degrees K color temp explained

[BVH]

Here's an explanation of the term "degrees Kelvin" as it relates to bulb color temperature we all talk about.

When a piece metal is heated up to "X" degrees Kelvin, the color of the metal is observed and called "X" degrees Kelvin. So for example, if this piece of metal is heated up to 6000K, it will be the glowing the color of the lovely 6000K HID bulb we all know and love. And when its heated to 4300K, it's glowing the color of the HID bulb many of us prefer.

Just for fun,

4300 Kelvin is equal to 3,280 degrees F, or 2,371 C
5000 Kelvin is equal to 8,540 degrees F, or 2,760 C
6000 Kelvin is equal to 10,340 degrees F, or 3,315 C


I'm not a mathematician, but I could not see, using the online conversion calculator, where there is any linear increase similarity between K and F. Meaning every time there's a 1 degree rise in F, it does not result in same incremental rise in K

It must be late for me to do this mindless stuff.

[StarHalo]

Without Googling it, I do recall that Kelvin and Centigrade are 1:1 (one degree more C is exactly one degree more K and vice versa regardless of the number), it's Farenheit that is the oddball. F and C intersect at -40.

[brickbat]

Originally Posted by BVH

4300 Kelvin is equal to 3,280 degrees F, or 2,371 C
...

Those conversions are wrong.

4300 K = 7280 F = 4027 C

[jtr1962]

K = °C + 273.15 = °F/1.8 + 255.37
°F = 1.8 x °C + 32 = 1.8 x K - 459.67
°C = (°F-32)/1.8 = K - 273.15

Note that the use of degrees with the Kelvin scale is considered incorrect. 4300 K is pronounced simply "4300 Kelvin", not "4300 degrees Kelvin.

[2xTrinity]

When a piece metal is heated up to "X" degrees Kelvin, the color of the metal is observed and called "X" degrees Kelvin. So for example, if this piece of metal is heated up to 6000K, it will be the glowing the color of the lovely 6000K HID bulb we all know and love. And when its heated to 4300K, it's glowing the color of the HID bulb many of us prefer.

An imporant point to note here:

You will not find a metal that will not melt or vaporize at 6000k, or even 4300k. Tungsten -- the metal with the highest melting point, melts at roughly 3700k at normal atmospheric pressure.

The sun for example actually does have a surface temperature of 5800k, however, it is certainly not solid, nor is a portable "sun" feasible for fixed lighting. For a higher color temperature to be achieved, either light must be subtracted through filtering, or light must be generated by other means than blackbody radiation (color of a heated up object).
For example, a fluorescent light uses a narrow-spectrum light source, such as a mercury vapor, to create UV, which is then converted to other colors using phosphors. While it may have an apparent color temp of 6500k, its spectrum is not at all similar to natrual daylight. This is the reason for the CRI unit -- to compare the color rendering of an actual lightsource, to ideal blackbody source.

A few other interesting things about color temp:

Planck's Law actually predicts the spectral characteristics of a blackbody object at different temperature, shown by the following curve. Note that as the temperature goes up, the total amount of energy emitted per unit surface area goes way up -- proportional to the 4th power of the temperature. This is why say overdriving an incandescent lamp that normally runs at 3000k up to 3500k will result in a huge increase in brightness for two reasons:

1. the total number of power emitted from the filament roughly doubles
2. The higher color temp filament emits colors that our eyes are more sensitive to.

This means a more than doubling of lumen output. However, expected lifespan of the bulb will drop from thousands of hours, to tens of hours.




Another interesting law is called Wien's Law. This says that there is a linear relationship between the temperature of a blackbody radiator, and the dominant wavelength emitted by that object.



This means at 2900k, the most dominant wavelength in the spectrum is 1000nm (Infrared)

For an object at 5800k like the sun at double the temperature, the dominant wavelength is exactly half, or 500nm (or roughly in the middle of the visible spectrum).

[balou]

Originally Posted by BVH

When a piece metal is heated up to "X" degrees Kelvin, the color of the metal is observed and called "X" degrees Kelvin.

Please note that it needs to be a perfectly black piece of metal. That's why it's called blackbody radiator

[Ken_McE]

Originally Posted by 2xTrinity

The sun for example actually does have a surface temperature of 5800k, however, it is certainly not solid, nor is a portable "sun" feasible for fixed lighting..

They all laughed when I talked about a portable sun at the academy, but I'll show them, I'll show them all... (heads off to high voltage forums)

[jhgreenstein]

Water freezes at 0 C which is equal to 273 K. Kelvin and Celcius use the same intervals so 1 C is 274 K and water boils at 373 K

The value of 0 Kelvin is the theoretical absolute zero point when atoms stop moving.

[LukeA]

Originally Posted by BVH

I'm not a mathematician, but I could not see, using the online conversion calculator, where there is any linear increase similarity between K and F. Meaning every time there's a 1 degree rise in F, it does not result in same incremental rise in K

If you want a scale with Fahrenheit increments that goes to absolute zero then you want the Rankine scale.

[LukeA]

Originally Posted by jtr1962

K = °C + 273.15 = °F/1.8 + 255.37
°F = 1.8 x °C + 32 = 1.8 x K - 459.67
°C = (°F-32)/1.8 = K - 273.15

Note that the use of degrees with the Kelvin scale is considered incorrect. 4300 K is pronounced simply "4300 Kelvin", not "4300 degrees Kelvin.

Exactly. We don't say "degrees Kelvin" because we don't measure Kelvin, we only calculate it.

The same is not true of the Rankine scale for whatever reason.