Going green

When not to use CFLs

Letter the other week in The Age:

Not worth the cost

AM I the only person having trouble with expensive ”green globes” alleged to last 10,000 hours? Used eight hours a day, a globe should last more than three years. I have replaced the globe in one lamp four times already this year. I want my cheap, long-lasting, environmentally unfriendly globes back.

Leone Garro, Northcote

My CFLs are lasting ages… provided they’re in the right places.

I would bet the globes referred-to above are switched on for short durations (less than 15 minutes), many times a day. That kind of usage is bad for CFLs, and it’s precisely why I’ve avoided using one in the toilet, for instance, and also in the bedrooms — our particular usage there seems to predominantly involve ducking in to get something/drop something, then out again.

For those types of spots, it’s far better to stick to non-CFLs, such as the energy-saving incandescents still available. (My local supermarket has the Philips EcoClassic products, which for instance provide 100w of light but burn 70w of power.)

Evidently this message isn’t isn’t getting through.

Or possibly Leone’s light fitting or wiring is faulty, but I’m betting it’s the former.

(Previous blog post on this topic)

By Daniel Bowen

Transport blogger / campaigner and spokesperson for the Public Transport Users Association / professional geek.
Bunurong land, Melbourne, Australia.
Opinions on this blog are all mine.

15 replies on “When not to use CFLs”

While i agree in principal with your situation I think your major problem is in the quality of lamp you purchase. CFL lifespans are based around the quality of it’s ballast. The cheaper globes use cheap quality parts and hence never reach their intended life if used in areas such as toilets. Personally I have decked my entire home with Phillip tornado lamps. The ballast is of very quality. The lamp itself feels very solid and so far I have had no issues… Additionally 4 years ago I replaced my parents only remaining incadescent globes (toilets, laundry and garage) with CFL’s they are all still going strong… Lamp price was $8 each back then, they have reduced since then obviously, and you can now get the Phillip tornados for $6 each.

One last thing to note is that the overall life cost including capital cost of the globe is still less than incandescent, even if your only globe lasts 5000 hours

The problem of short lasting CFL’s may also be the voltage level supplied to the house, which depends on how close you are to the street transformer. If you live close to the transformer (as I do) then the voltage may be up to 7% above nominal (i.e. up to 256V) – the higher the voltage the more stress placed on the lamps and the quicker they fail.

What Lachlan says tallies with my experience. There is a transformer across the street, and our power comes from that. I’ve had terrible luck with CFLs, all good brands (or so I thought), treating them how they are supposed to be, and many only lasting two or three years. Both hallway lights went out within a few days of each other just last week. Occasionally lamps will “die” then come back again, only to eventually die properly.

@enno: you won’t be overcharged. The old mechanical ones use a setup with coils inducing eddy currents in an aluminium disc to integrate the product of voltage and current. Electronic ones use dual-slope integration. Either way, it’s going to calculate energy usage correctly.

Higher voltage will cause lamps (incandescent as well as fluorescent) and toasters to fail quicker, and will cause appliances that present a resistive load to use more power (e.g. incandescent lamps, heaters, kettles and electric stoves). If it’s a heater with a thermostat or an automatic it won’t really make a difference, as it will just hit the cutout point faster. “Universal” motors (e.g. in your power drill) may consume more power, but induction motors will generally just operate with a poorer power factor (more current, but no more real power). Electronic devices with regulated supplies may waste more power as heat depending on the design of the power supply.

Incandescent lamps produce a whiter (less yellow) colour at higher voltage, too. You can calculate how much more power your incandescent lamps will use – it’s roughly proportional to the square of the voltage. According to AS60038-2000 the voltage must be 230V +10% -6% true RMS, so it could range from 216V to 253V before you have the right to complain (the old AS2926-1987 was 240V). A lamp rated at 100W on 230V would consume about 121W at 253V.

Sorry for the rant, but I hate to see blatant misinformation being spread, and I have a degree in EE, so I may as well use it for something.

The meter will calculate the energy flowing into your house correctly. But if the voltage is too high, you will consume more energy, some of which may be useful but a lot of which isn’t useful.

Devices like electric heaters and light bulbs will use more power, which you will have to pay for even if the extra heat or light isn’t beneficial.

Electronic devices like a computer which have a regulated power supply will waste more power. For example if the power supply of your computer reduces the voltage to 6V and then regulates it to 5V exactly, you are wasting power equivalent to that 1V difference. If your supply voltage is too high, the transformer output is , say 6.5V and still getting regulated to 5V. The result is the wasted power is equivalent to 1.5V which is a 50% increase in wastage at that point.

@enno: you don’t know how a switch-mode power supply (SMPS) works. The SMPS in a computer rectifies and filters the mains directly, then passes it through a high-frequency inverter, steps it down with a small ferrite core transformer rectifies and filters the output. The voltage is regulated by varying the duty cycle of the inverter so as to keep the logic voltages accurate, while the other voltages are derived by turns ratio on the transformer (a modern PC will get +12V, +5V, +3.3V and -12V rails from the PSU – regulation the +5V and +3.3V rails is critical, but the 12V rails can vary quite a bit). Varying the supply voltage (within reason) won’t make it use more power.

What you described is a linear regulated power supply. They not very common any more as SMPS have become very cheap and efficient over the last couple of decades. They are only used in devices with relatively low power consumption anyway, as they generate a relatively large amount of heat. A linear regulated power supply for a modern computer would generate too much heat to be of practical use.

Almost all heaters have room thermostats these days – even my cheap Kambrook fan heater has one. At higher voltage it will cut out quicker. Modern kettles cut out when the water boils, so they’re the same deal. Electric ovens have thermostats. Electric irons have thermostats. Even electric cooktops have crude temperature regulation mechanisms.

If your incandescent globes are running brighter, you can buy ones rated at lower power. They’ll still fail quicker, so you’re out of luck there, but you can enjoy the better colour temperature. And my pick for CFLs is E-Star Daylight – excellent colour temperature and quick startup.

You’d better stop arguing before you run out of cred to burn – I’m not kidding about having an engineering degree with first class honours. ;)

Well you have agreed that your appliances are going to draw more power at higher voltages, which you are going to have to pay for. Which is what I said.

And I am sure that Bunnings will sell me a 93.742 Watt light bulb to enable me to compensate for the fact that a regular 100 Watt will use too much power at an elevated supply voltage. And then exchange it for a 95.23425 Watt bulb when the voltage wanders around some more.

As for the details of exactly how the modern computer’s power supply works, I will take your word for it, it is not my current field. When I hand-built one of the first personal computers in the country, things worked a bit differently then.

And I will see you First Class Honours degree in Electrical Engineering, and raise you a University Medal, two Masters and a PhD in Mechanical Engineering.

Yeah, 8-bit computers often had linear regulators. My Pioneer PX-7 does, and my Commodore 64 did as well. But they didn’t draw much power compared to the monsters of today. The Commodore 64 could supply 25W on the 5V rail, and would be wasting about 7W as heat in the linear regulator. That’s actually very inefficient in percentage terms. Modern PC PSUs can often deliver over 500W – just think how well they would work as fan heaters if they used linear regulators. Personal computers were the main contributing factor to driving down the price of SMPS.

Yes, I agree that some appliances will use more power. However, you will not be “overcharged” by any definition of the word, and for most appliances that actually use more power, it won’t make any difference because they use thermostats or some other kind or regulation or cutout mechanism.

Lights are the one exception, but IMO the increased power consumption isn’t the biggest issue – reduced lamp life is far more troublesome (but I do like the whiter light). If you’re having real issues with this, there are two things you can do. You could get it tested to see if it’s out of spec, but I don’t know how much luck you’d have trying to convince the supplier to do something about it if it is. The more practical and more expensive option is to install an autotransformer or variac in your lighting circuit to reduce the voltage to a more suitable level. Don’t try this on your own – get a competent electrician to do it, and try not to scream when you see how much it will cost.

I started doing a PhD in wireless telecommunications, but I realised early on that it was all about the university’s quest to sneak possibly patent-worthy IP into IEEE standards while doing the bare minimum to satisfy the requirements of a PhD (you’ve probably heard about CSIRO’s recent victory using this technique). I figured if I was doing commercialisation, I may as well be paid properly for it and quit. I’m pretty disillusioned with the whole scene, and with the way higher education is going in general in this country.

Voltage coming from the power lines in the street will not only vary with the distance from the transformer but it can also vary somewhat throughout the day as the load on the system increases or decreases.

My condo building where I used to live in the US would buy light bulbs in bulk for the common areas that were rated at 130 volts as these were supposed to last longer with somewhat less brightness at the normal 117 to 120 volts of US power.

I have used CFL bulbs for years in the US and in Australia and find that some will last years and some just burn out sooner. I had a light fixture in my bathroom in the US that took 8 bulbs. I replaced them with CFLs all at once and I found that most of them lasted many years while 1 or 2 burned out much sooner.

Might I just say I love seeing blog commentary with a proper, educated, nerd debate. See what sort of riff-raff you attract Daniel? Thanks enno/Vas – for a change, I felt my IQ *increase* reading something in a blog. :-)

Might I just say I love seeing blog commentary with a proper, educated, nerd debate.

My reaction was the exact opposite. Reading academics squabbles is a complete snorefest.

*unzips pants*

Yeah, and my phd is bigger then yours too.

Where can I get an LED bulb? I’d like to know when it comes time to replace the CFLs in high switching situations.

Most affordable current generation LED globes produce rather harsh, unnatural light. That said, there was a place I used to walk past on the way home from work on the north side of Spencer St, around where it mingles with Miller St, that had a sign up saying they specialise in LED lighting, but I can’t remember the name or exact address. That was a year ago – it might not be there any more.

Comments are closed.