Zalman CPU & VGA Power Consumption Meter

[MikeC]

http://www.silentpcreview.com/Zalman_CP ... ower_Meter

[scdr]

Wonder how much power the meter uses itself.

[MikeC]

Wonder how much power the meter uses itself.

Not enough to be significant. Typically such devices pull at max maybe a watt.

[frenchie]

Good article. But I'm missing something here. What's the use of such a device for amateur silencers like most of us ? I see the point for reviewers, but other than that...
Can someone explain why you'd want to constantly display the power your CPU or your GPU use at any given time ? It looks "modern and cool" I guess, that's about all I can think of...

[Monkeh16]

The device seems entirely pointless to me. It's incapable of measuring the full power draw of the GPU, and I suspect there are boards (other than the mentioned LGA1336 example) which draw power from the main power connector, making it incapable of measuring true power draw for the CPU also.

[lm]

I would rather put that money towards a clamp meter. Can anyone figure out a valid use case besides reviews such that it makes sense to invest in this rather than just an AC power meter for the whole system?

[Monkeh16]

I would rather put that money towards a clamp meter. Can anyone figure out a valid use case besides reviews such that it makes sense to invest in this rather than just an AC power meter for the whole system?

Geek cred? Oh, wait, it'd have to work for that.

[scdr]

Wonder how much power the meter uses itself.

Not enough to be significant. Typically such devices pull at max maybe a watt.

Reading the specs at Zalman I see that it only measures down to 10 watts, so I guess it isn't much use for measuring idle power on the more efficient CPUs.

http://www.lostcircuits.com/mambo//inde ... mitstart=4

Does it continue the same accuracy (circa 1%) down to low loads (e.g. at 10 watts)? Since the meter itself is drawing power from the same source it is measuring, wonder if they corrected the reading for the meter draw, or just aimed it at heavy load conditions (where the meter draw would be swamped by the load).

[MikeC]

The meter's lowest readout is always 10W, I can confirm this. So it's really not quite as good as a separate shunt resistor + accurate low voltage multimeter, which allows much lower readings. But as mentioned in the article, such meters are not cheap. Ours cost several hundred $$. Not sure how much of a limitation this really is, though. I have to admit that we've never tried to get low power AUX12V readings at idle before. AC power seems much more important & useful.

As for the overall usefulness of this meter, for most folks it's low. Among SPCR readers, I suspect it would be higher.

[MoJo]

You could get a similarly accurate reading just by measuring at the wall when the computer is idle and under load with an ordinary power meter. Since you know the minimum power consumption of both the CPU and the GPU thanks to sites like SPCR you can then estimate the load consumption pretty well.