CFL consumption ?

[Nick]

I recently got a ac current and power meter and something has become apparent which makes me wonder if its significantly out of calibration... or just the type of load.

It seems to measure pure resistive and most other loads accurately, including those with SMPS like the TV / PC etc.

But try and measure a CFL or indeed any Fluorescent light and it is telling me it is consuming about double the power it should be.

- The 60 watt circular kitchen lights each consume around 120 watts

- The dual 58 watt (116 watts) fluorescent lights takes approx 170 watts

- the 18 watt CFLs take approx 34 watts In fact the Megaman 11 watt CFL has printed on it 96 mA which is 230 x 0.096 = 22.08 watts

Can anyone explain ?

Is the power not going into the tube not counted ? ( i.e what the ballast wastes not counted ?)

Thanks,

Nick

[meow2222]

Its reading VA, not power. Google "power factor"

NT

[<me9]

On AC the volts x amps is not necessarily the power consumed. Some of the amps are what is known as reactive. these do not contribute to the consumption of the device but 'circulate' and are present in the distribution system.

If it's like my device it will also measure power factor, which when multiplied by volts times amps gives the true power on which you are metered.

The figures you are quoting will be VA.

[Fred]

What's the model, have you looked at the spec regarding power factor. Unfortunately fluorescents don't always have very good power factor, even if they have a correcting capacitor. CFLs don't have the space.

[The Natural Philosopher]

'power factor'

[cynic]

Any flourescent light has a power loss in the choke/control gear. If you check in the OSG for a rule of thumb this is normally taken as not less than 80% of the rated power of the lamp thus a 100 watt lamp fitting will consume 180 watts of power or more. Your readings are entirely ok (For those with a copy of the 17th edition OSG its note 2 on page 96 dealing with assumed current demand)

[meow2222]

You'd be very hard pressed to find a ballast that bad.

NT

[John Rumm]

You seem to be confusing the current demand, and power dissipated in the load. The maximum circuit current may well be in excess of that expected by inspection of the lamp power alone, however not all of that additional current flow corresponds to matching power dissipation (and hence energy consumption) in the load.

That note is describing how to estimate the current demands on the circuit. i.e. if you have 1kW of strip lighting, then you might assume that a 6A MCB would be appropriate protection for the circuit. If that lighting was incandescent, then you would be right. However in the case of discharge lighting with poor power factors, you would actually need to design for getting on for 8A of load (i.e. protecting the circuit at

10A typically) even though the actual load is only 1kW plus a bit for thermal losses in the ballast.

[<me9]

VA will be 180, for cable sizing. power consumption(W) will be much nearer (but above) the 100 watts, 120 perhaps max.

[cynic]

I do know the difference and very much doubt the el-cheapo meter readout is actual watts power. It is likely to be reading va and the readings mentioned fit reasonably well with that. Maybe I should have given a fuller description :-(

[John Rumm]

I have one of the plug in meters somewhere (the one Maplin flogged for a while)... IIRC the one I have does make a stab at assessing power factor, however it is probably not much good on things that have non sinusoidal current waveforms like CFLs. It seems ok on basic reactive loads with straightforward leads or lags though.

(spose I ought to dig it out and work out why this house is so much more expensive than the last one for electricity use)

[Peter Andrews]

.

When my consumption jumped up I eventually, months later, discovered that the control knob on the chest freezer was broken and although it indicated the mid position the compressor was, in fact, running 24/7!!

Peter

[meow2222]

After a bit more thought I suspect the meter isnt reading VA after all. The reason is that the high current for the CFL doesnt tally with its wattage or VA rating. Since the current draw is in the high v part of the mains cycle for a CFL, it wont draw excess mean current compared to a resistive load. Rather it will draw higher peak current.

I suspect instead that like many a cheap digital meter its measuring peak current and doing a blind conversion to what will be power for resistive loads, but will be simply a wrong answer for other stuff.

NT

[<me9]

It will also be effectively delayed, and have a high harmonic content, both affecting the power factor.