I guess you've never worked a graveyard shift at a 24-hour porn store.
Guess what I do in 4 hours from now?
Guess what I weigh?
This study is just bullshit. Pure and simple bullshit with ZERO true controls (Yes, I read the study, I have full access as research director for a multinational corporation.)
This study is bunk and biased. PLUS I enjoy a better sex life versus the majority.
We don't do a full on-off cycle. We do a full wave half wave pulse. No flicker after start up, only cameras will detect it and only directly facing the lights themselves, and even then it's just like a desert mirage shimmer/flicker.
The reason for that is because they're actually under-driving the diode. It's a typical trick to extend the life of the diode and manage heat with their crap thermal designs.
Find one getting full power with a real heat solution and you'll see much different results.
"Go ahead and set that up some time, tell me how it compares to the lights on the market."
Works very well, as it's how my LED horticultural panels are made for the most part. Want a closeup of one of my boards so you can see the serial paths?
"You don't just use some random cheap LED, there are specific high intensity diodes that are used."
Wrong. We can use most any diode, even crappy 5mm tail-throughs. What matters is the overall PPFD, and I've got flashlights that will blind you using cheapass 0.02c diodes.
Then you run in to the problem of heat. LEDs are not uniform in this dissipation of heat and when a little diode is getting 3 watts of power, and you have a bunch of them, well that's an issue that requires some serious heat dissipation."
No more so than any typical computer heatsink/fan combo will handle.
"Then of course there's additional power supply requirements. Voltage conversion aside, a simple bridge rectifier wont' do it unless you care for flickering light."
After a couple of seconds of operation, diodes quit flickering as they reach their junction operational temperatures. Yes, my LED panels flicker as they go from dim to bright, but once they've hit peak in about 3 seconds, there is no more flickering. It's not an issue.
"Gets even more complex if you want a dimmable light"
Depends on your config. With my given serial setup, all you need is a potentiometer to drop the available current. Works just fine.
"You need PWM control circuitry to handle it."
Again, depends on the setup. Remember, LEDs are DC devices. You can get away with simple current limiters instead of pulse-width modulation, if the setup is made for it.
"Its luminous efficacy is equal to or above CFLs (which isn't true for many LEDs)"
This isn't true and hasn't been for a few years. Cree smashed 100 lux/w for usable human lighting years ago, and just recently they popped past 220 lux/w.
Most typical LEDs from Cree, Epistar, Edison, Nichia, Seoul Semiconductor, etc. that are out now offer 110-130 lux/w, with fluorescent's PEAK being near 95 lux/w in the most efficient fluorescent lamp type, the T5HO.
LED has been down in price, you're just not looking at the right bulbs or looking in the right places.
Avoid Home Depot and big-name stores. Just go find a manufacturer in China since that's where Home Depot is guaranteed to be getting it. You might want to pay a visit there (or ask me since I've been through Shenzhen and toured hundreds of places doing quality checks before signing contracts for production) to find a good manufacturer, but they usually have no MOQ and the only steep issue you might encounter is customs duties.
"generally, fluorescent bulbs like to stay on. they'll last longer than incandescents if you turn them both on and keep them both on until they burn out"
Nope, that tungsten filament will burn FAR longer than the electrodes in the CFL will last, assuming you don't degrade the color phosphors beyond usability first by running it nonstop.
Nope. It's just a small transformer and a starter cap in the ballast. Pretty hard to fuck those up. What fucks up are the electrodes or the excitation amalgam, or the tube construction.
"Seriously. Incandescent, aside from being What You're Used To, really is NOT very good light!"
That's actually WRONG. Let's see, first, tungsten has a few nice higher emission peaks around the green and yellow, which is what your eyes are most sensitive to. Next, incandescents have great IR output and 630nm/660nm output (again due to tungsten) despite their inefficiency, which makes them suitable as supplementary lighting for horticulture.
These lights also work very well in colder climates where CFL and some induction lighting/HID lighting will not work.
And most incandescents, properly manufactured, are IP-65 rated minimum for dust and water ingress protection.
At 5,000 RPM what is going ot really determine noise is the force of blades against air at a given RPM - given the shown shape of this design, it looks like a modified slim blower fan. These run much quieter than your typical axial fan.
Slashdot Mirror
I guess you've never worked a graveyard shift at a 24-hour porn store.
Guess what I do in 4 hours from now?
Guess what I weigh?
This study is just bullshit. Pure and simple bullshit with ZERO true controls (Yes, I read the study, I have full access as research director for a multinational corporation.)
This study is bunk and biased. PLUS I enjoy a better sex life versus the majority.
PURE glass (SiO2) does NOT block UV. You need doped glass, usually borosilicate, to begin blocking UV.
And I have a UV-only quantum meter - UV is getting through despite how good the glass or filter coating might be.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2829662/
Those of us that build UV-emitting medical devices know better, and I design lamps from the ground-up.
We don't do a full on-off cycle. We do a full wave half wave pulse. No flicker after start up, only cameras will detect it and only directly facing the lights themselves, and even then it's just like a desert mirage shimmer/flicker.
I've got spotlights using fans smaller than an inch in diameter over a micro-finned heat sink with high surface area.
12w flood lights up most of my 50x20 back yard better than the 100w incan flood that was there.
The reason for that is because they're actually under-driving the diode. It's a typical trick to extend the life of the diode and manage heat with their crap thermal designs.
Find one getting full power with a real heat solution and you'll see much different results.
"Go ahead and set that up some time, tell me how it compares to the lights on the market."
Works very well, as it's how my LED horticultural panels are made for the most part. Want a closeup of one of my boards so you can see the serial paths?
"You don't just use some random cheap LED, there are specific high intensity diodes that are used."
Wrong. We can use most any diode, even crappy 5mm tail-throughs. What matters is the overall PPFD, and I've got flashlights that will blind you using cheapass 0.02c diodes.
Then you run in to the problem of heat. LEDs are not uniform in this dissipation of heat and when a little diode is getting 3 watts of power, and you have a bunch of them, well that's an issue that requires some serious heat dissipation."
No more so than any typical computer heatsink/fan combo will handle.
"Then of course there's additional power supply requirements. Voltage conversion aside, a simple bridge rectifier wont' do it unless you care for flickering light."
After a couple of seconds of operation, diodes quit flickering as they reach their junction operational temperatures. Yes, my LED panels flicker as they go from dim to bright, but once they've hit peak in about 3 seconds, there is no more flickering. It's not an issue.
"Gets even more complex if you want a dimmable light"
Depends on your config. With my given serial setup, all you need is a potentiometer to drop the available current. Works just fine.
"You need PWM control circuitry to handle it."
Again, depends on the setup. Remember, LEDs are DC devices. You can get away with simple current limiters instead of pulse-width modulation, if the setup is made for it.
What part of SEPARATE do you not understand?
I've said for years we need a PHYSICALLY SEPARATED NETWORK that has zero access to the global internet.
IBM researchers took control of a nuclear power plant in just a couple of WEEKS.
What the fuck happened to AIR GAPS?
"you can use a dimmer switch with them"
Only specialized PWM dimmers. No regular in-wall dimmer compatible LED exists yet (at least, nothing has been delivered to my company for testing.)
LED is better than halogen - you don't irradiate your child with UVB and UVC (and that glass ain't filtering all of it, FYI.)
LED is also much more viable. Keep an eye on the market, it changes every couple months with each new advancement.
LED? Complex? A simple 4-diode bridge rectifier and a series string of lighting to match 120v is complex? Really?
"Think of all the millions of accountants trained in tax law that would be without a job."
Any real professional knows to keep up in their industry or get left behind.
If the bean counters can't keep on top of their game, they DESERVE to lose their jobs.
Dimming only requires a minor bit of PWM, easily implemented and we've done it for years with dimmable CFL.
"Its luminous efficacy is equal to or above CFLs (which isn't true for many LEDs)"
This isn't true and hasn't been for a few years. Cree smashed 100 lux/w for usable human lighting years ago, and just recently they popped past 220 lux/w.
Most typical LEDs from Cree, Epistar, Edison, Nichia, Seoul Semiconductor, etc. that are out now offer 110-130 lux/w, with fluorescent's PEAK being near 95 lux/w in the most efficient fluorescent lamp type, the T5HO.
LED has become a direct replacement for HID.
All that wonderful UV just flying around.
No thanks. Halogen lamps need filters and other stuff to be safe for human use. That only adds in cost and its STILL using a tungsten filament.
$100 or more per dimmable LED downcan?
Slashdot, I think you guys just need to defer to me as the LED lighting guru, here. You are getting ripped off on prices to no end.
Thermodynamics says bullshit.
LED has been down in price, you're just not looking at the right bulbs or looking in the right places.
Avoid Home Depot and big-name stores. Just go find a manufacturer in China since that's where Home Depot is guaranteed to be getting it. You might want to pay a visit there (or ask me since I've been through Shenzhen and toured hundreds of places doing quality checks before signing contracts for production) to find a good manufacturer, but they usually have no MOQ and the only steep issue you might encounter is customs duties.
"generally, fluorescent bulbs like to stay on. they'll last longer than incandescents if you turn them both on and keep them both on until they burn out"
Nope, that tungsten filament will burn FAR longer than the electrodes in the CFL will last, assuming you don't degrade the color phosphors beyond usability first by running it nonstop.
"CFLs are complex enough that quality matters."
Nope. It's just a small transformer and a starter cap in the ballast. Pretty hard to fuck those up. What fucks up are the electrodes or the excitation amalgam, or the tube construction.
"Seriously. Incandescent, aside from being What You're Used To, really is NOT very good light!"
That's actually WRONG. Let's see, first, tungsten has a few nice higher emission peaks around the green and yellow, which is what your eyes are most sensitive to. Next, incandescents have great IR output and 630nm/660nm output (again due to tungsten) despite their inefficiency, which makes them suitable as supplementary lighting for horticulture.
These lights also work very well in colder climates where CFL and some induction lighting/HID lighting will not work.
And most incandescents, properly manufactured, are IP-65 rated minimum for dust and water ingress protection.
Really, the only downside is in how you use them.
Then perhaps you should talk with me as $40 per bulb is outrageous and I know I can do tons better.
"Overclocking you CPU beyond spec"
Got some news for you, we can go that high because the processors are already shipping under-specced.
http://forum.coolaler.com/showthread.php?t=251959
Oh look, 5 GHz by UNDERVOLTING. Sorry if you can't read Chinese.
At 5,000 RPM what is going ot really determine noise is the force of blades against air at a given RPM - given the shown shape of this design, it looks like a modified slim blower fan. These run much quieter than your typical axial fan.
Fin depth doesn't need to be deep when you have so much surface area for the heat transfer.
Low power density? Petrol fuels have more than enough, such that a tiny amount can kill you if properly utilized.