Sunday, December 6, 2009

Another LED tube from China

The vendor of the LED tube exchanged the temporary tube with another using surface-mounted LEDs, which is an improvement due to consistency of the pick-n-place machine instead of manual assembly using cheap Chinese labor. I insisted on getting the SMD version because its spatial light distribution is more suitable to my kitchen ambient lighting.

SPECIFICATION:

Input power: 20 watts, measured 22 watts
Input voltage range: 110-220VAC
Number of LEDs: 288
Power factor: > 0.9; measured 0.8
Color temperature: 4000K

My comment:

Given the input voltage range, its power supply is most likely a flyback converter. That means that there is an output electrolytic capacitor that will limit the life of the lamp to about 5 years, more likely less than 5 years. Flyback power supply is not the most efficient power supply money can buy, therefore you get what you pay: power supply efficiency lower than 80%, more like 75%.

The lamp comprises paralleled banks of 12 LEDs in series with a current equalizing resistor. There're 288 LEDs, thus each LED is rated for less than 76 milliwatts, because the series resistors eat up some power.

Again the uninspired manufacturer went for the obvious: use the extant form factor of the fluorescent tube the LED tube is designed to replace.

What is wrong with that approach?

First, the use of 288 small LEDs means that the life of the LED tube is much lower than the rated life of the LED used, which is typically 10,000 hours, not the 100,000 hours for the high-brightness LEDs which are designed with appropriately low thermal resistance to last that long.

Second, the LED has more glare than the fluorescent tube it replaces, due to the much higher brightness of the LEDs used vs. the low-light-density of phosphors used in fluorescent tubes. It is not economical to use any sort of optics or lenses to control the beam spread to say 60 degrees vs. the normal 180 degrees. That means that if used in a hallway, from far away the persons to try to see are masked by the glare from the tubes.

In summary, the inherent glare and the relatively short life of the built-in flyback power supply will not make this LED tube a good financial move.

Wednesday, November 11, 2009

LED tube, reality versus advertised spec

So I bought a LED tube to verify their spec against measurements. The company's web site talks about >100 lumens per watt, power factor >0.9, which are very impressive.

Reality, measured with Kill-a-watt:

Apparent power: 25 VA
Active Power: 20 watts
Label power: 15 watts

Thus measured power factor = 0.80, not the advertised >0.9

With power supply efficiency of 75%(= 15/20), there is no way their lighting efficacy can be greater 100 lumens per watt. Using about 500 tiny LEDs stuffed by hand, their real lighting efficacy would be about 75 lumens per watt, if that much.

As I lower the tube's voltage with a variac, the input power actually increases until the lamp starts to flicker.

My guess is they use a linear current driver, due to the small size of the "power supply", which may be just a linear regulator such as LM317, costing maybe $0.20 in volume production.

BTW tiny LEDs as mentioned in an earlier post, are rated for only 10,000 hours. With 500 of them, their expected life will be much less. I'm returning the tube and getting my money back! BTW Costco sent their LED light bulb PAR-38 purchasers a notice to allow them to return their purchases for full refund, because they also figured out (maybe by reading my blog) that such a bulb can't last 10,000 hours versus 35,000 hours advertised on the package. Indeed, their LEDs are connected in series/parallel combinations, and I don't see any LED by-pass circuit should a LED fail by open circuit.

One thing that bothers me is the narrow light distribution of the LED tube, about 45 degrees. That means I can't use it in my kitchen light where two fluorescent tubes are used. Retrofit both of them will cost over $150 plus labor, therefore I'm reluctant to do it, in addition to my objection to narrow beam pattern, which in all honesty, can be an advantage in many applications.

Monday, September 28, 2009

CFL: Advertised spec versus reality

We bought a bunch of "Kill-a-watt" instruments to measure up to 8 parameters of electricity going into appliances. So one of the first things we do is to measure the real power consumption of the CFL used in the lab. Relevant data:

Power consumption: 17 watts (advertised 13 watts)
Power factor: 0.60
Measured VA (Product of actual input current and voltage) 26 volt-amp

Thus hundreds of millions of CFLs out there are costing the power plants twice the current as advertised. And when the power companies install digital power meters, the consumer will be paying for the VA consumption and not the real watts read by the current old-fashion power meters.

And remember that more than half the light flux from a CFL goes into the shade surrounding it. And they contain mercury. I personally was an early adopter of CFLs back when they looked like a long U. Many of them have failed and taking them to the right disposal site is a pain, so they just form my collection of dead CFLs waiting for surgery, i.e. someday I'll try to fix them. After all, the fluorescent element is perfectly recyclable. It's the inverter (the electronics that transform rectified AC into high-frequency current to feed the fluorescent tube) that fails, most likely due to a capacitor that is the weakest element in the whole assembly.

CFLs have saved us a bundle over the last 15 years. But we went through quite a few of them. LED-based light bulbs won't have any disposal problem and last virtually forever in a home. We do use a cheap LED bulb purchased at Costco as night light to see how long it will last (we wrote the date of purchase on it) with 8 hours of operation every day. That one can't be used for reading due to its 100-lumen output.

Friday, August 21, 2009

CFL versus LED-based Bulbs

CFL (Compact Fluorescent Lamp) have been very popular at Costco and Wal-Mart. They are heavily subsidized by the local public utility companies such as PG&E in the West Coast. Due to subsidies, they're good deal for the consumer. However, many inexpensive imported CFL didn't last as advertised due to attempts to save every penny out of their cost, from manual labor to cheap PCB (printed circuit board) made out of paper. What has been you experience with CFL life?

Many web sites such as Eco-Geek or EDN.com have regular articles and associated comments on CFL. There're a lot of hated-it and love-it, so what is the truth? Please comment on your personal experience with them.

But what about the economic aspect of CFL versus P-30 series of LED-based light bulbs? The P-30 series seem to have higher price tag so does it make economic sense to replace CFL with P-30 series bulbs?

For an extensive comparison between CFL and P-30 bulb, please send an email to the.led.expert@gmail.com for the report in PDF format.

Wednesday, August 19, 2009

China-made LED-based Fluorescent Tube Replacement versus TLED™

There’re many China-based LED-based fluorescent tube replacement manufacturers. Typically they put hundreds of small LEDs into a tube, and sell the tube with a universal power supply to drive the LEDs. What’s wrong with that picture?

First, with 276 small LEDs connected in series, all it takes for the tube to fail is any one LED fails, out of 276.

Second, small LEDs are designed and rated for 10,000 hours, not 100,000 hours for high-brightness LEDs that are more expensive. Due to their heat management ceramic substrate.

Third, a “flyback” power supply used to drive the LEDs will not last 3 years, due to the use of two electrolytic capacitors. When it dies, you have to look for a replacement, pay for its cost and its shipping and the labor cost to replace it. Not a good deal isn’t it?

Fourth, a flyback power supply’s efficiency is about 80%. You’ll be wasting 20% electrical energy in the form of heat. Energy means money in the long run.

Fifth, their power factor is typically 0.5, as opposed to 1.00 ideally. Green products are supposed to have high power factor > 0.9

Sixth, their overall lighting efficacy is not great:15 watts input for 1000 lumens for model # LJ-SF10-12-W15. That’s 70 lumens per watt.

Seventh, no optics to focus beams to prevent glare

Eighth, the power supply is to be installed in the light fixture. Do you know how to do it or are you willing to pay an electrician do do it for you at maybe $70/hour rate plus travel time?

Would you buy such a product and do it yourself? And repeat the process every other year, i.e. paying for all the costs all over again and again?

The model’s specification:

Part number: LJ-SF10-12-W15
Input: 110V/220V
Power: 15W
Luminous Flux: 1000 lumens
Color Rendering Index: 80
Color Temperature: 3000K-6500K (this doesn’t mean much. Correlated Color Temperature is not specified)
Lifetime: 30,000 hours (would you believe that number now?)
Housing: Glass/PC/Aluminum (Glass can break)

TLED specification:
Part number: TLED-1500 (others available)
Input: 110V
Power: 15W
Luminous Flux: 1500 lumens
Color Rendering Index: >80
CCT: 2700K (warm white) or 6000K (cool white)
Lifetime: >50,000 hours
Housing: Acrylic/Aluminum
Power Factor: >95%
Warranty Period: 5 years

Contact: 707-795-7567

PearlLED.com

Sunday, August 9, 2009

A typical LED product from China

Here's their ad, picked randomly from the web:

==============================================
LED Bulb (PAR30-E27-7W)
Product Description

3-year total unconditional guarantee
MKS-LED unconditional 3 year guarantee allows the consumer to return the broken or damaged light and get new lights immediately in-store, with no questions asked and no down time.
1% free
Furthermore, we supply 1% free light with your order if your order is bigger than 100sets.

Detailed product description:
Model number: MKS-PAR30-E27-7W
Base type: E27
Power: 7W
Input voltage: 85 to 260V AC
Light source: 7 x 1W high-power LED
Luminous flux: 500lm
Body temperature: Less than 48
Temperature: -40 to +70D
Brightness degradation: 3% to 5%/1kh
Lifespan: 50, 000 hours
Aluminum alloy curst and faster scatter thermal diffusion
No UV or IR radiation
Full range of colors: Cool white (6000 to 6, 500K), warm white (2, 800 to 3, 300K), blue, green, red and yellow
Applications: Replacement bulbs for general lighting, especially for museums, art galleries and cosmetic counters where UV or IR radiation is undesirable
============================================

My analysis:

Input voltage: 85 to 260V AC >> This means they use a flyback power supply (PS) to drive their LEDs. Such a flyback PS has at least two small electrolytic capacitors, one for filtering the rectified AC current, the other at the output for smoothing the output voltage of the flyback PS, hence the LED current. They dry up over time, therefore the LED lamp won't last long. Furthermore, its power factor of less than 0.5, which increase power transmission losses from the power plant. Flyback PS typically has lowest efficiency amongst all PS topologies.

Luminous flux: 500lm >> That is 50% dimmer than a conventional PAR-30 it's supposed to replace. Therefore it can't really replace a PAR-30

Lifespan: 50,000 hours >> This is an outright lie, due to the short life of their PS's electrolytic capacitors.

Conclusion: No one should waste their money buying this product.

Thursday, August 6, 2009

Indian Market

It's huge, it's starving for energy.

We signed up a company there to represent us in India, the New Delhi area to be specific.
They have 5 hours of black-out a day on the average, therefore they can use energy efficient LED-based bulbs.