Close
0%
0%

Which colours are LEDs sensitive to ?

Measurement of the spectral sensitivity of several LEDs

Public Chat
Similar projects worth following
Measurement of the spectral sensitivity of several LEDs

Which colours are LEDs sensitive to ?

LEDs are nice for illumination, and sometimes, LEDs are used as a colour sensor. But I wished to use a handful of LEDs as a spectral sensor with more channels than just red, green blue...

So I soldered 14 different visible LEDs, two infrared LEDs and two photodiodes to a microcontroller board, using the same measurement principle as in https://hackaday.io/project/19244-leds-as-photodiodes-for-1-kb-challenge

Basically, every LED is connected with both anode and cathode to GPIO pins, anode is set low, cathode is set high, and then cathode is switched to input. Photo current discharges the junction (and wire) capacitance, and the time until the cathode input pin toggles is measured.




Channel list (1 to 6 are normal brightness tinted LEDs, 11 to 18 clear high brightness LEDs):

01 Red
02 Red
03 Yellow
04 Green
05 Blue
06 Blue
07 IR 880 nm SFH485
08 IR 940 nm
09 Photodiode
10 Photodiode BPW34
11 Darkred EPD-660-3-0.9
12 Red EPD-630-394
13 Red
14 Yellow
15 Red
16 Yellow-Green
17 Green
18 Blue

Here are their forward voltages, measured with a constant current of 4.2 mA, sorted for ascending voltage:

(10) 0.63 V Infrared Si photodiode BPW34
(09) 0.65 V Infrared Si photodiode
(08) 1.08 V Infrared from an old TV remote control
(07) 1.22 V Infrared SFH485
(11) 1.55 V Barely visible very dark red EPD-660-3-0.9
(15) 1.68 V Red
(01) 1.72 V Red
(12) 1.75 V Red EPD-630-394
(03) 1.85 V Yellow
(13) 1.87 V Red
(02) 1.89 V Red
(14) 1.90 V Yellow
(16) 1.92 V Yellowish-Green
(04) 1.92 V Green
(18) 2.68 V Blue
(17) 2.70 V Green
(05) 2.76 V Blue
(06) 3.17 V Blue

Error of forward voltage measurement should be in the +- 0.03V range.

My assumption was that the LEDs with direct bandgap semiconductor material would be roughly visible for their own emitted wavelength, and the indirect bandgap ones would be sensitive to wavelengths slightly shorter than their own emission.

With a nice spectrometer available in laser spectroscopy lab at university, I measured the emission spectra:

Note that the spectrometer is configured for Raman spectroscopy with a 473 nm laser, therefore, it is equipped with a lowpass interference filter, which cuts off wavelengths shorter than about 482 nm. The blue LEDs spectra are clipped for this reason.


For a first small test to confirm the wavelengths the LEDs might be sensitive to, I took my decorative LED lamp I build many years ago which contains red, yellow, green, blue and ultraviolet LEDs. With this, I illuminated all the sensor channels and took notes which is sensitive to which colour.

I was very surprised, but observations first:

Red illumination:    Reaction on channels 1           | 7 8 | 9 10 | 11 12 13 14 15 |       18.
Yellow illumination: Reaction on channels     3 4     | 7 8 | 9 10 | 11 12 13 14 15 | 16    18.
Green illumination:  Reaction on channels     3 4     | 7 8 | 9 10 | 11 12 13 14 15 | 16 17 18.
Blue illumination:   Reaction on channels     3 4 5   | 7 8 | 9 10 | 11 12 13 14 15 | 16 17 18.
UV illumination:     Reaction on channels   2 3 4 5 6 | 7 8 | 9 10 | 11 12 13 14 15 | 16 17 18.

Note that the diffuse LEDs 1, 2, 5, and 6 had a much lower sensitivity than the other ones. I took a quite long discharge time in order to see if there is any sensitivity.

Surprises:

* The tint of "01 Red" must act as a narrow filter, so 01 is sensitive to the red illumination only.
* "02 Red" is sensitive to ultraviolet only despite being a red LED.
* "18 Blue" is slightly sensitive to everything, despite being a blue LED.

Clearly, my former rule of thumb that a LED is sensitive for its own wavelength or slightly shorter (more blueish) does not hold in all cases. Therefore, I cannot easily determine the LED sensitivity by measuring its emission spectrum. I need a tunable monochromatic source of light for measuring the actual sensitivity spectrum of the LEDs. Unfortunately, I could not take apart the expensive spectrometer in university lab..


Finally, I found a surplus grating monochromator with manual mechanics giving...

Read more »

Monochromator-LED-Messung-Durchblick.tar.gz

Raw data for the second experiment and Gnuplot scripts to recreate graphics

gzip - 6.25 MB - 01/02/2020 at 16:17

Download

LED sensitivity.tar.gz

Raw data of the measurements and Gnuplot scripts to recreate all graphics

gzip - 8.88 MB - 09/28/2019 at 13:46

Download

  • Second experiment with known part numbers

    Matthias Koch01/02/2020 at 16:10 0 comments

    Time for a second revision, this time with known part numbers:


    This is the list of LEDs used, from left to right:

    ---------------------------------------------------------------------------------------------------------------------------------------------
      #  Colour  Reichelt part     peakwl Vf min  typ  max     Crystal material    Angle   Brightness  Manufacturer part number          Notes
    ---------------------------------------------------------------------------------------------------------------------------------------------
    
      0    IR    SFH 4346            950 nm   -   1.6  1.9  V   ?                   20°                Osram OS     SFH 4346             Tinted
      1    IR    SFH 4356P           860 nm   -   1.7  2.0  V   ?                   20°                Osram OS     SFH 4356P            Tinted
      2    red   RND 135-00006       700 nm  1.8   -   2.6  V   GaP                 40°      10 mcd    RND          RND 135-00006        Tinted
      3    red   LED 3-3500 RT       660 nm   -   1.85 2.5  V   GaAlAs              50°    2300 mcd    Kingbright   L-934SRC-J4
      4    red   LED 3-4500 RT       660 nm   -   2.1  2.5  V   AlGaInP             34°    4500 mcd    Kingbright   L-7104SRC-J4
      5    red   EVL 264-7SDRC       650 nm  1.7  2.0  2.4  V   AlGaInP             40°     100 mcd    Everlight    264-7SDRC/S530-A3
      6    red   KBT L-7104SURC      640 nm   -   1.9  2.5  V   AlGaInP on GaAs     34°    2500 mcd    Kingbright   L-7104SURC-E
      7    red   KBT L-7104SEC-H     635 nm   -   2.2  2.8  V   AlGaInP             34°    1000 mcd    Kingbright   KBT L-7104SEC-H
      8    red   VIS TLHR 4900       635 nm   -   2.0  3.0  V   GaAsP on GaP        16°      13 mcd    Vishay       TLHR 4900
      9    red   SLK 3MM RT          627 nm   -   2.0  2.5  V   GaAsP on GaP        50°      60 mcd    Kingbright   L-934EC
     10   green  WUE 151034GS0300    525 nm   -   3.2  3.8  V   InGaN               30°   10000 mcd    Würth        151034GS03000
     11  orange  LED 3-7800 OR       611 nm   -   2.2  2.8  V   AlGaInP             34°    7800 mcd    Kingbright   L-7104SEC-J4
     12  orange  LED 3-3000L ONG     605 nm  1.6  2.0  2.4  V   ?                   20°    3000 mcd    Lucky Light  LL-304UAC2E-4AC
     13  yellow  LED 3-2800 GE       590 nm   -   2.0  2.5  V   AlGaInP             34°    2800 mcd    Kingbright   L-7104SYC-J3
    
     14  yellow  NTE NTE30032        592 nm   -   2.0  2.4  V   AlGalnP on GaAs     10°    2500 mcd    NTE          NTE30032
     15  yellow  LED 3-1300 GE       595 nm   -   2.0  2.4  V   AlGaInP             50°    1300 mcd    Kingbright   L-934SYC
     16    ?     VIS TLHY 4205       585 nm   -   2.4  3.0  V   GaAsP on GaP        22°      20 mcd    Vishay       TLHY 4205            Tinted, does not light up
     17 yl-green KBT L-7104MGC       574 nm   -   2.1  2.5  V   AlGaInP             34°     700 mcd    Kingbright   L-7104MGC
     18 yl-green KBT L-7104GC        565 nm   -   2.2  2.5  V   GaP                 34°      60 mcd    Kingbright   L-7104GC
     19   green  LED 3-14000 GN      520 nm   -   3.2  4.0  V   InGaN               34°   14000 mcd    Kingbright   L-7104ZGC-G
     20    red   WUE 151034RS0300    628 nm   -   2.0  2.6  V   AlGalnP             30°    2600 mcd    Würth        151034RS03000
     21   green  EVL 204-10 UBGC     502 nm   -   3.5  4.3  V   InGaN on SiC        20°      50 mcd    Everlight    204-10 UBGC/S400-A6
     22 trqouise RND 135-00151       500 nm  2.6  3.2  3.6  V   ?                   30°    8500 mcd    RND          RND 135-00151
     23   blue   LED 3-3700 BL       465 nm   -   3.3  4.0  V   InGaN               20°    3700 mcd    Kingbright   L-7104VBC-D
    
     24    IR    BPW 34                                         Si                  65°                Vishay       BPW34                PD 600-1050 Peak 900 nm
    
     25   blue   LED 3-2000 BL       465 nm  2.9  3.3  3.6  V   InGaN               30°    2000 mcd    ?            L-314BC-A3
     26   blue   LED 3-3000 BL       458 nm   -   3.2  3.7  V   InGaN               20°    3000 mcd    Kingbright   L-7104QBC-G
     27  blueish EVL 1254UBC         428 nm  3.0  3.8  4.5  V   GaN on SiC          40°     400 mcd    Everlight    1254UBC
     28    UV    LED LL 3-120 VI     400 nm  2.8  3.2  3.8  V   InGaN               20°     120 mcd    LuckyLight   LL-304UVC4B-Q5D
    
     29   green  KBT L-7113ZGC       515 nm   -   3.3  4.1  V   InGaN on Sapphire   20°   14000 mcd    Kingbright   L-7113ZGC            5 mm
    

    Sensitivity measurements were done exactly as in the first revision of the experiment, as explained in the Project Details. As I changed lab in the meanwhile, I cannot use the spectrometer anymore, therefore no measurements of the emission spectra this time. The peaks drawn in red denote the peak wavelength given in the respective datasheets.

    All parts were sensitive to direct sunlight, but with varying orders of magnitude of their discharge times. Therefore, some parts are not sensitive enough for this experiment, other parts are highly sensitive, but for wavelengths outside of the range of my monochromator.

    Data is sorted for emission wavelength.


    Infrared:

    BPW34 is the same photodiode as in the first run of the experiment, which is sensitive to the complete...
    Read more »

View project log

Enjoy this project?

Share

Discussions

Dan Maloney wrote 09/30/2019 at 16:43 point

Interesting results. Nice work, and good job cobbling together the test instruments for this.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates