Compact Fluorescent Eff. compared

tonyb

Enlightened
Joined
Aug 31, 2002
Messages
311
Location
PA
Does anyone know where compact fluorescents stand against HID and Led in terms of light ouput in lumens per watt. I heard HID is about 40 lumen per watt and leds are about 20 lumen per watt but what about fluorescents lumens per watt input power?

tony.
 

Ocelot

Newly Enlightened
Joined
Dec 31, 2003
Messages
115
Location
Corvallis, Oregon
50 to 100 lumens per watt for a fluorescent, depending on the bulb.

Note that CCFL would probably be lower.

High pressure sodium can be ~140 lpw (or so I've heard).

Scott
 

tonyb

Enlightened
Joined
Aug 31, 2002
Messages
311
Location
PA
Thanks for the info!
I had recently purchased a compact fluorescent flood light for outdoor use that was rated at 20watts and was surprised at how bright it was and began to wonder why I havn't made a bike light with a 10watt bulb and a custom inverter board. They seem to be almost as efficient as a HID but not as expensive.
 

PeLu

Flashlight Enthusiast
Joined
Jul 26, 2001
Messages
1,712
Location
Linz, Austria
There are millions of postings in bicycle forums about people who wanted to use fluorescent tubes for a bicycle light and all were abandoned. There was even a comercially made fluorescent bicycle light one time (with a 100V/6V dynamo).
 

NewBie

*Retired*
Joined
Feb 18, 2004
Messages
4,944
Location
Oregon- United States of America
CFL and CCFL are two different monsters.

The Cold Cathode Fluorescent Lamps I have used typically run from 75-90 lm/w. They are typically used in TFT displays, and often the designers use a metal reflector wrapped right around the lamp, which, once the plasma inside gets going, capacitively couples alot of the energy right to ground, which adversely affects the efficiency.

For an example, see http://www.linear.com/pdf/an55fa.pdf figure 16.

An interesting company for CCFL bulbs is http://www.lcdl.com/ultrabrite4_1.html
 

mattheww50

Flashlight Enthusiast
Joined
Jun 24, 2003
Messages
1,048
Location
SW Pennsylvania
The gold standard in HID's is 100 lumens per watt. There aren't a lot of them that actually make 100 lumens per watt, but most of them are pretty close.
HPS lamps generally do better, the bigger they are, the better they are. 1000 watt GE Lu1000 is 140 lumens per watt, most do better than 100 lumens per watt.
The highest efficency (and worst color rendition) is LPS,
the best can put out about 200 lumens per watt.
Fluoro vary considerable, depending upon the color rendition you want, and how much money you want to spend
For example a GE Chroma 75 has a CRI of 92 but only gives you about 45 lumens per watt in an F40, while an SPX35 in the same F40 is about 90 lumens per watt (and you will definitely pay for the honor of having an SPX series lamp). Most of the CFL's are around 60 lumens per watt.

In general CFL's and Fluor's don't care about orientation, and in general HID's care. A HPS, LPS, Fluoro or CFL or incandesent generallt doesn't care if it is base down, base up, vertical or horizontal. Many HID lamps have base up versions ,base down versions, and horizontal burning versions. You usually get less output from the Horizontal versions. Lamp life suffers considerably (as does output) from the wrong orientation.

As for Bicycle use, Fluoro and LPS envelopes tend to be 'delicate',
HID and HPS envelopes tend to be pretty robust because conditions inside the arc are pretty ugly. You also need a lot of surface are for a fluoro. A 35 watt D2S is pretty small. The surface area of fluoro putting out about the same number of lumens is about 1 square foot! (1 inch diameter x 48 inches long)
 

Ocelot

Newly Enlightened
Joined
Dec 31, 2003
Messages
115
Location
Corvallis, Oregon
FYI, Doug Strait (who posts here, and was the co-designer of the Li4430 light) designed a fluorescent CCFL caving headlamp quite a few years back. I believe it was a state-of-the-art light at the time.

It was quite bright, and had a custom reflector to focus the light (one of the difficulties of CCFL as I recall). He had a custom driver circuit for high efficiency and clean startup. I once looked at buying one from him, however they were pretty expensive at the time, and he recommended powering it with 8 D cells soldered together in order to cut out IR losses and to get a long run time. Since I like helmet-mounted batteries or a smaller hip pack, I decided to forgo it.

Doug has since switched to LEDs even with their lower efficiency; perhaps he can let us know some of the nitty gritty details of CCFL (or CFL) design?

Scott
 

Doug S

Flashlight Enthusiast
Joined
Jun 20, 2002
Messages
2,712
Location
Chickamauga Georgia
[ QUOTE ]
Ocelot said:
FYI, Doug Strait (who posts here, and was the co-designer of the Li4430 light) designed a fluorescent CCFL caving headlamp quite a few years back. I believe it was a state-of-the-art light at the time.

It was quite bright, and had a custom reflector to focus the light (one of the difficulties of CCFL as I recall). He had a custom driver circuit for high efficiency and clean startup. I once looked at buying one from him, however they were pretty expensive at the time, and he recommended powering it with 8 D cells soldered together in order to cut out IR losses and to get a long run time. Since I like helmet-mounted batteries or a smaller hip pack, I decided to forgo it.

Doug has since switched to LEDs even with their lower efficiency; perhaps he can let us know some of the nitty gritty details of CCFL (or CFL) design?

Scott

[/ QUOTE ]
Scott, I must say that you have a remarkable memory. I believe that you may have set in on a presentation that I gave over 10(?) years ago. To set the record straight, one of that small batch of prototypes [HCFL, not CCFL] built in 1991 is *still* my primary light. In fact, I was underground for 9 hours just yesterday with it. For efficiency in a 2.5-5W range headlamp, LEDs still cannot match it. I am eagerly awaiting the day when LEDs catch up. I don't think it will be long.
As to compact HCFLs, the three efficiency figures I remember off the top of my head are 50 lm/W for the PL-5 (S-5) 5w twin tube CFL, 57 lm/W for the PL-7 (S-7) 7W twin tube CFL, and 67 lm/W for the PL-9 (S-9). These are lamp only efficiencies. Losses in the lamp ballast can vary widely with design. In general, all else being equal, efficiency increases with tube length. This is true for both CCFL and HCFL. The reason is that the (non-productive) voltage drop at the cathodes [called "cathode fall" by FL lamp engineers] is fixed regardless of length. For a given tube length, HCFL tends to be more efficient than CCFL. This is because HCFL cathode fall is usually less than 10V per end and on a CCFL it is usually greater than 50V per end. Back in the early 90's when I was heavy into engineering with HCFL, CCFL with the 75-90 lm/w quoted by Newbie above did not exist, at least not in "reasonable" lengths. Those efficiencies with early 90's technology were only possible with *very* long tubes. If those efficiencies are now available in shorter CCFL's, we have the laptop computer industry to thank.
@Newbie: do you have any links to Manufacturers' data showing those 75-90 lm/w CCFLs?

@Scott: I went to helmet mounted battery packs about 5 years ago for my Flourescent. I get about 8 hours runtime on "Low" [2.8W] from a pack of 4 18650 Li-ion cells.
 

PeLu

Flashlight Enthusiast
Joined
Jul 26, 2001
Messages
1,712
Location
Linz, Austria
I remember that time we discussed the use of CCFL lamps there efficiency was pretty poor, the smaller ones even lower than halogen bulbs (especially halogen bulbs with an Willie LVR). So I'm somewhat puzzled reading form NewBie
[ QUOTE ]
The Cold Cathode Fluorescent Lamps I have used typically run from 75-90 lm/w.

[/ QUOTE ] Doug, you probably remember that several cavers wanted to use this CCFL as a caving light and we told them not to do it.

A completly other story are the HCFL. IIRC, I made my first one with a 4W tube somewhen in the 70ies, the circuit for it was published 1970 in Elektor (famous Dutch electronics magazine). I think, I used a germanium AD130 for better efficiency.
Unfortunately I was never able to get one of your 'Awesome Lights' (was it the name?), but there was also the NevTec I almost bought. It was a little bit compareable with your light.
The main problems always were that the smaller tubes have/had lower efficiency and it dropped at lower temp. And many implemantations were fragile.
Sorry for digressing.

NewBie wrote further:
[ QUOTE ]
An interesting company for CCFL bulbs is http://www.lcdl.com/ultrabrite4_1.html

[/ QUOTE ]
I bet these are not CCFL tubes (but I may loose here .-)
 

NewBie

*Retired*
Joined
Feb 18, 2004
Messages
4,944
Location
Oregon- United States of America
[ QUOTE ]
PeLu said:
I remember that time we discussed the use of CCFL lamps there efficiency was pretty poor, the smaller ones even lower than halogen bulbs (especially halogen bulbs with an Willie LVR). So I'm somewhat puzzled reading form NewBie
[ QUOTE ]
The Cold Cathode Fluorescent Lamps I have used typically run from 75-90 lm/w.

[/ QUOTE ] Doug, you probably remember that several cavers wanted to use this CCFL as a caving light and we told them not to do it.

A completly other story are the HCFL. IIRC, I made my first one with a 4W tube somewhen in the 70ies, the circuit for it was published 1970 in Elektor (famous Dutch electronics magazine). I think, I used a germanium AD130 for better efficiency.
Unfortunately I was never able to get one of your 'Awesome Lights' (was it the name?), but there was also the NevTec I almost bought. It was a little bit compareable with your light.
The main problems always were that the smaller tubes have/had lower efficiency and it dropped at lower temp. And many implemantations were fragile.
Sorry for digressing.

NewBie wrote further:
[ QUOTE ]
An interesting company for CCFL bulbs is http://www.lcdl.com/ultrabrite4_1.html

[/ QUOTE ]
I bet these are not CCFL tubes (but I may loose here .-)

[/ QUOTE ]

The ones in the picture are HCFL, you can see this by the dual wires on each end, to supply power to the filaments. They also make them in CCFL, and even serpentine. I've used them before (over 20,000 of them).

BTW, you'll find LCDL's bulbs used here:

- Airbus Corporation  
- Applied Display Technology  
- Astronautics Corporation
- BAE Systems 
- Barco View   
- B/E Aerospace  
- Boeing  
- Bruce Industries
- Eaton Corporation  
- Elbit Systems Ltd.  
- General Dynamics  
- Goodrich Avionics  
- Honeywell International  
- Interstate Electronics  
- Kaiser Electronics
- Korry Electronics  
- Lowrance  
- L-3 Communications   
- Meggitt Avionics 
- Northrup Grumman  
- Parker-Hannifin  
- Rockwell Collins   
- Rogerson Kratos   
- Thales (Sextant)  
- Smiths Industries  
- Symbolic Displays   
- UPS Aviation Technologies
- Universal Avionics Systems Corporation  
- White Electronic Designs Company

Commercial Aircraft:
- MD-11 
- MD-90 
- MD-900 
- B-727 
- B-737 
- B-767 
- B-777

Military Fighter Aircraft:
- F-15 
- F-16 
- F-18E/F 
- F-22 
- F-111 
- EFA-2000

Military Cargo Aircraft:
- P-3 
- KC-135 
- C-130H 
- C-130J 
- C-141

Helicopters:
- Bell 412 / 430 
- CH-46 
- CH-47 
- AH-64D 
- RAH-66

Army Ground Vehicles:
- Bradley Fighting Vehicle (M2-A3)

Space Shuttle:
- Flight Deck Displays





I've also used these HCFL, which are based off the Candescent and Winsdor lamps:

80W @ 10,000 footLamberts (34260 nits)

http://www.thomaselectronics.com/avionics/10f024_overview.php

FYI, they also do a number of custom sized units, as well as a CCFL version.

10f024.jpg



Another flat lamp is made by OSRAM/Philips/Sylvannia, but it has some drawbacks (and advantages). The PLANON source generates light using a xenon excimer discharge. The efficacy of PLANON systems is currently on the order of 24-28 lumens per watt (lpw). Long life ? 100,000 hours

http://www.sylvania.com/press/05302001b7.html
http://www.sylvania.com/business/planon/
http://www.sylvania.com/business/news/pdfs/planon_tec.pdf

05302001b7a.gif



Also, PeLu:
Take a look at this app. note. I've seen a good 25% loss in the "reflector" interacting with the electrical, and another 20% loss in the inverter, that could be dialed out.

http://www.linear.com/pdf/an65f.pdf


Interesting, here are some really odd looking CCFLs rated for 70.7 lm/W.
http://www.nationalcathode.com/
 

Ocelot

Newly Enlightened
Joined
Dec 31, 2003
Messages
115
Location
Corvallis, Oregon
[ QUOTE ]
Doug S said:
Scott, I must say that you have a remarkable memory. I believe that you may have set in on a presentation that I gave over 10(?) years ago.

[/ QUOTE ]

No, I just remember it from a 3-day Lech trip quite a while back.

I always thought CFL-type lights were the way to go based on efficiency, however even on low the lights were very bright (couldn't dim them down any more), and the beam angle was a bit too wide for my tastes (although that could probably be addressed via the reflector).

I often cave with <1W of light, because the cave is either light enough or small enough. I rarely run with 2.5W with an LED, unless in larger lava tubes. Although, with 2x greater efficiency at 2.5W, you are getting 1.25W "free" compared to an LED...

Scott
 

NewBie

*Retired*
Joined
Feb 18, 2004
Messages
4,944
Location
Oregon- United States of America
[ QUOTE ]
Doug S said:
Back in the early 90's when I was heavy into engineering with HCFL, CCFL with the 75-90 lm/w quoted by Newbie above did not exist, at least not in "reasonable" lengths. Those efficiencies with early 90's technology were only possible with *very* long tubes. If those efficiencies are now available in shorter CCFL's, we have the laptop computer industry to thank.

@Newbie: do you have any links to Manufacturers' data showing those 75-90 lm/w CCFLs?


[/ QUOTE ]


Here you go, they are available in U, Serpentine, and straight. You can run an array at low power on each tube for some rather high efficiencies. FYI, stray capacitance to the chassis will kill the efficiency. It is also often not apparent at first to folks, that peaking the inverter (typically a "royer" based design) efficiency is usually not the peak power input to light output efficiency point. Jim Williams of Linear Tech fame has lots of great app. notes regarding this.

ccfl.jpg
 

Doug S

Flashlight Enthusiast
Joined
Jun 20, 2002
Messages
2,712
Location
Chickamauga Georgia
Newbie, Thanks! Those are good efficiencies. From the drawing the tube length looks to be about one meter. That qualifies as *very* long to me.
 

PeLu

Flashlight Enthusiast
Joined
Jul 26, 2001
Messages
1,712
Location
Linz, Austria
Thanks NewBie for all that info. Although this tube looks really a little bit large for caving use .-)

BTW, Doug, could you post an image of one your caving lights. I'm shure several would be interested.

After not looking at flurescent tubes for a while, I'M amazed how much I missed. And I wonder how the LED manufacturers want to beat them in efficiency for LCD backlights (There are similar messages from time to time in the news).
 
Top