Data on power draw?

[rogerh]

Can't do better than 1 without going into the domain of complex numbers.

Perhaps it's what you get with a simple PF meter when the current leads the voltage, that is when the PSU is presenting a capacitative rather then an inductive load.

 

[Ericloewe]

Perhaps it's what you get with a simple PF meter when the current leads the voltage, that is when the PSU is presenting a capacitative rather then an inductive load.

What kind of measurement system would cause that?

 

[rogerh]

What kind of measurement system would cause that?

V. brief googling suggests a polarised vane type of meter could do it, but it should probably be calibrated as .98 (perhaps with a minus sign) rather than 1.02.
Edit: but more likely it is just a digital type meter coping badly with non-sinusoidal currents.

 

[MMacD]

Total power draw depends on a lot of factors.

Yes, I know. But I was thinking of the basic case, the one with which one does acceptance testing or debugging: mb, cp, cp fan, mem.

We have draw data for every one of those components except the motherboard. We have TDP for the cp, which I suppose is at least standardised within vendors even tho not between. Most memory vendors provide power info for their DIMMs. Digikey provide wattage information for their 92mm Delta fans, and fan tech is very stable so there's probably not a lot of between-vendor differences.

So the only thing we don't know is the mamaboard draw, and since motherboards differ greatly in features, it's reasonable to think they probably differ a lot in power consumption too. But maybe they don't! Maybe the difference is only a few watts regardless of the difference in features. I'd bet it isn't, but it'd be very nice for planning purposes to know one way or the other.

 

[cyberjock]

Sir...a PF of 1.02 doesn't even make mathematical sense, since it is the sine of an angle.

Sir.

You are correct. Used to Navy terms. For me it would be 0.98 lagging to 0.98 leading. Basically extremely close to unity.