Should I switch lights off for short periods

[Racerbear02]

Sorry if this has been answered already but I did a search and couldn't find any results.

It was always given as a reason for office block lights to be left on all night that it was cheaper than switching them off and then re starting them due to the starters using more energy to start the tubes than was used overnight.

1. Was this ever true?

2. With modern fluorescents and starters is this still true?

3. Does this apply to low energy bulbs as well?

4. If any of the above IS true, what is the time below which it is cheaper to leave the lights on than re start?

[Ivan]

I've no idea why this fallacy has perpetuated so well.

No, there is no truth in it. Probably just used as an excuse for laziness.

Fluorescent tube starters run current through heaters in the end of the tube for a few seconds on startup. The amount of power used is approx three times that of continuous running. So if it takes a tube 5seconds to start up, then it consumes the same amount of electricity as it would in running for 15seconds. So if you are leaving the room for 15seconds or less, then you should leave the lights on!! If you are going for more than 15seconds, then you will be saving energy if you switch the lights off.

There are some other factors -eg if you switch fluorescents on/off very frequently, they will fail sooner, but not dramatically so. I've been told that using compacy fluorescents in PIR detector lights is a bad idea as the fail early due to the constant on/off cycling, however, I have one that was installed about 5 years ago, and it is still working fine.

Incidentally, there is a similar fallacy about computers - there was an argument for leaving computers on all the time, rather than switching off when not in use. The reason, apparently was that over a period (probably a decade or more), the ICs would 'walk' out of their sockets due to the expansion/contraction due to changes in temperature, eventually losing electrical contact with the socket. I've never seen this, and I don't think many computers have chips in sockets these days, anyway, as they tend to be surface-mount.


Ivan

[wyleu]

The IC walking effect is genuine. We had several large picture bending engines housed in 19 inch racks with 25 or so vertical circuit boards of about 2/3 of a metre high. They contained several racks of IC's in sockets and it was common practice to push the chips back into the sockets as the first element of any fix. ( there wern't many they were prety good boxes) They were left on all the time and it tended o be that once they had been turned off the chips crept until you got it all stable again when it would run continually. Probably more likely that the thermal expansions and contractions of the boards produced the problems but they were pretty inconsequential and made for an easy reputation as an 'Engineering Genius' by saving some advertising  edit session in 5 minutes flat. I wouldn't be surprised if this sort of fixing techinque ( the boxes consummed 3.5 Kwh's each ) dribbled down from these sorts of environments to PC's. The first stage of any fix I do is to just check the chips are located properly in sockets, and you do get some movement, but I would'nt put it down to one particular cause. This effect and saving the cost of a socket is probably why you don't see many socketed boards nowadays.

By the way the picture bending effects of these engines are now available on your average PC graphics card. The only difference is the D/A converter filters are probably not as good, which is why HDMI has better picture quality, if one is willing to ignore the Digital Rights management handcuffs.

[Capcave]

The main dis-advantage I have found with using CFL's in PIR detectors is the slow warm up time especially as PIR's are usually outside. I usually use a normal bulb set a low time out so you get the light instantly but it does not stay on long. Where I have used a CFL they seem to last as long as others not in a PIR.

[Ted]

As Ivan says it's a complete fallacy. I was badgered by my wife into testing the latest batch of CFLs that I bought a few months ago.  They drew about 250% of their rated power for the first 2 seconds after power on. That's it.

For larger fluorescent tubes the figures will be different - especially for the latest high frequency tubes (they power on with no flicker) - but not by much.

Common sense should tell us that the myth was just that. Lighting circuits are protected by fuses/MCBs and if a light took 30 minutes worth of power in the first few seconds after being turned on then the trips would have been triggered all the time.

[Ivan]

Yep, they are definitely not as bright in the short term - but I just fitted a slightly higher rated CFL bulb to my PIR - I think I have something like 18w CFL instead of a 40W tungsten and it's slightly brighter even though it's only just fired up. I also used the instant -light CFLs to reduce the delay in firing up.

[SteveH]

 I thought this idea went back to incandesant bulbs, were the hot & cold cycling & the large initial current surdge would stress the filiment material & cause it to fracture... I seem to remember seeing derating information for life expectancy of these bulbs based on flashing or intermitent use..... Wouldn't suprise me if it was one of the electricity suppliers who discovered this....

[Racerbear02]

Thanks for the replies

Just goes to show how urban myths are propagated.

[wyleu]

You can get rough usage  filament bulbs, for a little extra cash you get a better quality filament with an noticably increased life.

[Ivan]

aren't they slightly less efficient?

[martin]

it may have been from tvs in the 50's and 60's - you switched it on, and waited an eternity for it to fire up, usually accompanied by the characteristic smell of dust buring off the valves, often with a trail of smoke as something fried..........the attitude was "if it's going, tiptoe away quietly, and don't disturb it" - knowing full well that switching on was the time components fried with monotonous regularity

[wyleu]

I don't know about the efficiency of them but I know that when you have four floors worth of office suites to change you could always spot the non rugedised ones because they were the first to blow. Given we were swallowing 400 Amps three phase to produce a milliwatt of video signal as output, we would have looked at other areas to reduce the costs which in the scale of things were pretty low. Half the electricity went into the equipment and the other half went into the air conditioning to cool it all down.