/diy/ - Do It Yourself

[Anonymous No.1215761]

why is nobody manufacturing everlasting light bulbs?

is it possible to diy?

[Anonymous No.1215762]

>>1215761
>everlasting light bulbs
not profitable

>possible to diy
yes

[Anonymous No.1215826]

>>1215761
You can essentially make any incandescent lamp last forever if you power it with extremely reduced voltage.

Lower filament temperature = slower evaporation = longer life

The problem is that lower filament temperature also equals lower efficiency, and the electricity used to run incandescent lamps costs more than the incandescent lamps themselves. Therefore, it makes more economic sense to design lamps for high efficiency and short lifespan and just replace them more often. Long life lamps only make sense in critical applications (e.g. traffic lights) or when the lamps are particularly difficult to change (e.g. high ceilings, and traffic lights again).

[Anonymous No.1215833]

>>1215762
This. Look up Phoebus cartel OP.

[Anonymous No.1215846]

>>1215761
Incandescents rated for 10,000 hours exist, and you could probably squeeze 10x the life out of them if you put a resistor (to reduce the operating voltage slightly) and an NTC thermistor (to block inrush current) in series with the socket.

[Anonymous No.1215869]

>>1215761
easiest way to extend the life of filament-based incandescents, don't switch them off. Or on :) The record holding lifspan bulbs (100 year +) are, somewhat counter intuitively, constantly on. Pro light rigs where incandescents will be switched a lot (stage lighting, etc.), most of the dimmers use a low current pre-heat, which prevents reduced filament life expectancy that results from lots of switching. None of this (pre-heat, constant op) saves money either, but, it should extend the lifespan some.

[Anonymous No.1215873]

>>1215761
Use LEDs. They can last 50k hours easy. Though, the measurement of lifetime and lumens is a bit different than incandescent.

[Anonymous No.1215914]

>>1215761
>The Lightbulb Conspiracy

[Anonymous No.1215916]

>>1215846
>using a resistor as a light dimmer
Good idea if you want to waste a bunch of electricity creating heat, or possibly start a fire.

A transformer or proper light dimmer (with soft start, if possible) will do nicely.

>>1215869
Those "record holding lifespan bulbs" are such because they have thick ass filaments that burn at a very low temperature and suck down a ton of electricity just to produce a dim orange glow.

More often than not a bulb will fail when it's turned on, but the shock really just puts the final nail in the coffin. They would fail soon enough even if they were left on. I've had plenty of them fail suddenly during operation.

A soft starting system will always help a bit, but I can't imagine it being worth the investment for a residential application.

[Anonymous No.1215930]

>>1215916
>Good idea if you want to waste a bunch of electricity creating heat, or possibly start a fire.
You just need to drop 5% of the voltage or so. For a 60 watt bulb, that means your resistor is only dissipating 3 watts. You'd probably lose more than that using a transformer, and unlike your dimmer, there's no "smart" electronics that will fail when you least expect them to. Plus, a resistor that size is like ten cents. Sometimes the simplest solution is the best one.

[Anonymous No.1215936]

>>1215930
5% voltage drop is not going to make a 1,000 hour bulb last 10,000 hours. Not even close.

https://en.wikipedia.org/wiki/Lamp_rerating
I actually have a spreadsheet template for running these calculations quickly and easily. I have some sort of autism for incandescent lighting.

In my experience, they are fairly accurate in practice, especially if you don't go into extreme adjustments.

A 5% voltage drop (down to 114V for a 120V lamp) doesn't even get you to 2000 hours. In order to break 10000 you need to reduce all the way down to 99V, and you end up with rather poor efficacy, as expected.

Doing any better than this would require a lamp of superior design than the norm to begin with. Lowering the voltage is not magic, it's a trade-off that kills your filament temperature and efficacy in exchange for lifespan.

[Anonymous No.1215940]

>>1215936
That's why I said start with a 10,000 hour bulb.

[Anonymous No.1215945]

>>1215846
Pic related

[Anonymous No.1215946]

>>1215940
Any 10,000 hour bulbs I've seen have had very poor efficacy to begin with. They're basically just sold to people who don't understand how incandescent lamps work and think that most bulbs only last 1,000 hours purely because of some cartel and planned obsolescence. What these people end up with is a dim, yellowish light that costs them more in electricity than they ever spent on all those 1,000 hour bulbs they were going through so quickly.

There are genuinely superior bulbs out there, but they're few and far between and hard to get a hold of. The halogen-filled bulbs that became mainstream a few years back have pretty good efficacy and respond well to dimming, but in the most recent year or so, manufacturers actually downgraded them to make them less efficient, while essentially blatantly breaking federal law for efficacy standards, but nobody seems to care.

[Anonymous No.1215975]

ive only seen one oe 2 replys about leds itt. surely undervolting them and building the driver circuit robust enough will guarentee them to last decades, right?

[Anonymous No.1216002]

>>1215833
>Phoebus cartel
>Members' bulbs were regularly tested and fines were levied for bulbs that lasted more than 1,000 hours

[Anonymous No.1216026]

>>1215761
I bought a desk lamp and a bulb freshman year in College, 6 years later, same lamp 8-10 hours a day on average open, still working.

WTF, I moved 4 times since then as well, and survived

[Anonymous No.1216046]

>>1216026
At our old house we had a bathroom light fixture with four bulbs, and one of them kept working for like a decade while all the others had been replaced multiple times. I think it was still working when I just replaced all four with new halogens.

He was a good bulb.

[Anonymous No.1216060]

>>1216026
Byron the bulb

[Anonymous No.1216069]

Actually if you read about the subject a light bulb can last for 100 years. But they realized they would be out of business, so they designed them to fail....see http://www.centennialbulb.org/

[Anonymous No.1216070]

oops didnt read thread

[Anonymous No.1216338]

A very high quality LED could probably be everlasting.

[Anonymous No.1216958]

>>1216338
Why anyone would still be using lightbulbs is beyond me.

[Anonymous No.1216960]

>>1216958
Lightbulbs have a smoother spectrum, LED can look weird and artificial and fuck with colors

[Anonymous No.1216962]

>>1215946
>downgraded them
I fucking knew it but I had no way to quantify it.
I bought like 2 dozen quality brand halogens a few years back that were cheap because of a pricing error.
I bought some more like last year and two have already died, while the older ones are still running. It's not like they're even getting used different, they're all in my garage on the same circuit.

[Anonymous No.1216975]

>>1215761
There is a /diy/nosaur that makes vacuum tubes for fun.

He shows his stuff some times and to be honest, considering the curves he gets from his homemade tubes, his stuff is fucking relevant.

To be honest, if there is one fag here that knows his shit about filaments and glasswork, he's the one !

Just summon him, and wait for him to happen in your thread :)

[Anonymous No.1216993]

>>1216962
>I fucking knew it but I had no way to quantify it.
I'm just going by the lumen numbers on the packaging. They literally tell you flat out that they have less lumens.

For example, the 72W 1000 hour halogen lamps used to be 1490 lumens as a standard, but now they're around 1200-1250. They're using the more lax "modified spectrum" regulations to skirt the law, but many of them aren't even modified spectrum in the slightest.

The 72W halogen lamps have always been lame, though, because even 1490 lumens isn't really "100W equivalent" since good 100W lamps were in excess of 1600 lumens, sometimes even 1700. Now that they're only ~1200 lumens, they're closer in brightness to old 75W lamps than 100W.

[Anonymous No.1217009]

>>1216960
It depends, LEDs with color rendering index as good as the incandescent lightbulbs ones do exist.

[Anonymous No.1217016]

>>1217009
There's not an LED in existence that has a 100% pure light spectrum when compared to a blackbody radiator. That being said, there are many that are damn close and good enough even for the most autistic of lighting snobs.

The problem is that you can't just walk into a store and buy LED lamps that are consistently good in every way.

Tints vary from model to model or even between examples of the same lamp, many models them have 120Hz flicker, light distribution characteristics are all over the place, etc.

With incandescent lamps, light spectrum will vary between brands and models due to differing filament temperature, but regardless of the filament temperature, the light will still conform purely to the natural spectrum of light at that temperature (there are exceptions like GE Reveal lamps, but that's an intentional thing that's clear and upfront and you don't have to buy them). There will be no worry of flicker, because the filament inherently responds too slowly to flicker in any significant way at mains frequency. The light distribution is nearly perfect by nature, because the filament just glows in all directions, unlike LEDs which are directional and have to be engineered for proper light distribution. And while they generally should be avoided for reasons of efficiency and quality, all of these points apply to even the cheapest piece of junk incandescent lamps you can find.

What hurts LEDs is not technological limitation, but poor design and quality control by manufacturers.

[Anonymous No.1217046]

just get leds, they last 20 years

[Anonymous No.1217046,1]

LED lights produce blue light underneath the white.
Specific Blue light frequencies affect the nervous system in a negative way, not to mention the increased chances of Glaucoma
Now you know why the push