Fixture socket say only use led equivalent to 75watt incandescent

[jaggedben]

See my post #29. Your mileage may vary with a different LED bulb, but in my experience the driver section gets much hotter to the touch than the base. Just doesn't seem to me like the base conducts that much heat.

 

[wwhitney]

See my post #29. Your mileage may vary with a different LED bulb, but in my experience the driver section gets much hotter to the touch than the base. Just doesn't seem to me like the base conducts that much heat.

While it may not conduct much heat, I expect conduction is still the primary heat dissipation mechanism from the base. So the base may get quite warm from a fairly small source of heat within it.

Cheers, Wayne

 

[hbiss]

See my post #29. Your mileage may vary with a different LED bulb, but in my experience the driver section gets much hotter to the touch than the base. Just doesn't seem to me like the base conducts that much heat.

That's understandable. See my post about where the driver PC board is located. It's well above the base and the only connection is a pair of wires probably 3 inches long from the PC board down to the shell and center contact for power. So, there is no heat sinking to speak of.

-Hal

 

[jaggedben]

While it may not conduct much heat, I expect conduction is still the primary heat dissipation mechanism from the base. So the base may get quite warm from a fairly small source of heat within it.

Cheers, Wayne

Not sure I follow, and also this seems speculative. What are you calling the base? I'd call the screw-in part the base. That's not where the driver electronics are on these bulbs necessarily, they are in the next section, as much as halfway out toward the front of the bulb from the base. Then there is the outer bulb of translucent plastic for diffusing the light. The base (screw part) doesn't get got. So it seems to me like the driver section - which is usually made of a different type of plastic that almost feels like porcelain - is radiating most of the heat. Maybe I need to learn more about heat conductance but I would think that if the base were conduction the majority of the heat away from the bulb it would be closer to the temperature of the driver section.

 

[wwhitney]

Designs vary, I was thinking of a design with the driver in the base, where most of the driver heat generation would be in the base. Not sure what the thermal distribution would look like if the drivers are higher up.

As LEDs get cheaper, bulb life can be extended simply by driving the LEDs with less current. I assume that the driver circuity if reasonably designed would likewise generate less heat that way?

Cheers, Wayne

 

[Flicker Index]

Fixture say either 75watt incandescent or led equivalent led wattage.
So looking around 8-11 watts for led…..
Why would a led 100 watt equivalent not work still going to be way under the incandescent wattage.

seems dump to me

LEDs produce considerably more heat that has to be disposed through conduction and LED elements and ballast components are a lot more sensitive than incandescent lamps.

The energy going into an incandescent lamp that is not emitted as visible light is mostly disposed as infrared. The energy that is radiated as infrared correlates to surface temperature. So the amount of heat that an LED lamp can dispose of as infrared is entirely dependent on surface temperature. LED ballast and LED elements can not handle the temperature anything close to that of an incandescent lamp.

The incandescent lamp's bulb can be though of as a glass jar filled with really hot oil far hotter than the thermal tolerance of LED related components.

 

[Flicker Index]

Designs vary, I was thinking of a design with the driver in the base, where most of the driver heat generation would be in the base. Not sure what the thermal distribution would look like if the drivers are higher up.

As LEDs get cheaper, bulb life can be extended simply by driving the LEDs with less current. I assume that the driver circuity if reasonably designed would likewise generate less heat that way?

Cheers, Wayne

the modern topology both for reducing ballasting loss as well as reducing LED ballast circuitry cost is to match the LED string voltage as close as possible to line voltage.

With 12v strip lights, you'll commonly find 3 o 4 LED elements per rung with a ballast resistor. This arrangement is inherently inefficient. So, many LED screw-in lamps run each string in three digit voltage, some as high as 400v... and some as low as 65v.

The flicker index of many LED apparatuses though often rivals that of 1960s F40T12 technology or even worse. Actually earlier electronically ballasted PL-C CFL and 32WT8 had among the best flicker quality with many having less flicker than 100W filament bulbs.

 

[jaggedben]

LEDs produce considerably more heat that has to be disposed through conduction

No, they don't. Not per lumen. Only as percentage of power consumption, which doesn't mean much when the power consumption is so much lower.

 

[Flicker Index]

No, they don't. Not per lumen. Only as percentage of power consumption, which doesn't mean much when the power consumption is so much lower.

Yes they do. LED lamps have to dispose nearly all of whatever that is not emitted by conducting it away from the LED elements and ballast components and into a heatsink for cooling by convection. The exterior of dedicated LED luminaries are thermally connected to the ballast and LED elements. Their ability to directly shed heat as radiant energy is very limited.

A 25W light bulb and a 25W cement resistor behaves quite differently and LEDs behave more similar to the resistor. It maybe radiating perhaps 30-40% of incoming energy as radiant visible light. You put a 25W light bulb in a large globe and much of the energy would escape straight through the glove as radiant energy. If you were to paint the globe black, it would heat the entire globe and allow the whole globe to work as a heat sink. Well, if you run a 25W power resistor in place of the light bulb, it would overheat, because it can't effectively carry the heat out. You'd have to use fins and an internal fan to effectively transfer the heat to the outer globe. If you fed 25W of power into it, it would continue to get hotter and stays at a temperature at which point equilibrium is reached, but this point is much above the allowable temperature of the component.

Heat dissipation through conduction is quite small for incandescent lamps. Vast majority of wattage is released as radiant energy mostly in infrared range. Some directly by the filament, some by heated argon/nitrogen mix which then heats the outer bulb(this is why gas filled bulbs get much hotter to the touch than vacuum bulbs)

The hot glass then re-radiates as long wave infrared. The radiant emission vs temperature is exponential and the entire glass bulb carries the heat.

 

[Flicker Index]

So... if you throw in an unthrottled 17W 100W equivalent LED A-bulb in a jelly jar or a pancake meant to take a 60W or a 100W A19, the integral LED ballast will likely fry out in a matter of two to three digit hours. If it doesn't, it's throttled. Many two digit wattage LED lamps are throttled so the output is modulated to hold the heatsink at a setpoint to avoid frying out. Throttled LEDs coupled with a heavy heatsink is great when it won't remain lit for a long time at a time. But if you leave it on past the foldback point, it will dim down and it's no longer a "100W equivalent" in throttled state.

 

[Besoeker3]

You can touch an LED with your base hands. I have done so. Don't remotely try that with an incandescent. That's as simple as it gets.

 

[wwhitney]

Yes they do.

My impression is that you may be right, but you haven't addressed the "per lumen" question.

Say LEDs generate 10 times the lumens/watt of an incandescent. Then the energy dissipation accounting might be (to make up some numbers):

Incandescent: 4% visible light 4% conduction 52% non-visible radiation 40% convection
LED: 40% visible light 50% conduction 5% non-visible radiation 5% convection.

In which case the LED is dissipating 5/4 as much heat through conduction "per lumen", vs incandescent. But if the figures were slightly different, the ratio would end up below 1.

Cheers, Wayne

 

[Flicker Index]

You can touch an LED with your base hands. I have done so. Don't remotely try that with an incandescent. That's as simple as it gets.

You haven't tried touching the heat sink part of an LED lamp that's been in an enclosed fixture.

 

[Flicker Index]

My impression is that you may be right, but you haven't addressed the "per lumen" question.

Say LEDs generate 10 times the lumens/watt of an incandescent. Then the energy dissipation accounting might be (to make up some numbers):

Incandescent: 4% visible light 4% conduction 52% non-visible radiation 40% convection
LED: 40% visible light 50% conduction 5% non-visible radiation 5% convection.

In which case the LED is dissipating 5/4 as much heat through conduction "per lumen", vs incandescent. But if the figures were slightly different, the ratio would end up below 1.

Cheers, Wayne

The values for LEDs are probably outdated, but this gives a good idea.

 

[retirede]

You haven't tried touching the heat sink part of an LED lamp that's been in an enclosed fixture.

I replaced 50W BA14 halogen incandescents in my kitchen with 6W LEDs. I can remove the LED bulb after being on for hours and barely detect heat anywhere on the bulb. The halogens would result in 2nd degree burns.

Edit: the LEDs have a slightly higher lumen rating.

 

[wwhitney]

The values for LEDs are probably outdated, but this gives a good idea.

That data is a good start, but it's at the filament or LED chip level, not at the bulb level.

Cheers, Wayne

 

[jaggedben]

Yes they do. LED lamps have to dispose nearly all of whatever that is not emitted by conducting it away from the LED elements and ballast components and into a heatsink for cooling by convection. The exterior of dedicated LED luminaries are thermally connected to the ballast and LED elements. Their ability to directly shed heat as radiant energy is very limited.

A 25W light bulb and a 25W cement resistor behaves quite differently and LEDs behave more similar to the resistor. It maybe radiating perhaps 30-40% of incoming energy as radiant visible light. You put a 25W light bulb in a large globe and much of the energy would escape straight through the glove as radiant energy. If you were to paint the globe black, it would heat the entire globe and allow the whole globe to work as a heat sink. Well, if you run a 25W power resistor in place of the light bulb, it would overheat, because it can't effectively carry the heat out. You'd have to use fins and an internal fan to effectively transfer the heat to the outer globe. If you fed 25W of power into it, it would continue to get hotter and stays at a temperature at which point equilibrium is reached, but this point is much above the allowable temperature of the component.

Heat dissipation through conduction is quite small for incandescent lamps. Vast majority of wattage is released as radiant energy mostly in infrared range. Some directly by the filament, some by heated argon/nitrogen mix which then heats the outer bulb(this is why gas filled bulbs get much hotter to the touch than vacuum bulbs)

The hot glass then re-radiates as long wave infrared. The radiant emission vs temperature is exponential and the entire glass bulb carries the heat.

PER LUMEN

60-70% of LED power consumption is only 9 percent of incandescent power consumption if LED consumes (generously) 15% of the incandescent.
Also, some of that can be irradiated from the heatsink as IR as well as convection.
In other words, there isn't 'considerably more' (non-IR) heat from the LED. It's about the same.

Do you actually engineer these bulbs or just install them?

 

[Besoeker3]

You haven't tried touching the heat sink part of an LED lamp that's been in an enclosed fixture.

Nothing remotely like an incandescent in an enclosure. It's far more Watts.

 

[jaggedben]

So... if you throw in an unthrottled 17W 100W equivalent LED A-bulb in a jelly jar or a pancake meant to take a 60W or a 100W A19, the integral LED ballast will likely fry out in a matter of two to three digit hours. If it doesn't, it's throttled. Many two digit wattage LED lamps are throttled so the output is modulated to hold the heatsink at a setpoint to avoid frying out. Throttled LEDs coupled with a heavy heatsink is great when it won't remain lit for a long time at a time. But if you leave it on past the foldback point, it will dim down and it's no longer a "100W equivalent" in throttled state.

This is interesting, although I haven't noticed it happening in my enclosed fixtures. But in any case, it implies that the fixture manufacturer shouldn't have to worry about the heat generated by the LED bulb if the bulb is going to modulate itself.

 

[Flicker Index]

This is interesting, although I haven't noticed it happening in my enclosed fixtures. But in any case, it implies that the fixture manufacturer shouldn't have to worry about the heat generated by the LED bulb if the bulb is going to modulate itself.

Not every lamp has a burnout protection. Those that do often do to prevent bad reviews and returns, but they often don't mention the feature, because they don't want to advertise their 1600 lumen bulb becomes a 600 lumen bulb in an enclosed fixture after it is fully warmed up. It's very much like computer clocking down to prevent damage in case of blocked vent or fan problems.

What's the rated input wattage of the lamp, fixture type and how long did you observe the true input watt for?

Earlier in this thread, tall girl made excellent points. In addition to the LED chip itself, the materials used for the solid state fluorescent lamp's phosphor blend, as well as the components of LED ballasts can not reliably handle the temperature needed to reach an equilibrium to dissipate the heat.