[Nut-upsdev] Re: CyberPower 685AVR and newhidups

[Charles Lepple]

On 10/31/05, Peter Selinger <selinger at mathstat.dal.ca> wrote:
> Charles Lepple wrote:
[...]
> > Peter, how are you calculating VA?
> >
> > VA really isn't going to be much help in the EU where power factor
> > correction is required (so VA and watts will be equal). However, we
> > only have a variable for "power", and often UPS manufacturers don't
> > distinguish between VA and apparent power.
>
> Unfortunately, NUT specifies that the unit of the "power" variable is
> VA, not W. My understanding (partly from the USB HID Power Devices
> spec) is that W is true power, whereas VA is "peak V" times "peak A".
>
> I have converted W to VA by multiplying by sqrt(2).
> So 390 * 1.4142136 = 552.
>
> Perhaps I should round this number to the closest 5, 10, or 25?

well... you can have peak-to-peak VA, and RMS VA.

You're correct that the "W" unit is for true power, and VA is voltage
times current, though.

For a light bulb, since it is a purely resistive device (for all
practical purposes), your voltage and current are in phase. I forget
whether this is "power factor 0" or "power factor 1" or whatever, but
it's a phase angle of 0 either way.

The problem is that some devices (most notably, motors and switching
power supplies) are more inductive than resistive, so their phase
angle approaches 90 degrees. The wattage may be low, because the
current and voltage are not in phase, but the RMS current times the
RMS voltage is higher than the RMS value of (instantaneous voltage *
current).

> The manufacturer indeed specifies 390W and also 685VA (see
> http://www.cyberpowersystems.com/CP685avr.asp); this does not make any
> sense to me. Certainly, we don't get any report from the UPS involving
> the number 685. In fact, the UPS claims it is a "BF700", not a
> "685AVR" or "BF685"; however, no such product exists on the web.
> Probably "BF700" was their original name, and some marketing types
> decided at the last minute that "BF685" sounds sexier, but they forgot
> to change the name on the chip.
>
> Since 390W is the lower of the two values, I assume it is probably the
> correct one. "Watts" is a well-defined and measurable unit, whereas a
> "VA rating" could be something from the manufacturer's dreams. So
> probably their "VA rating" is an exaggeration.

Basically, it's a question of "how much inductive energy can circulate
in the UPS?", so it is possible to have a higher VA rating than
maximum wattage. Wattage is limited by the amount of heat that you can
remove from the electronics, and VA is limited by the breakdown
voltage and maximum current of the power devices.

> In any case, none of these variables have any bearing on the
> functionality of NUT; they are just provided for information, along
> with the device name, manufacturer, and serial number.

Right.

I vote for no conversion; if we need a variable suffix to indicate the
quantity being measured (if not VA), then at least we won't be mixing
up VA and Watts when displaying values.

--
- Charles Lepple