Frugal lighting

[Ron Peterson]

That's because the rods of the eye are more sensative to blue light and the human eye has more rods than cones.

[Ron Peterson]

That also occurs. I am just repeating the recommendation of consultant that was given to our building managers. The recommendation was to increase the color temperature of the fluorescent lights.

And that also depends on the type of ballast used.

[Don Klipstein]

In , TKM wrote: (edited for space)

I would like to add: Watch out for overheating CFs by putting them in small enclosed fixtures and in downlights. Many CFs don't take this well, and there is a general trend that this is worse with higher wattages.

Some actually rated to take the heat of recessed ceiling fixtures are Philips SLS non-dimmables up to 23 watts.

I consider it notable that when I see recessed ceiling fixtures in commercial buildings with CFs, the bulbs usually do not have built-in ballasts and will not screw into "regular sockets". The ballast is somewhere else in the fixture where it will not get as hot.

- Don Klipstein ( [e-mail])

[daestrom]

Sorry to hear yours don't work out so well.

I buy the 'whatever' brand at Wal---- and they last several years for me. The ones in the kitchen are the shortest lived because the fixture has an enclosed 'globe' that seems to hold the heat in and shorten their life. But even at that I get 18 months to 2 years from them.

daestrom

[Don K]

Why can't they use components that compensate for the heat?

Don

[Don K]

twice had to return

recognised that some will

It's worth a try. I'll just take it back next trip to the 'depot. Thanks

Don

[JeffM]

Perhaps you mean *tolerate* the heat or *disipate* the heat.

The former could mean more exotic materials. Silicon Carbide semiconductors are being developed which operate over a wider temperature range. They will, of course, be more expensive than old-school parts.

The latter requires more surface area so that the heat will be more readily radiated to ambient. This means that the volume of the product must increase to accomodate a larger heatsink and/or venilation must be increased.

In a strongly competetive market, these options don't look especially attractive to most manufacturers. As the word spreads about the problems that these products have and if a repository of comparative data is established, perhaps things will change for the better. I won't be holding my breath.

[Don Klipstein]

I think that the problem is finding components that have low failure rate after 6,000 or 10,000 hours of working at internal chip temperatures roughly 150 degrees C or close to 300 degrees F.

The junctions within the semiconductor devices will be significantly hotter than the surfaces or the leads of those.

The surfaces and leads of heat-producing transistors (or other semiconductors) have to be hotter than the air inside the "base"/"ballast housing" that houses those things.

The air inside the housing has to be hotter than the inside surface of the housing.

The inside surface of the housing has to be hotter than the outside surface of the housing. Looks like I gotta take my "Ray Tek" infrared non-contact thermometer to see how hot that is when a CF is crewed into a recessed ceiling fixture or a small enclosedfixture...

Meanwhile, that has to be hotter than the air around that, and it gets awfuklly warm in the upper regions of recessed ceiling fixtures!

Heat has to "flow downhill" (from higher temp to lower temp). Transistors and other semiconductors with silicon main chip material and the usual epoxy molded surrounding body are usually rated to only 150 degrees C junction temperature. Some metal case ones are rated to 200 C junction temperature, but often (maybe normally) with unimpressive life expectancy figures when pushed that hot, as well as higher cost.

Silicon carbide semiconductors take much higher temperatures very well, but so far "The Market" has decided that in most cases silicon carbide power semiconductors (with cases and encapsulants that take such heat and tolerate warping from big temperature changes occurring unevenly throughout the semiconductor package) are too exotic and expensive in comparison to making do with silicon and epoxy molded packages for them.

The alternative is "inductor"/"choke"/"reactor" ballast rather than an electronic one, wound with "magnet wire" having insulation and having any interlayer insulation and insulation between winding and core and the winding bobbin rated for really high temperatures. Keep in mind that such ballasts have lower efficiency than magnetic ballasts and they can vibrate unless extra work is done to minimize that and a few people are bothered by the AC flicker from fluorescent lamps with ballasts other than electronic ones using high frequencies.

- Don Klipstein (not Don K) ( [e-mail])

[Anonymous]

It wouldn't matter how well the (CF) unit "dissipates" the heat if it's in an enclosed fixture. ( you'd have to look into how well the fixture dissipates the heat )

On the bright side, lets say a fixture is rated for "60 watt max" bulb, then a 60W CFL which only creates 18(?) watts of heat should have a pretty comfortable factor. ????

As for hi-temp components.... most commercial grade electronic components should be good up to "boiling" ( 212f ) except of capacitors.... ( I don't know about modern caps )

[Don Klipstein]

Turns out, a 42 watt CFL heats a fixture about as much as, or even a bit more than, a 60 watt incandescent. A CFL produces little infrared.

Of course, how much a lamp heats up the house is determined by wattage alone.

A downlight fixture at its temperature limit will almost certainly bake a CFL to death pretty quickly.

Most semiconductors are rated good to a junction temperature of 150 C, or 302 F. Oc course, the surface of the semiconductor has to be cooler than that in order for heat to flow out. And the air around the semiconductor has to be even cooler, the inside surface of the ballast housing has to be even cooler, and the outside surface of the ballast housing has to be cooler still.

Decent electrolytic capacitors are rated for 105 degrees C but usually only a few thousand hours life expectancy at that temperature. CFL manufacturers have put a lot of work into finding capacitors that hold up well enough to not do too much damage to their reputations.

- Don Klipstein ( [e-mail])