:

When the PC is off, it's consuming very little power, usually less than 0.5 watts ... which is basically a few cents a month in electricity costs.

with all the standby RGB it's like 1-2W judging from my meter

 

leave the PSU switched on while the computer is turned off is normal

and most of the time recommended (i think)

[akio123008]

It's normal to leave it plugged in. It's sort of what they're designed for. 

 

It doesn't matter too much but most people leave it plugged in as the power savings from unplugging it would be very small.

 

There is the argument of a plugged in PSU being a fire hazard, but that's only the case for an exceptionally bad power supply or one that's extremely old. 

[mariushm]

1 hour ago, Moonzy said:

there's also the charging and discharging of the PSU's capacitor, which isnt great

 

with all the standby RGB it's like 1-2W judging from my meter

 

leave the PSU switched on while the computer is turned off is normal

and most of the time recommended (i think)

There's a separate 5v stand-by power supply, and a separate power supply for the "heavy stuff" 

In a lot of designs, the PFC circuit is disabled so not even the big primary capacitor is charged. 

 

You have the equivalent of a usb phone charger in the power supply, with its own input and output capacitors and switching transformer. The "wear and tear" on capacitors in a modern power supply is not worth talking about it.

 

Nobody forces you to power the RGB stuff from 5v stand-by or from usb headers powered from 5v stand-by (you have option in bios to disable that feature)

 

[Ankerson]

4 hours ago, Newblesse Obblige said:

I always kept it plugged in from the wall. I heard it is bad but I want some confirmation from people who has knowledge when it comes to this.

 

Keeping it plugged into the wall is how you are supposed to do it.

 

Unplugging it drains the caps, not a good thing and hard on the PSU, so they have to charge back up in a sudden surge.

 

It's best NOT TO DO THAT.

 

 

[mariushm]

9 minutes ago, SGT-AMD said:

That is what a cap does.

Temp is more important than charge/discharge cycles.

A 10 degree c drop in temp, doubles the life of caps.

 

The temperature for a primary capacitor is not really a factor. 

 

Yeah, it's a simplification to say that for an electrolytic capacitor, its lifetime rating doubles with every 10c of lower temperature ... but the lower temperature is the temperature of the capacitor itself, which will be lower than the temperature of components around it most of the time.

 

So a big primary capacitor with lots of volume and easy to  cool itself simply by radiating  the heat.. you're not gonna get a big variation in lifetime rating.

For example, let's say you have a 4000h @ 105c capacitor ... a primary capacitor is probably gonna be at most at 50-60c ... so you have 8k @ 95c , 16k @ 85c , 32k @ 75c , 64k @ 65c .... there's 9k hours in a year .. so even if you'd run the power supply 24/7 with the capacitor at 60 degrees, it would be within 10-20% of its rated specs for 6-8 years. 

In real world, since you don't use the pc 24/7 and the fan cools the primary capacitor, it's 10-15 years at least for the primary capacitor, if it's from a brand name, quality one. 

 

ps. for polymer capacitors, formula is different, and the estimation is no longer correct.   They have a much higher lifetime rating 

For polymers, rating is   original lifetime  x  10 ^{(temp rating original  -  ambient temperature)/20} 

 

[mariushm]

1 hour ago, SGT-AMD said:

Q:  How does Corsair calculate the life expectancy of a capacitor when deciding what to use in a particular power supply?

A:  Capacitors have a few specifications and ratings. 

[...]

So what did I just say ... that is a simplification, a text written like "explain like i'm five" so that anyone can overstand. Also, it applies only to ELECTROLYTIC capacitors.

 

If you want to be anal about it, the formula for electrolytic capacitors is this :

 

Lifetime =  LifetimeOriginal  x (Vr / Va)ⁿ x 2^A x 2^B

where:

n = 1 for voltage >= 160v ,  0 for voltage < 160v

Vr = rated voltage

Va = applied voltage (in circuit )

^*Vr/Va is limited to 2:1 ratio.

A = (T rating - T ambient ) / 10

B =  (deltaTo - deltaT) / C

 

where :

C = 5 for 105c capacitors,  10 for 85c rated capacitors

deltaT0 = 5 for 105c capacitors,  10 for 85c rated capacitors, 0 for capacitors without ripple rating

deltaT = C x (Iapplied ripple /Irating ripple)²

 

But LIKE I SAID, for regular electrolytic capacitors, the ones you would normally find on an old motherboard, you can simplify that formula and just estimate that for every 10c decrease the lifetime rating of a capacitor doubles.

It's not quite the same with BIG PRIMARY capacitors, and which are away from hot heatsinks and mosfets ... the active PFC circuits can abuse them with high current pulses but if they're properly chosen they don't really get too hot (shouldn't be above around 60c)

 

Polymer capacitors (more and more used inside power supplies on the secondary side and on motherboards) are rated differently, like with the fomula I said above....

 

 

 

[Ankerson]

2 hours ago, SGT-AMD said:

Who told you that?

 

 

@jonnyGURU

 

Let him explain it to you.

[Guest]

39 minutes ago, SGT-AMD said:

In the radio business, especially older designs, Power supply caps were right next to the heatsink(s).

 

That is why I use computer grade caps where heat is an issue. More money, longer lasting, Worth it

 

 

Can't argue with that.

 

I think what has been explained has been taken out of context.  The CAPACITORS aren't damaged by the constant charge and discharge.  Like you said, that's what they're designed to do.  It's the components leading up to those capacitors.  The inrush current limiter, the mechanical switch, etc.  Unlike the caps, these parts DO have a limited life cycle.

[Guest]

1 minute ago, SGT-AMD said:

The Chinese vendor said that he had never heard of this issue.

 

Chinese vendor = "He lied".

 

I'm going to sound racist here, but I haven't encountered a 100% honest Chinese businessman in my entire life and I've been dealing with them for a couple decades now.  And after reading Paul Midler's "What's Wrong With China", I understand it's a cultural thing.

 

Good reading:

https://www.amazon.com/Whats-Wrong-China-Paul-Midler/dp/1119213711/

 

 

[ninioArtillero]

Sorry to revive this thread but I have a minor doubt left. I get hat the components leading to the capacitors get worn out by power on-off cycles, paraphrasing @jonnyGURU. But the question I have regarding all this is on the safety of leaving the PSU on or off-and-plugged-in during a thunderstrom. Shouldn´t the over current or overtension be eased to ground (in both cases)?

 

I understand that what @Dujith said in the first comment is common sense, but I would like to get detailed on the matter.

[Dujith]

On 8/18/2021 at 11:42 PM, ninioArtillero said:

But the question I have regarding all this is on the safety of leaving the PSU on or off-and-plugged-in during a thunderstrom. Shouldn´t the over current or overtension be eased to ground (in both cases)?

Short answer: No and no.

There are a few issues here:

• PSU's can have a over current or over voltage protection, tho that is meant for small power surges coming from the net. Not massive spikes.
• Alot of PC's are not grounded correctly. Be it dodgy wiring in the house or just no ground at the outlet. So the only path would be the inside.

But those 2 are moot anyway regarding a thunderstorm. A lightning strike nearby will kill of most appliances, there is no protection that can deal with that amount of energy in that short of time. Only defense is lightning rods or putting said appliance in a faraday cage  

 

 

[Offbeatvibe]

So I've been having to turn off my PC completely off for the last few months, not from the power supply switch, but from the power strip. Should I be worried that I've lowered the life expectancy of my power supply or something?

[ninioArtillero]

On 8/22/2021 at 3:31 PM, Offbeatvibe said:

So I've been having to turn off my PC completely off for the last few months, not from the power supply switch, but from the power strip. Should I be worried that I've lowered the life expectancy of my power supply or something?

Regarding the unloading and reloading of the capacitors in the power supply (PSU), yes: this does reduce life of the capacitors to a degree.

 

I do turn off the PSU switch off before unplugging the power strip (in my case a UPS) just to be extra careful but I do wonder if it really makes any difference though, considering it could mean just unnecessarily spoiling the switch. On the other hand, carving your question, I wonder if cutting power from the power strip rather than the PSU switch could jolt the capacitors harder.