Electricity consumption o. m. rig

[Laurijan]

Hi!

I measured the electricity usage of my rig and it was 9.2 Kilo Watt Hours for two days.. A lot of reboots were done in the second day.. And the measurement for the first day was 4.4 Kilo Watt Hours.

So i have to pay about 1.9€ a day if the PC is turned on all the time. Its quite much for a PSU having a efficiency of 85%..

I suspect my 3 HDDs are using a lot of the KWH measured

[Jizzler]

HD's aren't too bad. CPU's and GPU's are usually the biggest draws.

My box costs me about $13-14 USD/month.

[oli_ramsay]

Out of interest, what did you use to measure the power consumption?

[Sasqui]

Out of interest, what did you use to measure the power consumption?

Curious here too... my "Kill-a-watt" measures watt draw, but not cumulative.

Definitely the CPU and GPU drawing the most by far. HDD's draw less than 20w each (and that's on the high end) - if you look up the specs on your drives you can find out.

[Laurijan]

Out of interest, what did you use to measure the power consumption?

I used a Watt-Meter which was borrowed from the a local electricity company

[keakar]

well part of the cost is in you psu which is way way bigger than you need. looking at your specs it would only need a 400 watt psu to run everything with power to spare so a better sized psu can help you lower the cost a lot. that said the power your using is not that bad and i would guess its about average for most people. unlike some people think a psu is not only using the power it needs to provide whats required but it is nothing more than a transformer. it draws enough power to provide all the watts its rated for at all times and only generates heat when providing full amps so its the heat which is proportional to the power drawn from it not the use of electricity to the unit itself.

EDIT: further research proved the statement i just made above this is incorrect, please ignore it.


you could also cool your entire rig with just 1 rear exhaust fan if you do some proper airflow technics,
read my airflow guide. \/ \/

[Laurijan]

well part of the cost is in you psu which is way way bigger than you need. looking at your specs it would only need a 400 watt psu to run everything with power to spare so a better sized psu can help you lower the cost a lot. that said the power your using is not that bad and i would guess its about average for most people. unlike some people think a psu is not only using the power it needs to provide whats required but it is nothing more than a transformer. it draws enough power to provide all the watts its rated for at all times and only generates heat when providing full amps so its the heat which is proportional to the power drawn from it not the use of electricity to the unit itself.


you could also cool your entire rig with just 1 rear exhaust fan if you do some proper airflow technics,
read my airflow guide. \/ \/

I was told from a nice local computer shop that an high watt PSU wont heat as much as an avarage Watt PSU when it goes at its peaks and thereby would not use as much W as an 350W for example.

Can you prove them wrong more accuratly like with a link to a trusted site?

[newtekie1]

The PSU doesn't draw more than what is needed by the components connected to it, that is just wrong. My thermaltake 750w doesn't draw 750w constantly, my UPS confirms this. When the computer is idle the power draw is only ~240w.

The problem is, that even though your PSU is rated at 85% effeciency, that is just the best possible effeciency it will have. Most power supplies are only that efficient when you get close to putting them under ~80-90% load. As the load goes down, so does the efficiency. So if you have an 750w power supply, and are only putting it under 20% load, the efficiency isn't going to be anywhere near 85%.

[Sasqui]

http://www.motherboards.org/images/articles/guides/550W_Efficiency_chart.jpg

http://www.motherboards.org/articles/guides/1487_7.html

Edit: (Addition):

http://www.silentpcreview.com/files/images/corsair_psu/vx450w_efficiency.gif

http://www.silentpcreview.com/article751-page1.html

[newtekie1]

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story2&reid=7

The efficiency of the Thermaltake 750w peaks around 300w.

[Sasqui]

http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story2&reid=7

The efficiency of the Thermaltake 750w peaks around 300w.

I've read that most peak around the 50% load range (on average), so yea.

[strick94u]

I was told from a nice local computer shop that an high watt PSU wont heat as much as an avarage Watt PSU when it goes at its peaks and thereby would not use as much W as an 350W for example.

Can you prove them wrong more accuratly like with a link to a trusted site?

Its really nothing to do with the psu I could have a 10,000 watt psu and I would only use a faction of it to run my system its the amps that get you to figure watts you simplly need volt and amps Voltage x intensity= watts
12 voltsx4 amps =48 watts even if your running the psu off 120 volts us or 240? eroupe its whats going on inside the case that matters two overclocked 8800 gts 640 under full load on my old setup were hitting as high as 8 amps a minute 12x8=96 watts x60 5kw hour
good news is they were not always hitting 8 amps probly closer to 5 most of the time but shut off the overclock and it was under 3 amps under load. so I created my monster.But as you can see it was only bad during gaming and other intense graphic use. I had no way of measuring my overclocked cpu but judging from my heat it was bad. if your pc goes into sleep mode I don't think it will draw much, but since your a member here I would guess it don't sleep much. the newer stuff coming out now will be much better left at its stock configurations. My old AMD 3400+ laptop gets 45 minutes on a fully charged 3 month old battery my new centrion c2d gets 3.4 hours at 3 times the speed. My old laptop burns your leg my new one don't

[keakar]

I was told from a nice local computer shop that an high watt PSU wont heat as much as an avarage Watt PSU when it goes at its peaks and thereby would not use as much W as an 350W for example.

Can you prove them wrong more accuratly like with a link to a trusted site?

well after further review they are correct, psu use a variable load transformer not a fixed load transformer. it does draw only what you use so i was wrong about that.

you missunderstood them about it not drawing as much watts as a smaller psu would. big or small it draws the same watts because it draws what it needs and that doesnt change because of the psu size.

think of it in terms of a 6 cylinder engine and a 4 cylinder engine pulling the same load up a hill, the 6 cylinder goes up the hill without straining and not working hard helps keep it running cooler, where the 4 cylinder strains to get up the hill so its engine is very hot. this is a perfect example of how a bigger psu will help you reduce the heat output of your computer.

[Laurijan]

Does the draw include the monitor. If thats a CRT thatll use a lot of electricty.

Also you pay over 0.40+ euros per kWh of electricity? In london the average charge is around 12-14p per kWh and we have some of the highest charges in europe. In the states its far less than europe.


Funny avatar you got!

The measurement was taken from the computer + a printer in standby + a connected tv-stick (very hot) + a connected webcam + a backlights having keyboard + a mouse + the true flat trinitron CRT.
I have to messure the monitor alone since its a big CRT and another computer in the living room draws only 1KWH per 30 hours and it has a TFT..

[newtekie1]

think of it in terms of a 6 cylinder engine and a 4 cylinder engine pulling the same load up a hill, the 6 uses less gas and goes up the hill without straining and not working hard helps keep it running cooler, where the 4 cylinder burns a lot of gas and strains to get up the hill so its engine is very hot. this is a perfect example of how a bigger psu will help you.

That isn't really a good example, it depends on the load. A 4 cylinder will outperform the 6 in terms of fuel efficiency going up hill with lighter loads. The same is true about different power supplies.

If you are worried about power usage, set the machine up to use a little power as possible. Have it spin down the hard drives after a short time of inactivity, or even tell the machine to hybernate or go into standby when not being used. Don't have a screen saver, just tell it to turn the monitor off after a short while.

[keakar]

That isn't really a good example, it depends on the load. A 4 cylinder will outperform the 6 in terms of fuel efficiency going up hill with lighter loads. The same is true about different power supplies.

If you are worried about power usage, set the machine up to use a little power as possible. Have it spin down the hard drives after a short time of inactivity, or even tell the machine to hybernate or go into standby when not being used. Don't have a screen saver, just tell it to turn the monitor off after a short while.

ya the statement about gas was ass backwards so i edited it out. i need to quit trying today because im saying shit all wrong lol.

[Laurijan]

ya the statement about gas was ass backwards so i edited it out. i need to quit trying today because im saying shit all wrong lol.

I give you a thanks anyway..

[largon]

I measured the electricity usage of my rig and it was 9.2 Kilo Watt Hours for two days.. A lot of reboots were done in the second day.. And the measurement for the first day was 4.4 Kilo Watt Hours.

So i have to pay about 1.9€ a day if the PC is turned on all the time. Its quite much for a PSU having a efficiency of 85%..Dunno how you got that 1.9€ but it is infact far off the real figure... (1kWh = ~0.05€)

I have my PSU plugged into a powermeter all the time - had it that way for over 5 months now. During that time it has consumed 517.51kWhs that's 3.25kWh a day, which equals 0.16€ / a day on average.

[Laurijan]

Dunno how you got that 1.9€ but it is infact far off the real figure... (1kWh = ~0.05€)

I have my PSU plugged into a powermeter all the time - had it that way for over 5 months now. During that time it has consumed 517.51kWhs that's 3.25kWh a day, which equals 0.16€ / a day on average.

I have make sure its 0.46€ a KWH at day and 0.36€KWH at night - i called my landlord today and asked him about the electricity price here in Oulu Finnland and he gave me this numbers... not lol..

[largon]

Oon aikas varma että vuokraisäntäsi on nyt hiukkasen erehtynyt.
:)

[largon]

I pulled that 0.05€ out of memory.
I just looked it up on the powercontract:

0.053€ / kWh (= 5.3cnt) daytime
0.046€ / kWh (= 4.6cnt) nighttime

[Laurijan]

I pulled that 0.05€ out of memory.
I just looked it up on the powercontract:

0.053€ / kWh (= 5.3cnt) daytime
0.046€ / kWh (= 4.6cnt) nighttime

Maybe this confused woman ment 0.046€ and 0.036€ - Wimen!!

[suraswami]

This is what I do. All my 4 PCs have the same Power settings
Power Schemes: Portable/Laptop (So AMD C'n'Q will be effective. This will save me shit load on power consumption)
Monitor: Turn off in 5 minutes (All are LCD + one connected to TV)
HDD: Turn off in 10 min
Standby: 15 min.
No Hibernation.

Pretty aggressive power savings. All my machines are in standby and pretty much consume may be around 10W/hr?

My Media PC also has the same settings, it wakes up for scheduled recordings.

My 50" Sony TV and Sony DVD player are the ones getting beaten up now a days as my 2 yr old son watches his favourite and if somebody turns the tv off if he is doing something else, he will run back and switch it on. Its pretty much burning electricity all day. Thats the worst consumer in my home.:cry:

[Graogrim]

That isn't really a good example, it depends on the load. A 4 cylinder will outperform the 6 in terms of fuel efficiency going up hill with lighter loads. The same is true about different power supplies.

I think the moral of this story is to make sure that your power supply is well-matched to the load your system generates. Don't aim too low or too high. There are tools such as this (http://www.extreme.outervision.com/psucalculator.jsp) to help estimate power loads. Generate a range of power estimates from idle to max and then find a power supply that has maximum efficiency in that range.

Understand, though, that this is like buying a hybrid car simply to save on gas costs...the vehicle will have to have an exceptionally long life just to recoup the extra initial outlay. I'd suggest doing some math and figuring out exactly how much of a difference a slight (say 5-10%) efficiency increase will mean to your electric costs. If you can't make that money back within the expected lifetime of your system then it's not likely worth the bother.

[Laurijan]

This is what I do. All my 4 PCs have the same Power settings
Power Schemes: Portable/Laptop (So AMD C'n'Q will be effective. This will save me shit load on power consumption)
Monitor: Turn off in 5 minutes (All are LCD + one connected to TV)
HDD: Turn off in 10 min
Standby: 15 min.
No Hibernation.

Pretty aggressive power savings. All my machines are in standby and pretty much consume may be around 10W/hr?

My Media PC also has the same settings, it wakes up for scheduled recordings.

My 50" Sony TV and Sony DVD player are the ones getting beaten up now a days as my 2 yr old son watches his favourite and if somebody turns the tv off if he is doing something else, he will run back and switch it on. Its pretty much burning electricity all day. Thats the worst consumer in my home.:cry:

Good tips! Thx the HDD of my rig are password protected if they turn off the password kicks in and i have to reboot but otherwise your settings are mine from this moment on

[keakar]

Funny avatar you got!

The measurement was taken from the computer + a printer in standby + a connected tv-stick (very hot) + a connected webcam + a backlights having keyboard + a mouse + the true flat trinitron CRT.

this concerns me, if it is very hot then you are drawing too much power through it or it is defective and overheating. you should look at fixing this before it gets worse and you have it start melting on you.

electrical plugs, outlets, wires, or connections should never be hot, a little warm is ok but never hot.

[largon]

I use the "Sleep" mode in Vista Start menu.
It would seem like it powers down everything except RAM as power consumption plummets to less than 20W measured from the wall outlet. Then just hit the power button and you're immediately back on desktop with everything up and running exactly like the moment you put it in sleep mode. And btw, a typical ATX PSU consumes 10-15W even when it's not actually turned on. I mean just as long as it's plugged in an wall outlet and the I/O switch on the unit is @ I.This is what I do. All my 4 PCs have the same Power settings
Power Schemes: Portable/Laptop (So AMD C'n'Q will be effective. This will save me shit load on power consumption)
Monitor: Turn off in 5 minutes (All are LCD + one connected to TV)
HDD: Turn off in 10 min
Standby: 15 min.
No Hibernation.When setting HDDs to power down after such a short time as 10 minutes of idling keep in mind powering up stresses the motor more than continuous spinning. Pretty aggressive power savings. All my machines are in standby and pretty much consume may be around 10W/hr?I gather you mean 10kWh a day?

[suraswami]

When setting HDDs to power down after such a short time as 10 minutes of idling keep in mind powering up stresses the motor more than continuous spinning. I gather you mean 10kWh a day?

10KWH a day - don't know.

I have more than one HDD on most of my machines. So powering down the HDDs thats not used is better than spinning them constantly. Ofcourse the OS HDD is going to be spinning continously until S3 state kicks in.

[Laurijan]

Even if the power costs only 0.046€ at daytime and 0.036€ at nighttime.. i cosume about 5KWH a day which is 5x12x0.046+5x12x0.036 which is like 5€ a day.. Am i confused or did i calculate right.

[largon]

It goes like this:

Daytime (07:00-22:00):
kWh × €/kWh × (daytime/day)
5kWh × 0.046€/kWh × 0.625
= 0.14375€

Nigthtime (22:00-07:00):
kWh × €/kWh × (nighttime/day)
5kWh × 0.036€/kWh × 0.375
= 0.08625€

Total:
0.14375€ + 0.08625€
= 0.23€ a day

[Mussels]

I use the "Sleep" mode in Vista Start menu.
It would seem like it powers down everything except RAM as power consumption plummets to less than 20W measured from the wall outlet. Then just hit the power button and you're immediately back on desktop with everything up and running exactly like the moment you put it in sleep mode. And btw, a typical ATX PSU consumes 10-15W even when it's not actually turned on. I mean just as long as it's plugged in an wall outlet and the I/O switch on the unit is @ I.When setting HDDs to power down after such a short time as 10 minutes of idling keep in mind powering up stresses the motor more than continuous spinning. I gather you mean 10kWh a day?

vistas sleep mode depends on your motherboard - it has to support intel ViiV or AMD AWAY. its also called S4 standby in one of my bioses.

I too use a wall meter, and thats why i have so many PC's now - its cheaper to say, have my 8x hard drives in my storage system turned off 90% of the time, than have em on gaming/web browsing. :)

[Laurijan]

It goes like this:

Daytime (07:00-22:00):
kWh × €/kWh × (daytime/day)
5kWh × 0.046€/kWh × 0.625
= 0.14375€

Nigthtime (22:00-07:00):
kWh × €/kWh × (nighttime/day)
5kWh × 0.036€/kWh × 0.375
= 0.08625€

Total:
0.14375€ + 0.08625€
= 0.23€ a day

Ok i am a bit confused then because i went to a higher mathematical technical school for 7 years.

[Sasqui]

I had a brain fart, my Kill-A-Watt (Model=P3) does indeed measure KWH!!! :toast: I didn't see it before. So I'm up to 0.03 KWH after about 10 min.

I'm measuring ONLY my CPU, been itching to put it into the main (monitor, sound, printer), but thats :rolleyes: down by the floor and a tangled mess.

The CPU draws about 250w idle and around 360w gaming. If I run the automated clock utility in CCC, it goes up to 380+!

Now this is interesting: If I go into sleep mode, the CPU draw does not change! Anyone know what's up with that???

[largon]

Sasqui,
Dunno what's going on but 250W idle appears somewhat high for a 3.5GHz E6600 + HD2900PRO compared to for example my 3.6GHz Q6600 + 8800GTS 512MB (800/2000/1050) combo which draws 170W when on Vista desktop and 300W while playing Crysis. Load can be explained by your video card but isn't HD2900PRO supposed to be on par to 8800GTS 512MB (http://www.techpowerup.com/reviews/Zotac/GeForce_8800_GTS_512_MB/23.html) @ idle?

[Sasqui]

Sasqui,
Dunno what's going on but 250W idle appears somewhat high for a 3.5GHz E6600 + HD2900PRO compared to for example my 3.6GHz Q6600 + 8800GTS 512MB (800/2000/1050) combo which draws 170W when on Vista desktop and 300W while playing Crysis. Load can be explained by your video card but isn't HD2900PRO supposed to be on par to 8800GTS 512MB (http://www.techpowerup.com/reviews/Zotac/GeForce_8800_GTS_512_MB/23.html) @ idle?

Good question... the clocks on the 2900 are at XT speeds, but only during gaming. There isn't anything else signigifcant to explain the difference. I keep the rig OFF when not in use for the power savings - even if it is a fraction of my 'fridge :)

[swaaye]

You can save a lot by having the CPU power management set up. Cool'n'Quiet and Speedstep/C1E. These drop both CPU clock and voltage and it really reduces the power usage. It also helps to have an efficient video card like a GF8600 (or lower) or Radeon 38x0. 8800GTX pulls like 70W doing nothing, for example.

My Radeon 3850 draws as much at idle as the 8600GT I used to use. That's fairly amazing to say the least.

Hard drives are nothing. They each use like 8W or so when idling. Fans use even less (probably 2-3W unless you have a 3000 RPM 120mm hair drier or somethin).