Dissapointed from a LED light bulb...

[zahwi]

so its wire cutter's at dawn

Great!
Where can you get a ticket?!

 

[spec]

Great!
Where can you get a ticket?!

From either Nigel or me: £100 (50 for Nigel and 50 for me)

 

[Tony Stewart]

The design choices now are far better than what is discussed here. Unless you have DC with PWM dimmers, it isn't worth the expense of DIY crude FW bridge cap designs.
Not only will it be unreliable, it will flicker. Big brand bulbs now ONLY use SMPS in each bulb and cost at present $2/W. This will come down as chips are already < $0.1/W in this power range which come in a wide range of powers and voltages in one matrix chip.

 

[spec]

The design choices now are far better than what is discussed here. Unless you have DC with PWM dimmers, it isn't worth the expense of DIY crude FW bridge cap designs.
Not only will it be unreliable, it will flicker. Big brand bulbs now ONLY use SMPS in each bulb and cost at present $2/W. This will come down as chips are already < $0.1/W in this power range which come in a wide range of powers and voltages in one matrix chip.

Yeah, that is the proper way. The CCFLs that originally repaced incandescent lamps have a bridge rec, res cap and high voltage inverter, all stuck in the base of the lamp. They are a great source of components. I have a box full of worn out CCFLs for that reason.

 

[chemelec]

I Just bought these 60 watt, 110VAC bulbs. (10 Watt LED)
$3.00 in Canada at the Dollar Store. (or about $2.00 USD.)
(Same one is $3.69 or so, at Wallmart)

Quite a NICE Bulb, Especially for the Price.
Not as Yellow in color as a Incandescent Bulb.

Cutting the Top Bulb Off, this bulb puts out a VERY BRIGHT Light, Great for the Workbench.

 

[Tony Stewart]

LED bulbs have gone in the last few years from $2/W to $0.2W with approximately 70 to 100 lumens per watt for the cheapest. about the same as a high efficienct T8 4ft tube/ ( 88 LPW) ( previously stated incorrectly they were $2/W now)

These should bulbs and LED tubes should come down to $0.1/W in volume in near future with more options than CCT of ~2800'K(warm), 4500'K(Neutral) 6000'K (cool bluish) with the neutral being the most natural, but still slightly different than Halogen white in reflected brown shades.

Although the newer ones are dimmable if in parallel with Halogen , they will not dim the same and tend to be brighter at the low scale.


I recommend you get the dimmer compatible type which used to add $050 or so in cost,
Watch out for the marketing colour names when they don't specify CCT.

The 4 chip PAR30's are great replacements for Halogen ceiling pot lights in 8W or 10W.

 

[chemelec]

These are Rated at 10 Watts and draw 185 mA., Rated at 3000K and 800 Lumens.
Non-Dimmable, but that is not something I need anyway.

The actual LED Display is rated at 94 Volts.
The Driving Chip on the PCB is a BP2833A, and checking out the Data Sheet for it, It looks GOOD.

Yes I would like 6000K, But I am NOT prepared to pay the Price for those.

 

[Tony Stewart]

There is no cost difference in CCT for 3000Kto 4000K vs 5000K or 6000K. But 4000 to 5000K may cost more as this is more difficult to control as this a knife edge balance between cool and warm and the thickness of phospor is extremely thin compared to BLUE led substrate.

In fact 6000K is historically cheaper as it uses less phosphor and is easier to control process.

The efficacy of 3000K will be 10% to 15% less than 6000K due to phosphor absorption loss. Thus 6000'K is automatically brighter for the same LED substrate.
This is due to wavelength conversion process of phosphors that absorbs energy while shifting wavelength of energy.

 

[audioguru]

Why do "daylight" light bubs produce "cool blue"?
Today it is sunny with a blue sky. Everything does not look blue like they would be if lighted with a "daylight" colored light bulb.
My CFL bulbs are 3000k and 3500k and make pure white light that is not yellowy-pink warm and not cool blue.

 

[chemelec]

In Theory, That May be True, But NONE ARE AVAILABLE at the Price I PAID.
NOT EVEN CLOSE to that $3.00 Canadian.

 

[Tony Stewart]

if it appears cool blue then the "daylight" designation is incorrect or the product is inferior.
The colour rendering on cool white is not the same as Xenon bulbs that may appear cool white since they are broad spectrum and close to 100% CRI.
In fact CRI is worse (lower) than 70% and gets worse as CCT rises.

Be aware that when looking at the colour of the light is not an accurate assessment of it's colour. You must look at reflected colour from any object. Technical data always compares to NIST std's for daylight" with CIE correction factors for eyes.

THis is why we can see TV with just 3 narrow colour LEDs as white but when using this as a light source the reflected image has poor colour rendering since most of the spectrum is missing except those tuned to our eyes RGB rods and cones to give an "apparent" white.

Some of you may have seen my comparison of 4500'K tubes with daylight. in -powerpoint ( free viewer from MS if you dont have it.)
https://www.dropbox.com/s/oga2rljpz2l687a/Fluorescent light effects compare to natural sun.pps?dl=0

Notice the closest tube over counter gives a slightly "warmer" mushroom brown rather than grey taupe actual shade. and this is only 4500K. But 6000K would be quite bluish.

These tubes cost a lot more than standard daylight since they are tri-phospor rather than typical cheap 2 phosphor types.

Someday I'll compare with my "daylight" LEDs which have no apparent blue in them. ( I have 10 thousand of them here.)

 

[MrAl]

Hi,

I just had my 4100k 'curly' tube blow out the other day, have not obtained a replacement yet.
The color temperature was nice.

I like that color for looking at some color critical applications:
(1) Reading component color codes (orange and red can get mixed up with warm whites)
(2) Color of cooked meat, especially beef. Have to know if it is still blood red or not.

Now i see them calling the ones around 4000k "daylight" while the higher color temps are called "bright white".

What amazes me is that it takes the world decades to get stuff like this right.

Ever try looking for a "green" LED? Buyer beware, they all have their own idea what green is too. Stinky yellow green to nice saturated forest green. The stinky yellow green ones look like somebody got too drunk and puked on a white LED

 

[Tony Stewart]

"True Green" is 525nmD. and that's what all LED suppliers describe it as
Aqua green is 515 nmD can be more efficient and more popular in some sources.

The D stands for Dominant , which means CIE eye corrected as opposed to a flat response of a silicon sensor where they use nmP for Peak.
Always use nmD.

I remember being defined as colour-blind from a pastel colour test with hidden numbers by our C-MAC health nurse. The fluorescent light was bad too as I recall.
Pastel coloured resistor codes are even harder to read. (Yet I have been reading R codes all my life)

CRI ratings re based on an old method of using pastel samples. ( Color rendering index)

The better CCQ ratings endorsed by Wendy when she was at NIST are based on saturated colour samples which favour LED's over fluorescent tubes. She once told me since the FL industry was more influential, the CCQ ratings never got implemented.

This gives more info on rendering
https://www.google.com/patents/US20040218387