I think LED technology has changed a lot since the 70's. I have been searching LED data sheets for this nonlinear effect that shows increasing luminous efficiency with increasing current, and I haven't found an example for any LED produced in the last 20 years or so. At high enough current, all LEDs have nonlinearity that goes the other way (less luminous efficiency at higher current). I suspect this whole myth revolves around a particular short period in LED history that is equivalent to the point contact era of transistor technology. Things have changed.
I think more like congruency not yet confirmed or falling short of 100%.
Point taken - but I maintain that at least most of visually perceived gain in efficiency/efficacy of an LED by increasing instantaneous current with same average current and avoiding visible flicker is due to a radiometric gain.
I believe they merely failed to check that most of the gain was due to gain in radiant power output. LEDs of types that have more shift to shorter wavelengths at higher currents largely work more efficiently at lower currents. LEDs of types that appear brighter when fed pulsed current (with frequency high enough to avoid visible flciker) than fed steady current of same average current have the gain at least mostly radiometric. Radiometric gains from pulsing sometimes fail to get detected of the radiometric sensor or amplification afterwards runs into saturation effects during pulses of pulsed radiation.
I argue greater usefulness!
This business of pulsing causing an increase of sensitivity of the human eye to light even while visible flicker is avoided is at least mostly not true! I find this applicable to readers of sci.electronics.basics since a few of those would desire to exploit this supposed "phenomenon" to improve luminous efficacy of LED flashlights and LED bicycle lights. But let it be known - high brightness LEDs havwe a major trend of having little, zero or negative gain in efficiency or luminous efficacy from making instantaneous current higher than average current when average current is anywhere from "characterized" to "maximum continuous" or anywhere higher, as long as frequency is high enough or offtime is short enough to avoid visible flicker.
Point taken. However, I have seen a few datasheets for HP LEDs (all of chemistry available prior to the late 1990's, often of chemistries dating back to early 1990's or even older) indicating photometric output per millimap increasing with current. Meanwhile, the more recent higher efficiency (higher luminous efficacy) LEDs do well at having efficiency or at least luminous efficacy close to maximum at a current that they are rated to take continuously - and in some cases less than 20% of their maximum current.
No, but if the relevant points are available on the web, such as any HP or Agilent datasheets or application briefs, point these out and I will comment!
With exception of one LED chemistry largely not used in pulsed LED digital displays especially not during or before the early 1980's, LEDs used in digital displays were significantly more efficient with 40 mA at
5% duty cycle than with steady 2 mA. Efficiency could easily be doubled or more from pulsing with over 10-20 mA per LED chip compared to steady current of 2 or worse still often average curret of 1 mA per LED chip - awfully common when a segment of an LED digital display usually had more than one chip!
Go ahead and find an "original formula red" LED or LED digital display known to have chemistry GaAsP on GaAs substrate and with peak wavelength
660 nm (not 680-700 nm - that of a common "low current red" chemistry that has a broader spectrum). Feed the darn LED or a darn segment current that is regulated to 30-50 mA at a low duty cycle. See what happens when you put a big capacitor across the LED (CAUTION - have it not charged above about 2 volts). AND confirm with radiometric measurement *confirmed to not have saturation effects* from pulsing of the light. In my experience, a silicon solar cell usually does well when current it pushes it pushes through without needing more than .3-.35 or so volt - peak that is! In the general wavelength range from very-near-infrared to yellow-green, silicon solar cells have efficiency merely proportional to wavelength. So an LED changing from 665 nm to 655 nm due to instantaneous current will not cause a solar cell to work much more poorly. Also, I advise having one of those "clear CDs" that some recordable CD spindle packs have one of. Those make fairly good diffraction gratings, and those work a lot better still in front of my digital camera than in front of my eyes! Good for verifying sectral shifts or lack thereof!
I want to add that InGaAlP is a high brightness LED chemistry used in many modern LEDs and efficiency of even that one falls when current decreases to a fraction of "characterized" current. Agilent even recommends pulsing theirs to keep instantaneous current at least 10 milliamps if the average current will be less. One reason given was that one aging mechanism was due to a combination of temperature and duty cycle - apparently this aging mechanism is a fiffusion one that depends on electric field in the LED chip. Another reason I suspect (not that I know of Agilent saying this) is that aged LEDs could have high efficiency more dependent on current at least 10 milliamps than new ones. To a lesser extent this is also true of GaAlAsP and Agilent's similar AlGaAs. Efficiency is less there when current is a few percent or less than "characterizing" current.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.