Help me diagnose this shit

[Pentium100]

although, old components has a far higher chance of giving out than a brandnew component, with in mind that the average person upgrades once every 4years, or stingy persons sticking with theirs for over 10years(people who still uses pentium4s... no offense)

It also costs more than just keeping the old component. By the way, I have a server that uses Netburst CPUs. It is powerful enough for me at the moment and I didn't pay a lot for it, but it has a rackmount case (1U) which is nice.

sure you can repair the unit if it starts to give out but how much would it cost to repair(excluding the trouble of repairing them, since pretty much repurchasing one is equally troublesome)? at around 4years it's worth has been well used, selling it for 30-40%(70-80% if you're lucky) of the original price and repurchasing an either better one or the same thing for much less cost(since you're reinvesting the money back).

With my selling skills i would be lucky if I didn't have to pay someone to take it. Also, new capacitors for a PSU are not very expensive (especially compared to the price of a good PSU) - I could barely buy the cheapest possible power supply for that money.

theres also a good point of switching them instead of retaining them for so long, that is their efficiency diminishes over the years of usage that could've saved money through electricity bills.

I have not noticed that. I have a graph of the load on my UPS and whatever power consumption increase due to aging PSUs there is it most likely gets buried in the noise, so I guess it's not that much).

And yes, this is the realm of preferences. I like that companies sell old servers before thay wear out completely - that way I can get a nice server that is powerful enough for me without having to pay $3000.

[Ivon]

Thing to remember is the Certification of the PSU (80+ Gold, Bronze, Silver, Platinum). And never rely on PSU Calculators...most don't factor in the initial 'kick' of the system.

Get a Kill-o-Watt (ThinkGeek's) meter, you'll see the Wattage spike on the start up then simmer down.

I always recommend adding at least 100Watts to the "Recommended PSU" that the Video Card says to have. ...they're so cheap now and usually the last thing you every need to upgrade (like when they went from 20-pin to 24-pin).

[kitamesume]

^no because most, if not all, of the PSU has a peak non-continuous rating of 1.5-2.0 times their continuous capacity, thats how the generic PSU gets their 500watts max rating.

[Ivon]

^no because most, if not all, of the PSU has a peak non-continuous rating of 1.5-2.0 times their continuous capacity, thats how the generic PSU gets their 500watts max rating.

So you're saying you shouldn't have a buffer above what is recommended by the video card?
-Do believe that recommendation factors in a 'typical system' (CPU, RAM, HDD, etc) in addition to what the card will need.

Probably not understanding what you mean. For some reason when I red "how the generic PSU gets their 500watts max rating" I think of the Wattage listed on the side in large print....which would mean the continuous rating is something like 250watts.

[Pentium100]

Probably not understanding what you mean. For some reason when I red "how the generic PSU gets their 500watts max rating" I think of the Wattage listed on the side in large print....which would mean the continuous rating is something like 250watts.

That's exactly like it is with the really cheap ones. Try puling the rated power for more than a few seconds and the PSU might blow up (since it's too cheap to have overload protection).

[kitamesume]

^no because most, if not all, of the PSU has a peak non-continuous rating of 1.5-2.0 times their continuous capacity, thats how the generic PSU gets their 500watts max rating.

So you're saying you shouldn't have a buffer above what is recommended by the video card?
-Do believe that recommendation factors in a 'typical system' (CPU, RAM, HDD, etc) in addition to what the card will need.

Probably not understanding what you mean. For some reason when I red "how the generic PSU gets their 500watts max rating" I think of the Wattage listed on the side in large print....which would mean the continuous rating is something like 250watts.

did i ever say that you shouldn't add a buffer? where did i say it?
now a day's PSU are true rated at it's continuous output, GPU manufacturer's suggested PSU capacity factors in generic PSU's retarded labeling, if you've noticed why they even recommend you to use a 400watt PSU on a HD7750 that doesnt even use more than 60watts then continue reading below.

if you think the recommended PSU capacity is just about right, think again.

(click to show/hide)

notes: power consumptions            idle   avg   max
i5-3570K (stock settings)            55   82   109   watts    (interpolated - http://www.xbitlabs.com/articles/cpu/display/core-i7-3770k-i5-3570k_8.html )
i3-2100                     41   60   80   watts    (interpolated - http://www.xbitlabs.com/articles/cpu/display/celeron-g540-g440_7.html )
G540                        39   55   74   watts    (interpolated - http://www.xbitlabs.com/articles/cpu/display/celeron-g540-g440_7.html )
motherboard + ram power consumptions      25   50   50   watts (approximations)
HD7850 power consumptions            10   87   101   watts    (http://www.techpowerup.com/reviews/AMD/HD_7850_HD_7870/24.html )
HD7770 power consumptions            7   66   83   watts (http://www.techpowerup.com/reviews/Sapphire/HD_7770_Vapor-X/26.html )
HD7750 power consumptions            5   41   43   watts    (http://www.techpowerup.com/reviews/ASUS/HD_7750/24.html )
HardDrive 7200rpm avg power consumptions   ??   8   30   watts    (approximations)
Dell 18.5 inch IN1912H power consumptions   2   18   23   watts    (manufacturer)
-------------------------------------------------------------------------------------------
i5-3570K + MB + HD7850 + 1HDD + monitor   92   245   312   watts
i5-3570K + MB + HD7850 + 4HDD + monitor   92   269   402   watts
i3-2100 + MB + HD7770 + 1HDD + monitor      75   202   266   watts   
G540 + MB + HD7750 + 1HDD + monitor      71   172   220   watts   
G540 + MB + HD7750 + 4HDD + monitor      71   196   310   watts   

PS: avg power consumption means they're working at their normal state, not fully loaded, not idle.
PSS: i included the monitor, which obviously shouldn't be, although its there for that "extra baggage".

now tell me, how'd a 220watt setup got a 400watt recommended (or higher) psu capacity?

like i said, most true PSUs have a non-continuous peak output of 1.5-2.0 times higher than their rated continuous output, which means a 500watt PSU can output 750-1,000watt for a few milliseconds.
      spike surges from cold boot doesn't last more than a hundred millisecond, the spike surges aren't caused by the components of a computer but by the inductors when they charge up, kill-a-watt just has a slow poll-rate to even mark how long the spike lasted, you'll need an oscilloscope to do that.
this would mean the PSU can handle the aggressive spike without tripping the OCP, although tight OCP settings could interpret that surge as a short circuit.
you can run a generic PSU at it's rated wattage for not much more than a second or two, after that it fries.

[Pentium100]

      spike surges from cold boot doesn't last more than a hundred millisecond, the spike surges aren't caused by the components of a computer but by the inductors when they charge up

Capacitors. Inductors cause voltage spikes when they are diconnected, capacitors cause current spikes when they are connected.
Also, a hard drive uses a few times more power to spin up than when it is already spinning. 10 hard drives spinning up at once can cause a significant load (this is why server hard drives can be made to delay spin up or to wait for a command to spin up - then the controller can spin them up one by one instead of all at once). This surge lasts about 10 seconds.

[kitamesume]

we were talking about equivalent load resistances being too low thats causing the power consumption spikes, inductors acts as a short circuit when its not charged thats what i meant, same thing applies to capacitors.

how old of a harddrive are we talking about o.o, modern harddisk spin-up current draw rarely pushes over 2.5A, thats 30watts which isn't far off my approximations (although i guess i should drop the avg wattage of the hdd).
http://www.pcguide.com/ref/hdd/op/spinPower-c.html


PS: i've done this since quite a long time ago, i built that table so i can easily calculate wattage figures easily, imho since theres a table i could just fill in the blanks and do some simple math.

[Pentium100]

we were talking about equivalent load resistances being too low thats causing the power consumption spikes, inductors acts as a short circuit when its not charged thats what i meant, same thing applies to capacitors.

Inductor acts like an open circuit when it is not "charged". If you connect an inductor to a voltage source, initially there will be no current and the current will slowly increase until it reaches some steady state. Inductor resists the change in current, so when you plug it in it resists the change from 0A to whatever.

how old of a harddrive are we talking about o.o, modern harddisk spin-up current draw rarely pushes over 2.5A, thats 30watts which isn't far off my approximations (although i guess i should drop the avg wattage of the hdd).

0.5A normal, 2.5A on spin up - so that's 5 times more. If you have 10 hard drives, the normal power consumption will be ~60W, but when they all are spinning up at once it would be 300W...

[kitamesume]

ah right yes, i might've remembered it in the wrong way, since an inductor acts as a high resistance circuit to high frequency signals.

but i had my estimated power consumption of the harddrive at 25W means i've factored in the spin-up power consumption, and since its not exactly 2.5A my estimation of 25W isn't too far off.
can you please stop pushing the harddrives as the main cause, generally regular desktop doesn't even have more than 4harddrives in them, even if you factor in 4harddrives on my chart it still doesn't hit 400watts.
a NAS on the other hand is a fileserver.

edit: for simplicity's sake i adjusted the chart.

[Ivon]

^no because most, if not all, of the PSU has a peak non-continuous rating of 1.5-2.0 times their continuous capacity, thats how the generic PSU gets their 500watts max rating.

So you're saying you shouldn't have a buffer above what is recommended by the video card?
-Do believe that recommendation factors in a 'typical system' (CPU, RAM, HDD, etc) in addition to what the card will need.

Probably not understanding what you mean. For some reason when I red "how the generic PSU gets their 500watts max rating" I think of the Wattage listed on the side in large print....which would mean the continuous rating is something like 250watts.

did i ever say that you shouldn't add a buffer? where did i say it?
now a day's PSU are true rated at it's continuous output, GPU manufacturer's suggested PSU capacity factors in generic PSU's retarded labeling, if you've noticed why they even recommend you to use a 400watt PSU on a HD7750 that doesnt even use more than 60watts then continue reading below.

if you think the recommended PSU capacity is just about right, think again.

like i said, most true PSUs have a non-continuous peak output of 1.5-2.0 times higher than their rated continuous output, which means a 500watt PSU can output 750-1,000watt for a few milliseconds.
      spike surges from cold boot doesn't last more than a hundred millisecond, the spike surges aren't caused by the components of a computer but by the inductors when they charge up, kill-a-watt just has a slow poll-rate to even mark how long the spike lasted, you'll need an oscilloscope to do that.
this would mean the PSU can handle the aggressive spike without tripping the OCP, although tight OCP settings could interpret that surge as a short circuit.
you can run a generic PSU at it's rated wattage for not much more than a second or two, after that it fries.

When I read "^no because most", I applied that to the entire message. Including the part about the buffer.

I didn't know they used Generic. I figured it was the SLI/Crossfire capability. So if the card supports 4-way then the 'Recommended PSU' was the total maximum that four cards would need plus a little extra for the system.

I've always stuck with "nothing under 1000-Watts" when making systems, and always with Corsair ...until they start making shitty products then I'll look at other PSU.

As for a kill-o-watt meter, I've never used one because I can watch my UPS's LCD for by the second power fluctuations. I remember it spiking to ~350-Watts when I'd turn on the old CRT monitor I had running on the second machine. Then dropping to something like 212-Watts.

Figured most people don't think about the 'kick'. Figured a Kill-o-watt meter would provide the same function as my UPS, but for a cheaper price.

Though, originally I was thinking it might be a case of 'Bad Caps' those are still turning up on Mobos :\

[kitamesume]

so im already telling you, the recommended PSU capacity on the GPU manufacturer list is too farfetched to be relied on, you must do your own math. although it isn't bad to follow it, but you wouldn't be needing to add much of a buffer if you do so, as you may see, its 1.5-2.0 times over the average setup's max power consumption, thats already a 25-50% buffer.

[Pentium100]

so im already telling you, the recommended PSU capacity on the GPU manufacturer list is too farfetched to be relied on, you must do your own math.

Well, it is written with the cheapest power supplies in mind - the ones that can only deliver half of rated power. Then the manufacturer can reduce the number of people calling support because their card does not work due to underpowered power supply.

You won't need a 500W Corsair PSU to run that card, but you might need a 500W Q-Tec PSU...

[kitamesume]

exactly, so whats the point of quoting me for it?

[Ivon]

Can we just say...crisis averted?