Bad Power Factor Measurement

[ENG2009]

I rented a Fluke 435 to measure the power consumption of some big power supplies we have (Input: 480VAC, 60 Amps). I logged the data on 3 different models of power supplies. On one of them the power factor was 0.71, but on the other two the results are meaningless - the power factor was around zero, it even went slightly negative. It really justs looks like noise around zero. The voltage and current readings seem fine. With a zero power factor, I can't really calculate my KW usage. Any idea why this would happen?

-Scott

 

[avogel]

What kind of load is it connected to? could it be something that could be distorting the waveform? digital meters can be flukey like that.

 

[PHovnanian]

avogel may have a point about distortion. While this meter should be able to handle high THD voltage and current measurements, I see no reference in its specifications as to how it calculates PF, real and reactive power given non sinusoidal I and V inputs.

 

[ENG2009]

It's connected to a power supply about the size of a whole server rack. The power supply converts 480 VAC to 15 VDC 800 amps. I measured 3 different power supplies and it worked fine on the first. I don't think the other two are so different the meter would have a problem. I don't have the meter any more, so I can't redo it.

--Scott

 

[jghrist]

Were the power supplies loaded?

 

[ENG2009]

Yes.

 

[PHovnanian]

If the supplies are similar (i.e. all SMPS with similar characteristics or all transformer/rectifier/big cap) then I'd expect the readings to be close. If one is a switcher with unity p.f. correction and the other two not, this could expose limitations in the meter's capabilities.

The only other thing to have checked (but if the meter is gone, its too late) is the possibility of an error in connecting the 3 Voltage and 3 current inputs. Getting two phases swapped or a CT connected backwards couuld account for weird settings.