Calculators VS Kill-A-Watt --REAL Power Consumption

S

sir_rob

Limp Gawd
Joined
Jan 27, 2003
Messages
135
Question: When I upgrade, do I really need to get another 650+ watt PSU to compensate for inaccurate manufacturing specifications?

I have used many calculators to give me an idea of how much power my computer is using. They all state that my system uses 500 to 650 watts (or they recommend PSU's that will supply that much power).

When I plugged my computer into my Kill-A-Watt to see how much the PSU was drawing from the outlet, it NEVER exceeded 200 watts under load from before it posted through some stress testing! I ran Prime95 on each core and transferred data from a DVD-RW to a HD at the same time. I assume that if I were ripping from all three DVD-RW drives, running prime 95 on each CPU core, and transferring files between all 5 hard drives...etc, I could get it over 200 watts, but probably not by too much.

Knowing that the PSU is not 100% efficient, that tells me that even with a large number of components (but low end video card), my system is pulling well under 200 watts from the PSU. Am I missing something here? I have gone through two 550+ watt PSU's on various builds of this computer over the past 5 years. I use a UPS that does not condition the line. I understand that PSU manufacturers are notorious for overstating power ratings. One of the PSU's that failed was cheap, and the other one was a decent Fortron.

My current system is an Athlon 3800 X2 that has been overclocked at times, but is not currently. I will probably be upgrading to a decent Intel quad core soon. Except for the motherboard and RAM, everything else will stay the same.

Again: Do I really need to get another 650+ watt PSU? Am I missing something here?

Specs:
Athlon 3800 X2
MSI K8N Neo4
2 Gb RAM
Radeon 1300 Pro
(5) 7200rpm hard drives
8 USB devices
(3) DVD+/-RW drives
(1) 120mm fan
(3) 92mm fans
(2) PCI cards
 
I'm not going to comment definitively, but I was running a OC'd E6600 with dual X1950's in Crossfire, 3 HDD's, DVD drive, 5 fans, X-Fi with front panel, VFD display all on a OCZ Powerstream 510W PSU. In fact, that same system is still running today without the crossfire setup and has had zero issues. I used that PSU for 4 years in many different builds and never had any problems at all.

Like I said...more of a FWIW.
 
sir_rob said:
Am I missing something here?
Click to expand...

Yes and a quick search would turn up this topic a million times over. Here is the recap:

1) APFC can fool Kill-A-Watts into giving you abnormally low readings (some times giving better than 100% efficiency)

2) Power supplies derate with temperature anywhere from 2w/c above a nominal rated at value to 10w/c.

3) Kill-A-Watt's and most power meters sample to slowly to catch transient loads (the Transient load from our tests is 117w and is COMPLETELY missed by Kill-A-Watts).

4) Power supplies last longer if you stay in the 40% to 60% range of their output.

5) power supplies are quieter if you stay in the 40% to 60% range of their output.

6) Power supplies are cooler if you stay in the 40% to 60% range of their output.
 
Thanks! That really helps! Would it make any difference if I plugged the computer into a UPS and then into the Kill-A-Watt or will it always be fooled?

What do you recommend for my next power supply (power range for a decent PSU)?
 
sir_rob said:
Thanks! That really helps! Would it make any difference if I plugged the computer into a UPS and then into the Kill-A-Watt or will it always be fooled?
Click to expand...

The power meters in UPS software are just as bad. You have to spend some change before you get anywhere near an accurate power meter when your PSU has APFC.

What do you recommend for my next power supply (power range for a decent PSU)?
Click to expand...

Depends on the system.
 
I will keep everything the same except for the CPU, RAM, and motherboard. I will go with a decent Intel quad-core, and fairly high-end motherboard, and 4 Gb RAM.
 
Paul_Johnson said:
1) APFC can fool Kill-A-Watts into giving you abnormally low readings (some times giving better than 100% efficiency)
...

3) Kill-A-Watt's and most power meters sample to slowly to catch transient loads (the Transient load from our tests is 117w and is COMPLETELY missed by Kill-A-Watts).
Click to expand...

So the power reading taken with the Kill-A-Watt in these articles:
http://hardocp.com/article.html?art=MTUzMSw5LCxoZW50aHVzaWFzdA==
http://hardocp.com/article.html?art=MTU0OSw4LCwxNjA=

could actually be higher than expected?
 
Paul_Johnson said:
3) Kill-A-Watt's and most power meters sample to slowly to catch transient loads (the Transient load from our tests is 117w and is COMPLETELY missed by Kill-A-Watts).
Click to expand...

So how does one accurately measure the power consumption of their computer if not with a Kill A Watt?
 
Paul_Johnson said:
The ES-800 worked fine with the Kill-A-Watt when I had it.
Click to expand...

Ahh, thank you very much for the clear up.

panfist said:
So how does one accurately measure the power consumption of their computer if not with a Kill A Watt?
Click to expand...

You need a digital power meter. IIRC, Paul as well OklahomaWolf (of jonnyguru.com) uses a Brand Electronics Digital Power meter to measure power consumptions.
 
panfist said:
So how does one accurately measure the power consumption of their computer if not with a Kill A Watt?
Click to expand...

Spend a lot of money on better equipment for AC. For DC a good high quality (expensive) clampmeter.
 
Also, don't forget. Even if you've got a 650w power supply, your computer is only taking 200 from it. It's not just 'losing' power, it's not even generating that extra 450 watts. A lot of people reccommend a higher end PS for the fact of futureproofing. i7 is a prime example, with some of their chips running at a 130w TDP.
 
But shouldn't APFC error only effect power that's actually being power-factor-corrected? meaning power drawn by inducting components and the like? which wouldn't really be all that much compared to other loads in the system?

I am probably wrong...
 
Hi, can any one recommend a good calculator out there (if any) from where i can calculate the power supply requirements of my system. Below are the specs
Intel Q6600 Quad Core
Corsair 2 GB Dominator (1066)
XFX 260 GTX Black Edition
Western Digital Sata 500 MB Hard Disk
Currently i have a coolermaster 500W power supply and am confused if i need more or is it ok. I intend to get CM 690 nvidia edition, will the fans on the casing cause an increase in power requirements and how much.
 
Non-PFC power supplies similarly will make a cheap meter like the KaW read too high. KaW isn't all that useful for computers.
 
panfist said:
But shouldn't APFC error only effect power that's actually being power-factor-corrected? meaning power drawn by inducting components and the like? which wouldn't really be all that much compared to other loads in the system?

I am probably wrong...
Click to expand...

An SMPS is a nonlinear load. A non-PFC corrected power supply will have a power factor of ~0.59 to 0.65.
 
Paul_Johnson said:
And an Extech Power Analyzer.
Click to expand...

Definitely agree with Paul
a223c6b3710f85df22e9377d6c4f7553.jpg
6c1da886822c67822bcf3679d04369fa.jpg
0e3a37aa85a14e359df74fa77eded3f6.jpg
 
lazers said:
Hi, can any one recommend a good calculator out there (if any) from where i can calculate the power supply requirements of my system. Below are the specs
Intel Q6600 Quad Core
Corsair 2 GB Dominator (1066)
XFX 260 GTX Black Edition
Western Digital Sata 500 MB Hard Disk
Currently i have a coolermaster 500W power supply and am confused if i need more or is it ok. I intend to get CM 690 nvidia edition, will the fans on the casing cause an increase in power requirements and how much.
Click to expand...

Which model CM 500W PSU do you have?
 
You must log in or register to reply here.
Back
Top