Author Topic: Perceived relative LED brightness at same current: Blue is brightest?  (Read 3468 times)

0 Members and 1 Guest are viewing this topic.

Offline 741Topic starter

  • Frequent Contributor
  • **
  • Posts: 406
  • Country: gb
    • Circuit & PCB Design (small PCB quantities OK)
The following is pretty subjective...

I put a blue led in series with (in turn) yellow#1, yellow#2, green and red. The setup was simply a variable voltage source (0..5V) plus resistor (220R), and the leds obviously have the same current. The blue always looked brightest, regardles of drive voltage. The brighness of both seems to track quite well, the blue always looks brighter by the same sort of proportion. With the LED Vf drops, this means, I am running it at a low max current (1.3 mA measured maximum).

At the brightness I want (just 0.4mA or so), placing a 10k shunt across the blue evens out the brightness.
Blue Vf changes from 2.63 (no shunt) to 2.59V (with shunt).
Voltage across 220R series resistor moves from 0.1V (no shunt) to 0.121V (with 10 shunt across blue LED), corresponding to a current change in the whole series circuit of 0.021/220 = 0.1mA.
V(blue) = 2.6V, V (yellow) = 1.8V.

See also this link
https://electronics.stackexchange.com/questions/208523/human-perceived-relative-brightness-of-indicator-leds
Quote
"adjusted them to equal perceived brightness...The blue diode was the lowest intensity (mCd) which I could find, and you can see I had to scale its current way back."
« Last Edit: May 01, 2020, 09:31:03 am by 741 »
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 14117
  • Country: gb
    • Mike's Electric Stuff
Did you try (cold) white ?
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline 741Topic starter

  • Frequent Contributor
  • **
  • Posts: 406
  • Country: gb
    • Circuit & PCB Design (small PCB quantities OK)
Thanks for your suggestion - just tried it. I'd say the series combination W/B is better balanced brightness-wise than Y/B.

Offline mzzj

  • Super Contributor
  • ***
  • Posts: 1298
  • Country: fi
Reds and blues are typically lot higher efficiency than green-yellow, sometimes by factor of ten.
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15157
  • Country: de
Blue LEDs are pretty efficient, but the eye is not that sensitive to blue light. So some GaN based green ones (not the yellow green ones based on GaP or similar)  can be higher brightness to the eye, even though there energy efficiency is often lower.
Another big factor is the opening angle of the LEDs.
 
The following users thanked this post: tooki

Offline 741Topic starter

  • Frequent Contributor
  • **
  • Posts: 406
  • Country: gb
    • Circuit & PCB Design (small PCB quantities OK)
This was viewed directly above the LEDs. The blue seems considerably brighter than red, yellow or green at the same current.

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Those particular random LEDs just happen to be that way. LED efficiencies vary easily by three order of magnitude. Best are around 50% of electrical energy converted into light; some are like 0.1%, a lot worse than incandescent bulb.

Comparing any random LEDs is like comparing a skateboard and a Formula 1 car because both have four wheels.

Human eye is most sensitive on yellow-green range, and less sensitive on blue, but the difference here is something like 2-3 times, much less than a possible 10x-100x difference between the efficiencies of randomly selected LEDs.

The differences are especially very large in green LEDs; very low-efficiency green emitters are still widely manufactured and sold, and easily found in your parts bin, whereas very inefficient blue leds existed only shortly, and are rare in people's part boxes.

But you can get a green LED which is as efficient as the most efficient blue LED (or at least close), in which case it looks brighter at the same power input. For such high-efficiency green LEDs, the Vf is also close to the Vf of the blue LEDs.
« Last Edit: May 01, 2020, 01:34:05 pm by Siwastaja »
 
The following users thanked this post: Someone, tooki

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 7466
  • Country: hr
Like Kleinstein and Siwastaja said, green is absolutely brightest with same optical power output.
When making some start clocks for car rally races, I had to run green ones at very low currents in comparison to amber and red ones to have same percepted brightness..
"Just hard work is not enough - it must be applied sensibly."
Dr. Richard W. Hamming
 
The following users thanked this post: tooki

Offline 741Topic starter

  • Frequent Contributor
  • **
  • Posts: 406
  • Country: gb
    • Circuit & PCB Design (small PCB quantities OK)
Absolutely; it is well known the eye has best response to green. I naturally expected green would be brightest (which it was not) - but my own tests are very subjective, I realise that. I agree with the observation that blue LEDs are more recent - up to a point - but blue has existed for many years now.

Quote
But you can get a green LED which is as efficient as the most efficient blue LED (or at least close), in which case it looks brighter at the same power input

Um... I am running the same current in my tests, and that was my 'constant' for brightness comparison no attempt to equalise power was made... But still - is this something you can say you have tried for real - or that someone else has tried for real?

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 14117
  • Country: gb
    • Mike's Electric Stuff
Absolutely; it is well known the eye has best response to green. I naturally expected green would be brightest (which it was not) - but my own tests are very subjective, I realise that. I agree with the observation that blue LEDs are more recent - up to a point - but blue has existed for many years now.

Quote
But you can get a green LED which is as efficient as the most efficient blue LED (or at least close), in which case it looks brighter at the same power input

Um... I am running the same current in my tests, and that was my 'constant' for brightness comparison no attempt to equalise power was made... But still - is this something you can say you have tried for real - or that someone else has tried for real?
Deep green ( around 532nm) can be pretty efficient and look insanely bright
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 
The following users thanked this post: tooki, 2N3055

Offline DrG

  • Super Contributor
  • ***
  • !
  • Posts: 1199
  • Country: us
Although it may not be terribly relevant or useful, this thread reminded me of playing around with an RGB color mixer some five years ago. I started thinking about how I could get a handle on the equal brightness for the R-G-B elements.

The psychophysics of “brightness” is not simple. One way that I thought about was to attach a pot to each element and then adjust them until they appeared to be equally bright. Of course, the elements in an RGB do not make that easy. You could use three of the same kind of RGB LED and hope that the specs were close enough between them. Otherwise, I would need to alternately turn each on and off while trying to "remember" the brightness.

While the adjustment to achieve the perception of equal brightness is essentially sound, it was a PITA – more accurately a pain in the eye. I just did not feel comfortable staring into LEDS.

FWIW, I ended up using a BH1750 https://www.mouser.com/datasheet/2/348/bh1750fvi-e-186247.pdf  light sensor with different resistors that I had on hand. From the resulting lux values, I generated the graph below. The horizontal dotted line represented “equal brightness”.  The resistor values falling on that line were what I chose for the R-G-B segments, respectively.

Not saying it is the best way, but It worked pretty well for me.



- Invest in science - it pays big dividends. -
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
I use cheap bead type LEDs to repair light fittings. I like to replace the old LEDs with new one of the same color temperature.
To me the LEDs classed as neutral white look brighter than the warm white ones.
BTW, I have used over a thousand of the 3W types without a single failure so far. http://i1195.photobucket.com/albums/aa385/brlux/LED/Bead%20LED.jpg
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline rdl

  • Super Contributor
  • ***
  • Posts: 3667
  • Country: us
Older yellow-green LED on left at 15 mA next to a modern InGaN 525nm green LED at 300 uA (yes, micro amps).



Original post:

https://www.eevblog.com/forum/beginners/how-can-i-determine-the-resistor-value-for-an-unknown-type-of-led/msg2893554/#msg2893554
 
The following users thanked this post: tooki

Offline I wanted a rude username

  • Frequent Contributor
  • **
  • Posts: 662
  • Country: au
  • ... but this username is also acceptable.
Also depends on ambient brightness, because while the photopic (light-adapted) eye is most sensitive to 555 nm light (slightly yellowish green), the scotopic (dark-adapted) eye is most sensitive to 507 nm light (cyan-green).

This works in favour of the (still relatively inefficient) InGaN green LEDs compared to expensive but highly efficient phosphor-based true greens, if you're using them at night.

Also both the true and perceived colour of InGaN LEDs will change slightly depending on their brightness. When running one extremely dim (uA range) at night, it'll appear thoroughly cyan!
 

Offline tooki

  • Super Contributor
  • ***
  • Posts: 13157
  • Country: ch
Absolutely; it is well known the eye has best response to green. I naturally expected green would be brightest (which it was not) - but my own tests are very subjective, I realise that. I agree with the observation that blue LEDs are more recent - up to a point - but blue has existed for many years now.

Quote
But you can get a green LED which is as efficient as the most efficient blue LED (or at least close), in which case it looks brighter at the same power input

Um... I am running the same current in my tests, and that was my 'constant' for brightness comparison no attempt to equalise power was made... But still - is this something you can say you have tried for real - or that someone else has tried for real?
You’re still not getting it. You’re working from the assumption that current is directly related to light output when comparing LEDs. It’s not. As others have already explained, there are gargantuan differences in LED efficiency (that is, the amount of output per input).

Modern-type LEDs are insanely efficient. At the 20mA that used to be “normal” for an indicator LED, they’re downright unpleasantly bright.
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Quote
But you can get a green LED which is as efficient as the most efficient blue LED (or at least close), in which case it looks brighter at the same power input

Um... I am running the same current in my tests, and that was my 'constant' for brightness comparison no attempt to equalise power was made... But still - is this something you can say you have tried for real - or that someone else has tried for real?

High-intensity green Vf may be actually very close to the Vf of a blue led, so current is almost ~ power. Do note that different blue emitters vary between 3.0 and 3.7V on "typical" Vf. A high-intensity modern green is easily over Vf=3.0V.

And yes, I have actually, for real, used RGB emitters where the green is visually brighter than the blue, for the same current, or for the same power, as well. Human eye is more sensitive to green, so all that is needed to make this happen is to have the green as efficient as the blue.
« Last Edit: May 02, 2020, 03:21:08 pm by Siwastaja »
 
The following users thanked this post: tooki

Offline OldEE

  • Contributor
  • Posts: 18
Granted this is subjective but in our server room, a few years ago now, from 20-30 feet the blued leds appeared to be considerably brighter than other color leds.
« Last Edit: May 02, 2020, 06:56:05 pm by OldEE »
 

Offline bson

  • Supporter
  • ****
  • Posts: 2497
  • Country: us
Um... I am running the same current in my tests, and that was my 'constant' for brightness comparison no attempt to equalise power was made... But still - is this something you can say you have tried for real - or that someone else has tried for real?
I've calibrated the brightness of Kingbright APTD1608LZ (grn, 0603) and the red and blue counterparts to use for PCB indicators.

Here are the colors, If, Vf at If, and limiting resistor value for a 3.3V supply for me to empirically match their brightness to a reasonable level.  (If you feed them 1mA they're so bright you'll see spots.)  Again, this is for PCB indicator use, not a flashlight or other illumination.

GRN, 25µA, 2.75V, 22kΩ
BLU, 280µA, 0.80V, 8.87kΩ
RED, 67µA, 1.80V, 22kΩ

It's perfectly possible that the blue is barely conducting, and may well have a higher maximum brightness - at much higher currents.

This works great for me, because I tend to use mostly green indicators.  Red and blue is more unusual.
« Last Edit: May 02, 2020, 07:16:11 pm by bson »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf