I want to mention another factor: Blue light sources have visibility
and visual impact greater than would be indicated by their photometric
figures.
The main reason is that blue light has more "color impact" (my words)
than other color light of same photometric quantity.
One example I can cite is a calculation that I did where a mixture of
red, green, and blue LEDs added up to (providing adequate color mixing was
achieved) a shade of white with color temperature of 5000 K.
7% of the light by photometric measurements was from the blue LEDs.
29% was from the red LEDs and 64% was from the green LEDs.
This example (where % lumens from each color LED is not shown as of
11:11 PM EST 1/21/07 but is easily enough dervivable from) is the left
half of the 5000K row of the equal-current example set in:
http://www.misty.com/~don/ledrgb2w.html
The right side of that same row uses a different, more pure blue LED
(that appears to me to have run into disfavor from lower photometric
figures, despite being more suitable than "regular blue LEDs" in RGB
work). That specific example achieves 5000 K white with 5.1% of the
photometric content being from the blue LEDs, 66.6% from the green LEDs
and 28.3% from the red LEDs.
I have done some experimentation and personal studying into "blue
impact" translating to "noticeability" of blue LEDs.
So far, it appears to me that "regular blue " LEDs (nominal wavelength
470 nm) can easily appear 30-50% brighter than indicated by their
photometric data, and that "royal blue" LEDs (although coming up dimmer
than "regular blue" ones) can easily appear 45-90% brighter than indicated
by their photometric data.
Depending on ambient lighting and how adapted your eyes are to
brightness/darkness and whether you are viewing the light with central
vision or peripheral vision, scotopic vision ("night vision") may be
coming into play. Its alternative, "photopic vision", has a spectral
response well enough known and agreed upon at conventions to be basis of
definition of photometric content in a given radiometric quantity of light
as a function of its spectrum.
Should scotopic vision be having a significant effect, then the blue
LEDs can look extra-bright and the red ones a little dim. Keep in mind
that there is controversy over significance of scotopic vision when
ambient lighting level is high enough for photopic vision to be dominant.
So, blue LEDs can appear extra-bright or extra-noticeable.
One other factor - blue LEDs have a higher voltage drop than red ones,
and for equal current the blue ones get more power.
Another factor - if current is on the low side, most red LEDs (one
exception being a "low current red" chemistry) lose efficiency and blue
ones tend to not be so bad with efficiency loss at lower current - in
fact, with moderate underpowering most blue LEDs gain efficiency!
Another thing - choice of red LEDs. Ones with peak wavelength in the
630's and dominant wavelength in the 620's of nm have a high tendency to
have much more "luminous efficacy" (photometric efficiency) than ones with
peak wavelength around 660 nm. And Radio Shack does not appear to me to
have these really good red ones, despite the fact that they are now
common.
- Don Klipstein (
[email protected])