Re VON's double-quoted message below: There are all sorts of stories of
this nature going around. Most are urban myths, based on the fact that
the switch-on phase _used_to_ be highly energy intensive with early
types of ballast. All CFLs now on the market use electronic ballasts,
which are a big improvement on the old magnetic ballasts in terms both
of energy consumption at switch-on and of robustness. It's true that
lots of on-off cycles will shorten the life of the ballast, but mostly
we don't use lights like that anyway (unless we have children of a
certain age...). And it is possible to get the ballast and tube
separately if you have an application where you need to do that - then
you just replace the ballast when it wears out. For the inside of
fridges, we usually don't have much choice what sort of lighting to use
- though I think the best solution there would be a few white LEDs.
It may well be that switching on causes the same amount of wear and tear
[I note your correction, VON] as 3 hours or 5 hours or whatever of
actual use. That doesn't mean that if you only need it for 10 minutes
you should leave it on for 3 hours. [For some other types of ballasted
lighting (e.g. metal halides), it _is_ advisable to leave the unit on
for several hours at a time, then not switch on again for several hours.]
Re temperature, low temps shouldn't cause damage, just a reduction of
light output. But even so, I'm looking forward to the day when white
LED fixtures are a bit cheaper - they make much more sense for outside use.
I've been using CFLs pretty well to the exclusion of all other lighting
types since the late 80s - all but 20 months of that in England. In
England, I never had any problems with them of any description (apart
from on occasion having impatient people around who didn't like waiting
for them to "warm up"; and sometimes having to replace fittings that
weren't designed to fit them). In 15 years, I think I only ever had to
replace about 3 or 4. But since moving here, I've had several cease to
function well before the 8000 hours that the packaging claims (itself
down from the 10000hrs that at least used to be claimed in Europe).
That's probably because I took to buying the bargain basement ones from
Costco, which are probably Chinese-made and not subject to the same
quality control that a European or North American made CFL would be. It
may also be that the variability in the voltage is greater here in rural
SK - and a poorly-made ballast will not cope well with that.
As for the frequency distribution, I can see that you have a genuine
issue here, Scott. However, there is no reason in principle why the
manufacturers could not address it. In your position I would get
together with my opthalmologist (or whatever the correct medical
specialism is) to write to all the main manufacturers about it (Philips,
GE, Sylvania, Osram) and see what happens.
atb
Mark
Steven Kurylo wrote:
> On 3/24/07, VON <ad070@sfn.saskatoon.sk.ca> wrote:
>
>>
>> And I forgot to mention, for maximum use of CFL's (and other
>> ballasted lights
>> like 'regular' flourescents, mercury vapour, sodium, etc DEPENDING on
>> the
>> ballast used), the first time you turn on that unit, DO NOT turn it
>> off for 5
>> hours, let it run similar to a 'burn in'. After that, note that
>> turning on a CFL
>> uses as much energy as using it for 3 hours, and never turn it on,
>> then off
>> within 10 minuts of turning it on (shortens life by many years), so
>> not suitable
>> for closets, fridges, etc.
>
>
> I'm sorry, but you're going to have to back that up. Turning on takes
> three hours worth of power? That doesn't even pass the smell test.
>
> I've read some places suggesting you leave it on for at least 3
> minutes, otherwise a incandescent would be more effective. So a
> closet where you never leave the light on for more than few minutes
> might be better off. If you're doing 10 minute cycles, you're still
> better off with a CFL.
>
> Now when manufactures rate a bulb's lifetime, they have to estimate
> the amount of times you'll power cycle the bulb; since more cycles
> means it will fail sooner. I've read they use a 3 hour cycle for that
> test.
>
Received on Sat Mar 24 23:56:57 2007
This archive was generated by hypermail 2.1.8 : Sat Mar 24 2007 - 23:57:07 CST