Is the wattage meter faulty or what?

M76

M76

[H]F Junkie
Joined
Jun 12, 2012
Messages
14,039
I connected a meter to the rig in my sig, and I was shocked out of my mind.

It showed 450 W at idle, and 1600 W at load. Is that possible or is the meter faulty? I mean it can't be, since I only have a 1000 W PSU, and I have no stability issues or sudden reboots, so the reading isn't right, right?
 
It doesn't sound to far off to me. I know when i put my old video card (amd 6870) in the power usage tripled while sitting at the desktop. Having 2 in crossfire may make it 6x. Also the important thing to note here is you are measuring the power from the wall and the power supply is measuring the power it is able to put out. Not sure if you included your monitor in the power usage or just the computer.

I haven't put a power meter on my cpu lately but the setup is a i7-2600, amd 7870, 2 ssd and a mechanical hd. I know the 3 lcd took about 100w each to run and if i remember the computer was taking about 100 at idle with the old video card. It has a corsair cx 600 power supply for efficiency reference. If i remember I'll post my power usage after i get back from church this afternoon.
 
I know that there is a difference between what the psu can put out and what it draws. But it's supposed to be a 80+ PSU, so it should only draw 1200W at maximum capacity. The display is included but it only uses around 30W. So that leaves at least 350W unaccounted for, assuming that the PSU is at 100% load, which I highly doubt.

Edit: I tested the same system, before I installed the 2nd 290x, and the CPU was at default, so no OC, and it showed 600W under load. There is no way the OC and the second GPU draws another 1000W.
 
Last edited:
It's a faulty meter. No way in hell would a 1kW PSU be stable at 1.6kW.
 
M76 said:
I connected a meter to the rig in my sig, and I was shocked out of my mind.

It showed 450 W at idle, and 1600 W at load. Is that possible or is the meter faulty? I mean it can't be, since I only have a 1000 W PSU, and I have no stability issues or sudden reboots, so the reading isn't right, right?
Click to expand...

And the power supply is?

But in general, no.

Nasgul said:
Sure looks that way. I have one also and it's right on the dot with the LINK software. Time to get a new meter......eBay, $14.99.
Click to expand...

Matching up to the link software doesn't mean much as neither the power meter you are using nor the software is accurate or reliable ;)
 
you need to plug the computer cord directly into the meter (directly into the wall). not the power strip with monitor, speakers, etc. do just the tower, and tell us what you find.
 
[H]OCP has said that the Kill-a-Watt meters and other cheap power meters do not play well with APFC PSUs.
 
Paul_Johnson said:
And the power supply is?
Click to expand...
Lepa Maxgold G1000-MB

I guess it could be caused by the APFC, since I think when I did the last measurement I still had my old psu, which was many years old, and nowhere near the efficiency of current ones. But it was APFC too, Chieftec CFT-750-14CS
 
Nasgul said:
Surprisingly enough, when I was testing my RM 1000i, the meter said 86W, in the LINK it say roughly the same, within a 3-5 watt difference. So how can the software and hardware be nor accurate or reliable, at the same time?
And there was fluctuation in the software and meter, but not by thousand watts like M76's.
BTW, I tested the meter with other appliances, pretty accurate I'd say.
Click to expand...

Very easily. A broken clock is right twice a day. You are using a piece of hardware that has very low accuracy and software that pulls from a very cheap hardware (if it was actually an accurate and precise piece of hardware it would cost a lot more) as well. Corsair will, of course, argue that it is very accurate (as will people who don't know what they are dealing with) but that monitor is not nearly as accurate as a real power meter nor is it calibrated. I have tested a number of the software/hardware monitors out there and as yet none have been accurate and some were neither accurate nor precise. The way they work varies and a number of implementations don't actually read values at all under certain circumstances but use a lookup table to give a predicted value. As for your kill-a-watt the pictures are floating around of lots of incorrect readings that I and others have collected. In a couple weeks you'll get to see even more numbers.
 
Last edited:
What about one of these?
sencore.jpg


What about the software for the UPS, how accurate/inaccurate is that?

This is what the software says for my system. I was listening to music on iTunes and reading this thread when I took the snapshot.
Everything on my desk is plugged into the UPS, 4790K machine in my sig, LG 34" monitor, M-Audio BX5 speakers and BX 10" subwoofer, 7 port powered USB hub,
2TB USB3 External drive, WD 8 port gigabit switch, HP DV7 laptop (powered off but plugged in), 36" fluorescent light, desk thermometer/clock.
apc-backup-4790k.jpg


And this is the UPS on the 3770K setup, just the tower, 2 cable modems, router, and switch are powered on,
apc-backup-3770k.jpg
 
Last edited:
M76 said:
Lepa Maxgold G1000-MB

I guess it could be caused by the APFC, since I think when I did the last measurement I still had my old psu, which was many years old, and nowhere near the efficiency of current ones. But it was APFC too, Chieftec CFT-750-14CS
Click to expand...

Unfortunately, I have not reviewed that unit so I don't know for sure if it has any funky harmonics, etc. However, I know the platform and it is not that inefficient at full load. So, if just the power supply is plugged in it is the meter.
 
Zepher said:
What about one of these?
sencore.jpg
Click to expand...

Very old (by tech standards don't tell BillR) and unless it has been serviced and calibrated it is not nearly as accurate as newer test gear not least of which because the read out is actually stepped instead of continuous as it appears. It also has the issue of only being as accurate as it is going to be at a fixed voltage IIRC. Also, if you hang around old audio stuff the people that used them used to complain about diodes on them blowing a lot so longevity and functionality along with accuracy and precision if nothing has been done to it would likely be suspect. It's like old variacs. Sure they were built better than the new cheap ones (not necessarily better than the new good quality ones), but if you did no service on them like the brushes they are about as useful as a boat anchor (and likely heavy enough to be one).
 
Hey Paul, I updated my post above asking about the UPS software, not sure if you noticed it.
And that old Sencore was given to me, I don't even know if I still have it, may have tossed it when I redid my room a few years ago.
 
Zepher said:
Hey Paul, I updated my post above asking about the UPS software, not sure if you noticed it.
And that old Sencore was given to me, I don't even know if I still have it, may have tossed it when I redid my room a few years ago.
Click to expand...

I would have to have the UPS, the software, and disassemble the UPS to know how it was being done. That said, accurate power meters cost more than most consumer UPS's so......
 
Paul_Johnson said:
I would have to have the UPS, the software, and disassemble the UPS to know how it was being done. That said, accurate power meters cost more than most consumer UPS's so......
Click to expand...

maybe you can post a link to the meter just for educational reasons
 
primetime said:
maybe you can post a link to the meter just for educational reasons
Click to expand...

? There are tons. You can spend $1,000 or you can spend $10,000+. Is it a basic single channel power meter, or a more capable analyzer, or does it have a built AC source? If it has the AC source what capacity, etc. If you want one then HP/Agilent/Keysight, Chroma, Yokogawa, Fluke, Manta, etc make quality ones you'll just have to poke around until you fine what you want then request a quote. Or you can look at the second hand market (but it's always somewhat dicey) and you'll find them from nothing on ebay to $2000 to $30000. You can also pick up any number of Taiwanese or Chinese "brands" but even the good ones there will still be more than a UPS. The ones we have in the family shop are Chroma 6620(x) and there is a Yokogawa.
 
Last edited:
You must log in or register to reply here.
Back
Top