I meant to suggest light intensity measured (lux per wavelength-range?) by the sensor for a segment of the light spectrum. Light measurement units are many, and confusing (I usually fall back on photon-count).
Avoid Duracell and you won't have a problem!
I prefer Panasonic and Eveready these days - never had on leak.
It is quite rare now to find that 3 AA's & resistor configuration. Most have either a single AA cell or a pair of them and a voltage to current converter that drives the stack (or a single one) at constant current.
The ones that have 3 cells are ideal candidates for bridging the switch with a 1M resistor so that you can find them in the pitch dark.
They are increasingly rare at least in the UK. Most are now single cell.
I think it is the organic binder or glue that holds it all together that degrades to become darker with free carbon in it under the influence of such a high light flux.
Moisture ingress appears to be the main mechanism by which the high intensity ones degrade in the wild. The harder you run them the worse it gets with differential thermal expansion. eg.
The little key-ring LED things often have one cell, one LED, no resistor.
Still you'd expect a fairly large colour shift, which I don't think is observed.
Interesting, thanks.
Cheers
Phil Hobbs
It is the blue LED to phosphor interface degrading under the flux of blue photons so I'd expect comparatively little colour shift if the intensity of blue light reaching the phosphor is reduced by being absorbed by carbon black particles before it gets there.
Organic materials don't really get on with shorter wavelengths.
I'm not at all sure that charring is the mechanism, though. Usually polymer binders turn a dirty yellow, not grey.
Cheers
Phil Hobbs
For devices that are suppose to last for quite a long time, putting any light emitting devices in series is really stupid - for one of the reasons you point out. Same issue for the (LED or incandescent) Christmas lights. A vicious plot to drive the consumers crazy and increase profits for the mfg. I've repaired a few monitors that have a LED in series light strip. Sme stupid design. There are a few (2, 3) strips in parallel depending on the size of the monitor but if one strip goes out the brightness goes south quickly.
+1 on avoiding Duracell. I've never seen a battery other than Duracell where both the + and - endcaps separated themselves from the body. Enloop/Fujitsu or Amazon basic rechargables for the last 5 years for all my devices. Well there are the flashlights that use 14500. Never had any of those fail....
if any part of the long filament in a light bulb fails, it goes out and it is scrap ...
The old filament ones would be guaranteed not to work out of the box every year and require at a minimum tightening all the bulbs in their sockets and more likely a binary chop search to find the failed one. They also had full mains voltage across the break. Later versions had a conductive breakdown glass so that if one bulb failed it would provide a current path but no voltage drop so the rest stay lit but brighter. There was a fuse lamp to prevent runaway failure if several blew.
Curiously to save wire a lot of the modern LED ones do use long chains of LEDS and a common earth. I have a set under my desk from the Village Hall awaiting repair at the moment. Half the LEDs on it don't light.
LEDs are pretty reliable compared to filamen bulbs so it isn't all that crazy a design. The ones I object to are cheap and nasty Chinese rectified mains ones that put ~100 LEDs in series.
Due to the temperature coefficient LEDs need to be driven by a constant current source, not by a constant voltage source.
If you want to put multiple LEDs in parallel and feed them with a constant voltage, you need to put individual series resistors for each LED, this will limit the individual LED currents.
Putting multiple LEDs in series, a single constant current source is enough, like a single series resistor.
Right, I know/understand that. Different design approaches and those extra resistors amount to perhaps an extra $0.10 which probably adds $2.00+ to the cost of the final product. Yea gotta produce cheap, cheap, cheap and dupe the unsuspecting public.
I had a colleague many years ago (late 70's early 80;s) who worked at Zenith Corp designing cks for tv's and other electronic devices. His assignments typically consisted of: here is a circuit for a subsection of a TV...reduce parts count. Sometimes he was given the assignment of developing a new circuit, if his circuit added between $0.04 to 0.05 USD, he went through a 'design review' and in most cases was told to do it over again with no cost increase. He figures the money he saved went into paying the people that 'handcrafted' the sets....(there is a joke there if you remember those commercials...)
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.