Q: how find specific LED

snipped-for-privacy@columbus.rr.com wrote in news: snipped-for-privacy@googlegroups.com:

Yeah... I know... I know... I still thought it was funny.

Don't know what it will unduly run down, but it won't be your battery. :-)

This would have been a photodiode into a TIA... (no bias voltage :^) and fet opamp. With 1 Gig or 100 Meg ohm FB resistor for the lowest ranges. You can get 1 Gig R's for ~$0.2 in onsies. 10 gig cost several bucks... just went shopping.

George H.

Hi Martin, that's a nice simple model, but I'm not sure it explains the data. For the green led the slope is ~3/2 from ~100 uA to 100 nA. I mumble, hand wave and say 'non-radiative recombination', but I don't know. The only sorta model I can imagine is a bunch of resistors and series batteries with the battery voltage near the band-gap voltage. All that in parallel with the ideal LED.

But still not very satisfying.

At the physics level I think about the density of carriers in the conduction band and some time diffusing around looking for a hole to recombine with. If you think about each electron being in some area, that might give you a term at low current that goes as I_led^2. The LED manufacturers might have a good picture of what's going on... but it's hard to find that info on the web.

George H.

IME modern T1-3/4 LEDs usually have a knee at about 10 uA and are pretty linear above that. So once the OP figures out what he means by 'reasonably bright', especially the solid angle/brightness tradeoff, it's reasonable to take the datasheet candela value, multiply by the ratio

(50 uA - 10 uA knee)/(datasheet current).

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC / Hobbs ElectroOptics 
Optics, Electro-optics, Photonics, Analog Electronics 
Briarcliff Manor NY 10510 

http://electrooptical.net 
http://hobbs-eo.com

We used a 50G 30% 0805 awhile back, $3 in onesies. It was for that nice surface voltage tool for semiconductors.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC / Hobbs ElectroOptics 
Optics, Electro-optics, Photonics, Analog Electronics 
Briarcliff Manor NY 10510 

http://electrooptical.net 
http://hobbs-eo.com

Hi Phil, Got a name/ part number of a data sheet that I could look at? At one point in the past I wanted a light source that would be fairly linear in light output.... (which was why I took the LED data above.) It would be nice if some LED was linear in current from 10 uA to 1 mA... All the data sheets I look at have a lower current of 1 mA.

George h.

Haven't made a homebrew optocoupler in a decade or so, but these ones look pretty good.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC / Hobbs ElectroOptics 
Optics, Electro-optics, Photonics, Analog Electronics 
Briarcliff Manor NY 10510 

http://electrooptical.net 
http://hobbs-eo.com

Are you confident that the photodiode was linear? That could be tested with filters.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

One could argue that if there is any current, then electrons are crossing the bandgap, then photons will happen. Maybe. A good LED has picoamps of reverse leakage in the dark, so there may be not much ohmic shunting to prevent the electrons from navigating the bandgap.

We used to use the original blue Cree LEDs as VME access indicators, at 50 mA. As they improved, we were really annoying people. We run the Osram blues at 1 mA.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Get the smallest physical diode you can. So, you should look for lower maximum current, that might indicate a smaller diode. Don't look for the "brightest" at some test current.

A few manufacturers do have luminance vs. current graphs, so you might search for those and try to extrapolate.

Jon

an "in" and several "outs", all thru-hole connectorized to make things nea t and tidy. This was for low-voltage DC on the order of a couple of amps p er output, so some decent size connectors one could easily solder to the bo ards.

the power rail so we could tell at a glance whether the board was powered-u p, once installed in the product.

a microscope to solder.

There's TWO reasons that was bad design; not only is it hard to hand-place the parts, but a connector puts stress on the board, bending it slightly whenever your connectors are manipulated. A surface-mount part usually doesn't have muc h strain relief, you can crack the parts in normal operation. A bent-leads throug h-hole resistor has a lot of compliance to deal with stress (and with a suitable standoff, or lead-forming, so does a leaded LED).

Huh, I've never questioned PD linearity at low light levels. I've done at lot of switched gain testing (switching FB R) and it's always been linear. (you might have to subtract off various offsets at low current.) Reading of recombination models they are all proportional to the (minority) carrier density.. like here. tps://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-772-compound-semiconductor-devices-spring-2003/lecture-notes/Lecture18v2.pdf

Which I think implies it's linear (light level with current) all the way down.

So color me confused as usual.

It might be there is a lot of non-radiative recombo near the edges... that might give a 3/2 power... a one dimensional density of defects.

George H.

• That was one of the many CREE red, clear lens LEDs i measured at 10uA,

25uA,and 50uA. Useful at 10uA as an indicator, just bright enough to show. Fairly bright at 25uA and "wow" at 50uA. Would guess that it would blast the eyeballs at 40mA.

• I recommend you try one out at 50uA, it is quite bright and should do the job you want.

To do A-B LED visual comparison, put them in series, using a 500K resistor minimum with variable PS.

• A Huggins lamp (mercury relay pulsing an open coax) has light output linear with voltage and used to test risetime of a PMT. The eye sensitivity seems superior to the PMT; turning down the driver/coax voltage will show weaker and weaker light pulses,even 5nSec wide...until it manifestly drops out at about 30V; the ionization potential of mercury. Use pulses, the brain is a superior correlator. Bet the eye wins.

A 2.2Meg resistor takes no extra electronic fluff, and at 24V supply the 10uA is very close to zero load.

Robert Baer wrote in news:lGEaG.362476$ snipped-for-privacy@fx39.iad:

Also unable to illuminate the LED. D'OH!

Use the RIGHT current limit resistor and use THIN PWM 'on' pulses to make for 'dim' operation.

Does not explain the WIDE discrepancy i saw: Suggest you look at the "comparison" of a 45 mcd VS 5000 mcd LED. Everlight 45 mcd: 10uA reads 18.5Meg, 25uA reads 4.56Meg, 50uA reads

1.73Meg. ROHM 5000 mcd: 10uA reads 7.6Meg, 25uA reads 1.86Meg, 50uA reads 0.41Meg. Note the 45 mcd at 50uA reads near what the 5000 mcd at 25uA.

I like the detail they gave concerning lead bending/forming/protection.

...

I've used these for high efficiency indicators, although at 100uA.

also the flux efficiency is great. And it's specified in datasheets, which is rare for indicator LEDs.

(Eye is more sensitive to green, so I prefer those)

Green HLMP-CM1G 475lm/W 27-59cd @ 20mA Red HLMP-EG1A 200lm/W 12-20cd @ 20mA

--
mikko