Blue LED at 3.3V
in [Design]
|
From: Don Klipstein on 4 Jun 2010 00:33 In article <nadg06t4ibi8k42rh16suo81dkvj31gsjf(a)4ax.com>, John Larkin wrote: >On Thu, 3 Jun 2010 22:57:45 +0000 (UTC), don(a)manx.misty.com (Don >Klipstein) wrote: <SNIP to shorten a longish story about LEDs being bright at low current> >> I sem to figure the greens may be dropping 2.6 volts. So, I figure >>about 4 mA, 10% of the time to get an average current of .4 mA. Since >>these LEDs are probably a little more efficient at 4 mA than at .4 mA, >>I would expect brightness about typical for .5 mA. And I have some Cree >>and Nichia green LEDs that are plenty bright at .5 mA. > >My Osram greens, beautiful parts, are "dim" at 0.5 mA, 2.62 volts, and >"bright" at 5 mA, 2.9 volts. You or anyone else want LEDs that get nice-and-bright at low current? I just tested a Cree CP41B-GFS-CM0N0784 that I got from Digi-Key. That is a 4-pin through-hole part having characterization current of 30 mA and maximum average current of at least 35 mA IIRC. I would expect its ratio of light output to current to be maximized somewhere in/near the range of 2-5 mA. The test that I just did was to measure light intensity at 100 mm from the LED, with a Lutron LX-101Alight meter, with LED current of .5 mA. The reading was 0 lux when the LED was off, and 10 lux when the LED was on. As a result, it appears to me that this LED with nominal viewing angle of 70 degrees achieved 90 millicandela at .5 mA. Please don't blame me if repeating this test with different instrumentation and LEDs from a different lot being operated at current far from what they are characterized at only deliver 50 or 40 or 30 MCD at half a milliamp (still plenty good-and-bright for indicator LED purposes), especially for rated "viewing angle" wider than 60 degrees! How about a more extreme one? In the old-fashioned-through-hole 5 mm (T1-3/4) through-hole format? Nichia NSPG520AS! Sadly, only (at least officially) available from Nichia's sales offices, in minimum quantity of 100. IIRC, I got 100 of them for 60 cents each plus shipping. I tested a few (all from same lot) and achieved 25-40 millicandela at .15-.16 milliamp (with the variation mostly noise in my measurement with a rather inexpensive light meter). The nominal viewing angle of that one is 45 degrees. I found more like 48, maybe 50. There is a 3 mm version - NSPG320CS. I found the viewing angle to be more like 50-55 degrees, hardly narrower than most diffused LEDs rated for 60 degrees. (Viewing angle is usually defined as where intensity is halved by being half that angle off-axis.) My limited testing of a few units in one lot found 25-40 millicandela at .21-.22 milliamp. And, 10 millicandela appears to me to be somewhat maybe-fairly above median for indicator LEDs, and 15 MCD appears to me to be somewhat to maybe fairly above mean. ================================================================== How about a white one with an almost-Lambertian-wide viewing angle of 110 degrees (half intensity 55 degrees from axis) and extremely efficient and Digi-Key-available? That one is Cree C535A-WJN-CU0V0231. It is a 5 mm / T1-3/4 LED that is stubbier than the usual "bullet" shape ones. Minimum order of 1 at 58 cents plus shipping/handling and any under-minimum-order fee. That one deals with low current extremely well. I just tested one to have voltage drop of 2.6 volts at .5 mA. I found 30-40 millicandela, and it may take 15-25 to appear decently bright for a white LED as opposed to a deep color one that appears similarly decently bright at 10 MCD. Sampling - 1 unit of 1 lot with current 2.5% of "characterizing current". I disclaim warranty beyond what I got paid to post this if your LEDs fail to do what mine did. :) However, I expect extremely high rate of these "wide angle" white LEDs being "nice-and-bright even for white indicator LEDs" at 1 milliamp. Even after being filtered to yellow by yellow "plexiglas" (to a shade of yellow less orangish than usual of yellow LEDs - maybe pushing to "a hair greenish"). 3 milliamps has efficiency or at least "overall luminous efficacy" of these close to maximized at fairly close to 100 lumens/watt. Voltage drop of the one I just now tested is about 2.73 volts at 3 mA. My Lutron light meter says 31 lux at 10 cm - .31 candela, 310 millicandela. Probably around .8 lumen from close to 8.2 milliwatts! -- - Don Klipstein (don(a)misty.com)
From: Don Klipstein on 4 Jun 2010 20:03
In <slrni0h0ha.j9d.don(a)manx.misty.com>, I, Don Klipstein wrote in part: > You or anyone else want LEDs that get nice-and-bright at low current? > > I just tested a Cree CP41B-GFS-CM0N0784 that I got from Digi-Key. > <shortened to 90 mcd at .5 mA> Back to the topic of blue ones - there are blue versions of this part. I just tested a Cree CP41B-BFS-CG0H0454 of the less favorable-to-low-current lot of 2 lots that I have blue CP41B series LEDs from. I'm coming up with 10 mcd at .5 mA. With voltage drop of 2.53 volts. A 10 mcd blue LED is easy to see. I am finding that one easily visible even over a white background illuminated to extent of achieving 1,000 lux. -- - Don Klipstein (don(a)misty.com) |