Those not so green light bulbs

The energy savings are real, but I'm not so sure how "green" they really are. Some day in the future there will undoubtedly be a big uproar about all the extra mercury and phosphor ending up in our landfills and ultimately our water supplies when all these compact fluorescents end up there.

And yes, I know that by reducing energy we reduce emissions from power plants which also contains mercury if they were coal fired, but in many areas the power is not coming from coal, so the extra mercury is just extra. Where I live in California is an example. There is serious discussion on banning incandescent lighting, yet we have virtually no coal fired power plants here, so all of the extra mercury that will end up in our environment represents toxins that are not already here.

The MERCURY and PHOSPHORS are already here, all we are doing is utilizing them.

Yes, there is a problem with recycling, but it is really with the end user.

Most areas have a recycling program, may cost, but it IS THERE.

LED (light emitting diode) lamps already beat compact fluorescents in lumens/watt. They are still more expensive, but in 4-5 years they will mature to the point wher they kill them.

100,000 hour life = no more need for sockets. 35 tears of 8 hours/day 24/7

They actually do not fail after 100,000 hours

socketed LED lamps can be bought now to retrofit tungsten lamps. In time lamps will be made with no ability to change them

Yes, LED is the future light source.

I do not know if gallium arsenide is an environmentally friendly compound?

Still, I do not like the colour, there is a lot of near UV that make some objects a bit fluorescent.

Research is moving at an amazing pace for LEDs.

Well, never mind the mercury for a moment. What about all the substances in coal? Try running a Geiger Counter over a coal pile and see what happens...

"How many Californians does it take to change a light bulb?

50.

One to do it and 49 to share the experience!"

"How many Country and Western singers does it take to change a light bulb?

5.

One to do it and 4 to sing about how good the old one was!"

The best thing to do is recycle the bulbs, henceforth no mercury polution, except from the coal fired plants.

Better to rule in Hell, than to serve in Heaven !

Silver/mercury amalgam fillings are now considered a risk due to the mercury.

Mercury caused hatters to go mad.

Float glass used to be poured onto mercury but is now formed on meted tin.

Home owners will break bulbs, the mercury will not be properly cleaned up and end up in the carpets and cracks of the floor. What is the long term effect of mercury on the brains of developing children? What if the mercury from a broken bulb ends up in the food that is being prepared? It will happen. It is only a matter of when and how often. Those 4ft fluorescent tubes are required to be coated for food handling areas. How about the compact bulbs?

mercury boils below the temperature glass melts at.

http://hypertextbook.com/facts/2002/SaiLee.shtml

mercury did do that, but not lately.

no known risk from mercury amalgam. Just internet sites suspect that and scream about it.

I checked, you are right.

Now, I have a confused memmory of a picture in a textbook of a fellow lying ontop of a mercury filled tank . I thought the caption was with regards to making floatglass, but Pilkington made float glass in 1952-59, and the book could have easily been older than that since it was my grandfathers. Any ideas?

Back in the good old days people with money had mercury pools to float on. Then they found out about the toxicity and stopped that. Same with hatters.

float glass cheapened glass a lot as it was optically flat without the cost of grinding, and Pilkington grew enormously

"Float glass cheapened glass a lot"

Only high-quality flat glass. To this day, my greenhouse does not use float glass.

"and Pilkington grew enormously"

The glass industry used to be dog kick (but never quite eat) dog. The following story illustrates:
In the interests of keeping the float glass process secret until they were ready to announce (so they claim), Pilkington didn't delay acceptance of orders for orders for a single grinding machine - and they proceeded to manufacture and deliver those machines.

Subsequently, Pilkington seemed either to rest on their laurels (or perhaps rather to lose the plot) for a while, diversifying into areas they didn't really understand (blame BT for encouraging them??). They seem to be more-or-less back on track again.

Fyz

Pilkinton float glass uses Tin to float the glass on.

Here's a simple test to "get a feel" for the difference. Go to a lamp that has been on for a while, turn it off and grab hold of it. I suggest you only do this with a CFL since the 100 watt incandescent will burn you. Maybe you just wave your hand near the 100 watt incandescent. Then tell me which one is using more energy! Also think of this, if you use air conditioning you get the additional savings of not having to remove that extra heat.

In my bathroom, I have a fitting that had 8 x 20 watt GU10 socketed 230 volt halogen lamps, you just needed to put your head up near the ceiling to find a layer of super hot air!!!

I have replaced them with cheap GU10 230 volt LED Lamps from ebay at about €4 each!!! Each lamp takes 1.2 watts of power.....not as bright as the halogen, but more than bright enough!!

Ha, that really gets to me...

In a cold country...that 'waste heat' isn't waste , it's part of the heating, same way as heating engineers allow for the 500w (ish) per person.

And being bald (ish) the nice warm feeling on the top of my head when I'm getting my breakfast under those spotlights is just bliss in the winter!

So tungsten in winter, LEDs in summer maybe?...don't let the marketing men get hold of that idea, they'll be selling us all twice as much!

In Canada we heat our homes for 8 months of the year. Turn out the lights and turn on the electric heat. It is a wash.

For the air conditioning months it makes sense.

Natural gas heated homes are CO2 unfriendly. Electric power from nuclear, hydro dams, wind etc looks like a far better deal.

How much mercury from broken bulbs does it take to damage a child's developing brain? That mercury will be in the carpet or floor boards for years, slowly evaporating into the air we breath.

We go nuts over lead paint (how much paint do you eat?), asbestos, mercury batteries (now banned), watches that glow in the dark (banned), and genetically modified food, but suddenly it is ok to have a broken bulb with mercury deposited into the family carpet. Some people will become "mad as a hatter".

Formaldehyde insulation is banned, but our bodies produce copious amounts daily.

Rome thought they had a good deal with lead pipes!

Reminds me of the power smart blitz at a school. Normal procedure, turn out the lights and use an overhead projector. Power smart(?) way, turn on all the lights and write on the black board with chalk. Duh!

I find it hard to separate "power smart" from "political smarts".

Just call me skeptical.

I think that the new fluorescent lamps with mercury (the old ones had it too by the way), should get a strong but thin coating of flexible plastic over the fluorescing part, and that connects/sticks on to the base, that will contain all the glass and chemicals in the even of damage.....like a see thru plastic bag.

I hope that the design people are busy designing the mercury out of these lamps as quickly as possible too....

They will never design the mercury out of a mercury vapor lamp, which is what all fluorescent lamps are. That's what makes them work. They have reduced the amount of mercury in them. Once I came upon two high intensity mercury vapor lamps used in an old "Photostat" machine, the predecessor to modern photocopy machines. I harvested about three pounds of mercury from these lamps.

I think you had the best solution of protecting the lamps from accidental release and what will soon obsolete them are LED lamps. That will in all likelihood the technology that will take another 100 years to obsolete, or at least 10. With regard to the toxic elements in LED's it is encapsulated so perhaps somewhat less of an immediate concern. Maybe we should put some gold in them so that it makes it worthwhile for someone to reclaim.

Just another line of thought: I have recently been testing in my home for levels of electromagnetic radiation and it may interest you all to know that

compact flourescent lighting increases the EMR levels by four or five times

that from tungsten or LED's. In fact a compact lamp of 11W rating registered

more radiation than from my microwave [ at the same distance]!!!

Actually is not the same, Radiation at microwave frequencies is far more damaging than slightly supersonic frequencies that these bulbs use they work around 50Khz but they emit lots of armonics at decreased power levels.

Hi jondowne. I agree with you, here in the UK we have a group called Power Watch, and they are trying to get us to reduce the EMR in our homes. I know many sufferes from the elecro-smog are affected by these so-called energy saving light bulbs, but by 2010 we in the UK will not be able to buy any form of incandescent lighting by law! The problem as I see it is that our politicians have jumped on the green bandwagon to gain votes. These so-called energy saving bulbs affect my eyesight so much that after 2010 I will have to smuggle into the UK as many incandescent light bulbs as I can, thus if caught I will be labelled a criminal. The world has gone crazy. Spencer.

It's not the fluorescence as such - it's the 50Hz flicker. At the moment I'm mixing 50% incandescent and 50% fluorescent, which seems to work for me. Once you run out of incandescent lamps, and unless/until lamp manufacturers introduce bulbs with built-in high frequency ballasts, the only way I know around this will be to buy high-frequency fittings and separate bulbs.

.....or LED lights!! They are getting really powerful nowadays!

LED lighting will also need suitably designed controllers to avoid low frequency flicker. In fact, they could be even worse, because the response of LED's (current in to light out) is very fast indeed, and at least a proportion of the phosphors used for fluorescent bulbs have a noticeable smoothing effect.

This gives reason for concern that, if radiance measurement standards are left as they are, LED lamps that are suitable for direct connection mains supplies will be no better than the worst existing fluorescents - even though it should be cheaper to provide smoothing because of the reduced total dissipation

Fyz

LED's operate on DC

No flicker

Hi Techno

Please provide quote for two hundred million units, 10-Watt DC Voltage transformers, soonest.

Thank you for your attention.

Fyz

(That should drive Canada's LED requirements for a few months.)

A LED is a DC device, therefore the AC must be rectified and hopefully smoothed, therefore you should not see any flicker - if the manufactúrers do their homework correctly!

Historically, the efficiency of LEDs rises as current rises - up to some optimum after which it degrades again. I've no idea if this remains true for present-day lighting designs. But fluorescents can also be run quasi-DC - or even better at 400-Hz. But it saves money and results in a smaller base for the bulb if you don't do this, so pulsing miniature fluorescents are what is available.

Even if the LED doesn't require pulsing, you need to ask yourself why the manufacturer would apply smoothing if the standards don't give him any market advantage. However, there may be light on the aurizon - LEDs lose efficiency (and durability) if they are pulsed so slowly that the temperature rises significantly during the pulse - this may work in our favour

Fyz

Fluorescents work perfectly on DC, except that one end of the tube, due to the arc starting from the same end, and the electrons flowing from one end all, one end wears out first, which is why on DC Royal Navy ships 40 odd years ago, an electrician goes around and swaps the polarity of the tubes each day. this almost doubles the life of the tubes.

Its only a switch that reverses the polarity to a relatively large area of lighting....on a carrier, it takes about an hour to do the job as whenever possible, the switches are grouped together....

Modern camping fluorescents convert the DC to high voltage AC first, but for some reason, one end of the tube is always blacker than the other, which is why I swap the ends once a season!

Yes, that is true. But, SFIK, one blackened end is not necessarily a sign of DC excitation either - small asymmetries in the tube can grow towards the end of life.

BTW, which post were you really replying to - #49 was about LEDs

Fyz

Read #49 again, it mentions DC and fluorescents!!!

So it does - my oversight.

Fyz

No problem.....

The idea is to pulse the Led's on off at a high frequency so as to get more out put for a lower heat out put. Above 1 watt these things get very hot and require big heat sinks so they are not nearly so energy efficient as made out 40% at best.

You must have very unique eyes if you can see a 100 or 120 Hz flicker. It must make watching Television really hard since the frame rate is the same (100 or 120 fps depending on country).

Most people's retina has a reticence of about 1/20 sec. Faster than that cannot be resolved.

Hmm, don't see the flicker until I move something, like wave my hand, then it really is annoying.

I don't like the color balance either. Have you ever tried to match socks that were close in color? Try different light sources, and then you run out and buy new socks.

Yeah I solved the sock problem years ago. I have black socks and white socks. When I can no longer match those it simply won't matter what is on my feet. Anyway, this year I made it through winter wearing sandals so socks aren't as important as they once were.

"then it really is annoying."...perhaps you should try a decaffeinated coffee?

Not you too. My wife says I need to back off the coffee. She says everyone gitters when I have had about my 6th mug. I think they all just get difficult.

I do not panic when socks are not of exactly the same colour.

I do not mix white and black, at least up to now! When I am 80 years old in 20 years time, I am sure that even white and black will be acceptable too! I will just be happy to have warm feet!!!! get your priorities sorted......

Anyone who worries about certain people noticing the slight difference in colour, should wonder WHY the person concerned is checking his socks out and not looking him straight in the eye!!!

I am sure that the colour temperature of new lighting can be adjusted, but do not forget that normal bulbs have a poor colour temp if my memory serves me correctly, but we are all used to it!!!!

Your facts are partially correct, but largely irrelevant. First, the difference between the response of the rods and the cones - you can certainly see the flicker of a traditional television out of the side of your eye; that doesn't matter too much because it is designed for you to look at directly. But lamps are designed to illuminate your entire field of view. Another feature is that, contrary to some claims, the alternate half cycles do not produce the same output; assuming you have no means for measuring the flicker, this may be deduced by inspection of a used lamp. There is also a measure of randomness for each arc, whose perceived effects are much improved with higher frequency operation

However, these not by any means the sole effects that make it tiring to work under standard (low frequency) fluorescent illumination. The problem is that the eye works by scanning. What you see in the centre of the field of view doesn't flicker, but is strangely broken up (similar to what electrical engineers call "aliasing"). The information processing in your optical system then interprets this as best it can, but the processing load and the degradation in the quality of the information are indeed tiring.

Television is effectively a series of still pictures, so the processing effects are somewhat different - but even then, there is good evidence that the framing has significant effects; for example, the scanning behaviour of the eye is known to change when watching traditional televisions, and it has been claimed that this is responsible for the hypnotic effect of many television receivers (I am not however familiar with the evidence in this field - you would need to consult an ophthalmologic neurologist - assuming such a beast exists).

Fyz

How about completely irrelevant? I just measured the actual light output of some CFL's and the output is a constant level with a superimposed AC signal that has a frequency of between 60 kHz and 90 kHz. So it never goes to zero and the AC that is imposed modulates the light about +/- 15%. If your rods can detect that kind of "flicker" on top of a constant output I am impressed. I suspect that the reason it doesn't go to zero output is that the phosphors have a persistence that is longer than the pulse rate of the high frequency high voltage converter that is driving the tube.

If the light modulation spectrum really does include insignificant components below 800-Hz, then "completely irrelevant" will do nicely". But...

There are certainly compact fluorescents available (at least in Europe) that use "traditional iron" ballasts rather than electronic ballasts. These cannot fail to generate 100(120)Hz flicker - and probably also generate a noticeable 50(60)Hz component.

HF compact fluorescent ballasts inevitably generate components at high frequency - typically somewhere between 40kHz and 300kHz, and this should be somewhat filtered by the response of the phosphors. In any event, this is not a problem to the eye. Nor should it be a problem for IR controls for televisions etc. unless there is some additional modulation component. The fact that televisions have been affected suggests that there other frequency components present (modulation of the modulation).
The most likely cause for lower frequency modulation of these HF ballast lamps would be the supply from which the HF oscillator is driven. This could be a rectified-and-smoothed mains supply - in which case a visible (100or120Hz) component should be expected due to limited smoothing. I can believe that some CFLs are really well smoothed as you have measured (scope? LF modulation analyser?); but I would be surprised if that applied to the majority of price-constrained product that most people end up using - if only because that is what is available locally. Did you look at the cheaper ones yet?

BTW, did you look at the output on a slow scope sweep as well as whatever else you did?

Fyz

Yes, no significant low frequency modulation. Setup was a fast visible spectrum photodiode into a 100mHz digital sampling scope.

PS Not having suitable measuring equipment to hand, I took a look inside a dead HF ballasted 11-watt CFL. The most significant energy storage that I could find was a 10-uF, 150 Volt capacitor. That would store at most 0.11 Joules. As we are 50-Hz here, that means that this capacitor could barely store enough energy to run a full half cycle, let alone smooth the supply for that long. It's more than possible that I missed some subtlety, but I couldn't find anything that looked likely to stabilise the output against the inevitable supply ripple.

I'm almost at the point of begging to borrow an ex-colleague's measurement equipment...

Fyz

The function of a common wire ballast is not to step up the voltage, but to limit the current by inductive reactance to the optimum value for the bulb.

The circuit you see is the electronic analog of that, possibly a simple recurring on-off for a portion of the cycle while V is high enough to fire the arc.

A scope and a current probe should reveal this.

That was my understanding too - thank you for clarifying it. What I am trying to understand is how (if) the light output of a HF ballasted CFL can be maintained constant (apart from ripple at > 40-kHz) over a half cycle of the mains input. This is what some manufacturers' literature appears to claim, and also what rcapper measures. To my mind, this needs either:
substantial energy storage and smoothing, or
adequate storage with a modicum of smoothing together with some means to maintain a constant energy* in each HF cycle

Am I missing something fundamental, or is it that the CFLs that rcapper has measured have a controller that is missing from the samples that I found irritating and so examined?

Fyz

*I suppose that peak current would do as the measure if you are using an inductive ballast - there would be some phase modulation, but this would correspond to a very low level of 100Hz/120Hz ripple. Presumably this would be invisible show on a scope unless youused a low-pass filter?

you get what you pay for. The better ones make a high rail of AC at 20 Khz and then only a small inductor is needed. Cheaper ones do other things.

Both waste electricity to a degree.

A bit of coverage here.

http://en.wikipedia.org/wiki/Electrical_ballast

http://www.google.com/search?hl=en&q=%22electronic+ballast%22+%2Bfluorescent+%2Bwiki

"The better ones make a rail of AC at 20-kHz" (between 20 and 100-kHz, I'd have said). So you could tell I got that far. The example of a compact fluorescent ballast shown illustrated in the Wikipedia article is exactly the sort of thing I was describing. It could make a rail at 20-kHz - but not apparently at a constant level rail.

It shows a transformer, a small inductor, a few passives, a bridge rectifier, a storage capacitor, a few miscellaneous passives. The storage capacitor is nowhere near large enough to do much smoothing. Would there be control circuitry hidden somewhere (if so, what a nightmare to assemble).

How have you determined that "The storage capacitor is nowhere near large enough to do much smoothing."? If you do not fully understand the operation of the switching supply circuit, it would be hard to say that it doesn't. For example, a proper switcher could create a constant output by varying it's pulse width as the voltage on the source capacitor decays. Were you to measure the ripple on the cap, you would see a large 100/120 Hz component but that does not necessarily translate to a similar component in the output. I say that as someone with experience in designing numerous switching power supplies, from the ground up.

"How have I determined that the capacitor is nowhere near large enough to do much smoothing?" It can store a maximum of 125mJ. 11-Watt for 100-ms is 110 Joules, so it would need to be discharge quite deeply. Of course, there is a period where the current is taken from the supply, not the storage capacitor. Even accounting for that, and assuming the power drawn at the output of the capacitor is only 9-watts, I'm sure you too will find that the voltage on the 10-uF storage capacitor has to fall from 150-Volts to 100-Volts in each half cycle (at 50Hz supply).

Under these conditions, I think you will agree that something would be needed to control the power dissipation in the lamp. I appreciate that could simply be down to pulse width modulation of the driver. The problem is that the lamp ballast I disassembled definitely didn't have a controller that would do this. I'm not so certain about the ballast that aurizon referenced on Wikipedia - it's possible that the DIL on it is such an IC, in spite of the apparently small number of pins.

As I implied in the first instance: it didn't appear that all CFL manufacturers were designing for low LF ripple, and this resulted in many of the lamps being tiring to work with.

I would now add: if has become economic for all lamp manufacturers design for minimal LF ripple, the problem should go away; unfortunately, in the mean time, the technology has acquired a reputation for apparent dimness and being tiring for many early users. Denying that there ever was a problem (or alternatively that the problem was trivial and irrelevant) is not going to persuade any of these people to give CFLs a proper second try.

Regards

Fyz

well, the high rail will be a DC rail with ripple. From this the circuit will draw pulses to run the bulb and have a small inductor to limit the current. In a typical circuit the tube would short the rail after the inductor, thus extinguishing the lamp and allowing the voltage to increase until the tube conducted again. This form of oscillation may well be independent of the switcher that made the rail and as I think of it you might be able to make a simple full wave rectified high rail from 120 or 240 VAC and simply let the plasma tube auto-oscillate. This simple oscillating mode might be less efficient than a proper feedback regulated switcher to run the tube?

Running on 12 voltes, you would need a full fledged switching power supply

SFIK, auto-oscillation circuits either delivers output power that follows the supply (more likely V^2), or need a high Q resonance in the oscillator to filter the response. High Q corresponds to storing the energy in reactive components through several cycles - and the components used simply do not have sufficient energy-storage capability. rcapper's described solution of using pulse width modulation on the driver will definitely solve the problem - but there was no possibility that this was being done on the lamp I disassembled. It is possible that the DIL package on the CFL in your Wikipedia reference is a controller switch.

The difficulty now is to persuade people who have had bad experiences with CFLs like the one I examined to give them another try... Admitting there has been a problem (now corrected) seems the least-bad route to me.

Fyz

the rail is followed by an inductor followed by a capacitor followed by the lamp, and this forms a relaxation oscillator. As soon as the V is enough to fire the bulb, it fires, and the voltage then drops and the lamp goes out because the inductor has too much reactance for a large current to maintain the lamp.

Would you care to back that up with a model* that can be analysed? Without that detail it sounds as if it will be even more dependent on power supply Voltage than a simple linear circuit.

Fyz

*Circuit would do - I've found SPICE to be a convenient analysis tool for all sorts of systems over the years

the model is the fact that the arc strike and the arc extinguish are at differemnt voltages.

As you increse the voltage nothing will flow until the arc starts(is struck) at which point it drains the capacitor and the inductor blocks a large current flow through the arc and when the capacitor is drained the current is not enough to sustain the arc...whereupon it extinguishes and the process repeats. Changing the inductor and capacitor valus can change the frequency of this.

I tried a quick and dirty simulation with a first approximation of a schematic based on my understanding of your description (series hysteresis-Voltage controlled resistor, capacitor and inductor driven from a sine-wave Voltage source). The result was a variation in drive power to the CFL lamp of 25% for a supply variation of 10%.

Obviously, something is missing from my understanding. I would need to see a circuit that self-regulates to be convinced.

(The lamp I disassembled certainly exhibited LF (~100-Hz) flicker - demonstrated with a spinning top, so that was no use for extracting useful values)

Fyz

well, a relaxation oscillator is a de-minimis type of thing that can be set up to make a fluorescent tube flash at some rate.

To see a few variations on this look here.

http://www.google.com/search?hl=en&q=neon+%2B%22relaxation+oscillator%22&btnG=Google+Search

possibly with full wave rectification and no capacitor you can get a several flashes in each excursion above zero by using the inductive kick back to start each flash,

That's a raft of references. You are the person who claimed this type of system could provide stability. It's not down to me to search for one that might do what you claimed. So please tell me which if any show the stabilisation effect you are talking about.

N.B. Just in case they were more relevant than first appeared, I looked at the first few - they would all produce a stream of pulses that get somewhat longer with increasing supply Voltage, and whose frequency depends (reasonable closely) on the difference between the supply Voltage and the average of the strike and extinguishing Voltages. That represents a supra-linear power-out vs supply characteristic.

Fyz

Stability?? I would be the last to maintain that a relaxation oscillator is stable. It is one with a minimal part count and thus sacrifices efficiency and stability.

I commented that it was possible to make very simple circuits that would have low efficiency and low cost, and that it was possible these low cost circuits could be part of the problem with the early mad in China lamps.

By now they have a full understanding of how to manage this type of lamp and the design has devolved to a control chip and the minimum parts count for high efficiency and long life. Lack of standards enforcemen accounts for part of the problem. In Toronto you can get these lamps with FCC, CSA and U/L stickers for many different prices. SOme of them are genuine, many are counterfeit and the bulbs have recently been in the papers about fires being started and the discovery of their faked credentials followed.

Silly me - I was assuming that at least some your replies were relevant to the answered posts - i.e. (working backwards):

93 was in reply to 92: "The result was a variation in drive power to the CFL lamp of 25% for a supply variation of 10%."

90 "the model is the fact that the arc strike and the arc extinguish are at different voltages" was in reply to 89: "Would you care to back that up with a model* that can be analysed? Without that detail it sounds as if it will be even more dependent on power supply Voltage than a simple linear circuit."

87 was in reply to 85: "output power that follows the supply (more likely V^2), or need a high Q resonance in the oscillator to filter the response"

83 "This form of oscillation may well be independent of the switcher that made the rail" was apparently in reply to 80 "The storage capacitor is nowhere near large enough to do much smoothing."

Fyz

I believe the problem you would find is that the fluorescent tube requires more than 180 volts to strike an arc. It will sustain on that or less but it won't fire, which is the purpose of standard starter circuits. With a switching controller you can feed a train of high voltage pulses so the bulb fires and extinguishes at the 60-90 kHz rate. To make the relaxation oscillator scheme work you would require a higher voltage source.

So the UK and europe with 240 VAC could do it, but the USA/Canada would need a 'tickler', with a small capacitive doubler to strike the arc.

It would be hard to get high efficiency from a relaxation oscillator

I am not sure what the actual breakdown voltage is but I think it is typically about 600 volts. I imagine it depends on tube length. Maybe sometime if I have a bad one I'll measure it. The older type that use an interrupting type starter or push button use the inductor (ballast) both to provide current limiting to pre-heat the filaments and an inductive high voltage kick on circuit opening to fire the lamp. Then they remain in series so the impedance limits the current through the plasma. The type that use a combination transformer for a ballast set the filaments of the tube at a high voltage above ground and the lamp must be in proximity to a grounded conductor across the length of the tube in order to achieve initial ionization. They will not start out in the open. It took me a while to figure that one out once!

The FS1 and FS2 starters were always a nuisance at end of bulb life as they would get cooked by the failing bulb having a higher and higher voltage drop as they failed. The self starting type made a big difference in the number of people needed to maintain factories/offices full of bulbs. Many just changed all their bulbs every 2 years or so, as that meant they could get it done cheaply by an outside contractor.

that's very strange. I have some free-standing 4' fluorescents lighting my barn that use an old-fashioned starter rescued from a traditional conductive fitting - they seem to work perfectly well without the fitting

With a starter they do not need to conductive fitting. Lights for cool situations, signs, barns etc, use what are called a "high output ballast" which light down to -40

Thanks. I misread the previous post.

When I was in high school and worked in the local TV repair shop I built a circuit to strobe a four foot fluorescent black light for our clubhouse. I use an audio output transformer from a TV and applied pulsed 12 volts DC to the secondary which generated high voltage pulses on the primary of the transformer to flash the tube. We had the coolest clubhouse in the neighborhood.

Of couse this was when TV's used tubes and so the audio output transformer had a turns ratio of 25:1 or greater.

I never did that, but I remember well well the small plate to voice coil transformers used. I used to make shock books, with a similar set of batteries and a title (with foil stripes) that encouraged esager openers.

Miss Playboy 1958 or some such. They picked it up and opened it and got a shock...

I don't think so. There are much more sophisticated ways to do this. I have designed many switching power supplies. The inductor is not just useful for limiting current. In a switching supply, if you don't need isolation, which the lamp does not, I can build you a supply using an inductor to make high voltage from low, low from high or even reverse the polarity. By pulse width modulating the amount of time you pump current into the inductor, I can give you a very nicely energy regulated string of high voltage output pulses from a power supply that has lots of 100/120 Hz ripple.

From my observation of the nature of the light output I see a very constant train of pulses at 60 kHz to 90 kHz, depending on the lamp. This is quite consistent with a regulated switcher and there is no reason to do it in a less sophisticated manner because it is so easy to do this way.

From my observation of the current waveform going into the lamp with a high speed Tektronix current probe I would infer that there is about 10 to 15 volts of ripple on the filter capacitor. So far I do not see anything inconsistent with my expectation of how this design would work if it were done by any engineer with even a little experience in designing switching supplies. If you do not agree I can dissect one and give you a detailed analysis of the circuit but I see nothing inconsistent in the behavior of the device with my expectation of how it should work.

Yes, indeed, a well designed switcher is optimal for efficiency and long life. Sadly, well designed ones may not be the cheapest ones, of which there seems to be many low cost, low efficiency, early failing variants out there. The large sellers here in Canada have gotten pretty sick of these cheap ones as they cost them many millions of dollars in client returns. They now only sell units from GE, Phillips etc

Some of them had 5 year warranties. Just return the unit to us with $5 for shipping and we will send you another one for free. Since a new one costs less than $5 this was a non warranty.

Its probably because of your perepheral vision, which if I remember correctly only sees in B&W, but can notice sudden movement better, even in lowlight conditions....it might have save our primitive lives to see that Dinosaur's foot coming!!! (not that man was around with Dinos, but mans forefathers maybe!!)

Have you ever looked at stars and found that you could see certain low light stars better when you did not look at them directly!!!!

Has anyone worked out how much energy is needed and pollution created by producing a CFL compared to producing the number of incandescent bulbs needed to equal the lifespan of the CFL ?

What are the comparative costs of recycling ? Mercury was mentioned earlier.

This equation should also include the energy needed for manufacturing, transport and packaging of components and subcomponents for the bulbs (e.g. raw plastics, injection molded parts, etc.) as well as the actual production/assembly, assuming that some components of a CFL are not manufactured where the actual lamp is assembled.

maybe (just guessing) the old incandescent lamp is not so bad after all, if one takes all those factors into account !

In the history of man, there have been milestones in time labeled per man's tech advancement.

Such as "Stone Age", "Iron Age", "Industrial Revolution" and many others.

I think our generations will be remembered (I don't know exactly how to say it) as the

"Period of Detremential Concentration of Natural Substances".

One comment said that "all of these materials already exist", for example mercury and phosphor. This is true since man can only gather these substances from the planet "Earth" which Nature has widely distributed over Man's World. But man gathers these substances and concentrates them into hazardous piles which become dangerous. Note that these substances were already in our World but they were spread thin enough so that they caused very little harm.

EXAMPLE: A documentary on "three mile island" stated that the people that were exposed to radiation, due to the accident, recieved no more total radiation than people that live on the western slope of the Rocky Mountains in Colorado recieve every six months. I have never seen any evidence of health problems for people that live in western Colorado.

The problem is that man harvests, concentrates and utilizes these substances but he leaves them in dangerous concentrations for future generations of man and animals. One hot item right now is "GREEN" and one of the large problems is man's concentration of Carbon and Oxygen into CO2 and then into the atmosphere.

BIG CONCENTRATIONS, BIG PROBLEMS

Everyone who lives "on the western slope of the Rocky Mountains in Colorado" will die just like the rest of us. Life on this planet is messy; there is no option to not generate pollution. Even if we go back to burning candles, oops that's petroleum make that whale blubber, oops that doesn't work any more either, not enough whales.

The point is there isn't any going back. Once something new is discovered or invented the genie is out of the bottle and the only realistic way to deal with it is to make more advances. More advances in technology, in pollution control, in resource management and so on.

Unfortunately, humans are crisis motivated. Wait until it's too late and then fix it. We are getting a little better at being less that way but just like anything else it's slow to change. I think the best approach is to encourage solving the problems created by new technologies than shunning the technology. We are much better at recognizing the dangers sooner as is evidenced by this blog.

How far back would you want to go really? Get rid of the cars the power plants the refineries the reactors? Sorry, it isn't going to happen.

The arrow of time moves forward and so must we.

One other problem with dictats or politically motivated moves to use more environmentally friendly products is to discard or be forced to discard the older less friendly products before the end of their useful/natural life. Everything manufactured creates its own unfriendly emissions during its manufacture, therefore temporarily increasing the emissions signature until the change over to the new technology is complete, delaying the reduction of the total emissions signature until further down the time scale.

Every new bulb, motor car, cement based building, comes to mind. I understand the manufacture of a new lower emissions car produces similar totals levels of unfriendly emissions as the fuel it burns during its natural working life. Increasing the natural life/durability of a motor car therefore decreases total unfriendly emissions. Legislating on introducing longer life bulbs is a drop in the ocean compared to legislating to increase the natural useful life of motor cars, say 15 to 20 years, albeit with new engine tachnology at set times. When will that come?

So politicians, be very careful what legislation you rush through in an attempt to appear green and win votes. Those who are in the 'know', know how silly it makes you look.

Georgee

you hit a nerve

things like why does nearly every car take different shocks, brake pads,air filters powerplant, transmission..........

If you look @ say industrial electrical control components, a round base 8pin relay, the pin assignments are standardized, back to the time of vacumn tubes.

if anything should be legistated it would be conformance to [auto]industry agreed upon standards of configuration but then you're messing w/the outside vendors [& unions] $'s

Cf lighting hasn't got there yet, shattersheild coatings, flicker rate, em radiation, all make the difference between energy savers & $ wasters

CF is just an intermediate step, leds will win, ever see CF marker lights on a car? leds are more & more common

The chances of politicans being thoughtful on any subject, when it runs counter to their re-election [fund raising] chances are slim

....I thought that Xenon lights were gas discharge like CFs, am I wrong?

Xenon lamps are gas discharge. Since they are filled with Xenon gas, they emit a different spectral line than lamps based on mercury. Xenon lamps do not use a phosphor coating on the inside. The mercury based lamps produce ultraviolet light that is not useful for lighting hence the phosphor coating to convert the energy from ultraviolet to visible. Additionally, fluorescent lamps are made with regular glass so as not to transmit the short-wave UV that would be harmful. Xenon lamps are often used in a pulsed mode but I think some of the new automotive headlights use Xenon lamps in either continuous arc or at a high pulse rate I am not sure.

SFIK, the light-emitting mechanism is different. In fluorescent lamps, the gas is only moderately excited, so they generate a characteristic molecular-transition line that is used to excite a phosphor. In an "arc" lamp, the contents are very hot, and the electrons and ions radiate (effectively) thermally - therefore across a wide spectrum. Of course there will also be some contribution in spectral lines that are characteristic of the gas.

Fyz

Yeah I guess in the back of my mind there was a question about it being "white" which of course is not a spectral line! But it makes perfect sense now and that also explains why the xenon lamps run very hot (depending on duty cycle). Thanks for straightening me out!

Rest assured that, no doubt, doubts are being raised...but the perpetually techno-ignorant and sophomoric conveyers of "spoon-fed news" information are not among those who understand the issue...or who realize how closely the blind ambition and open-eyed ignorance of legislators (especially in California) go together. The real problem when bad ideas begin to stampede...the resulting din drowns out all better ideas--especially old tech ideas which have long been around; and which would actually work. Another problem: it pays just as much (wastes government money just as fast, and for longer period of time) to explore a bad idea as a good one. And what else are governments for than to make tax appropriations evaporate as quickly and as maximally as possible!

The best doubters' solution at this point is to let it be known loudly, that citizens will do anything but comply...unless government bears all costs, intended or otherwise...until any mandatory CF experiment fails and is repealed.

No way to not comply, you say? Already, a certain doubt monger I know has begun horde-ing incandescents: for use and for supplying other doubters, and hapless non-doubters, as well. The key is to make the potential price of enforcement as costly to government (and complicit industry) as possible--if black market it must be, so be it. At minimum, this is the way to ensure that any phaseout will not (as it ought not) last less than one or two generations.

Surely you jest. Perhaps you should also begin stockpiling aluminum foil for protective hats to shield you from the government's mind control devices. You know you are witlessly playing into their hands to even be logged on the Internet. You are supporting their evil plan and creating pollution. But hey just power up that over unity generator and we're all saved. The incandescent bulb black market, that's great. If you hurry you can probably get a deal on some used equipment to make your own bulbs but you'll have to hide it from the "revenuers" or they'll come and git ya!

I have changed every light bulb in the house, which I was told would cut my energy consumption by those that are promoting energy conservation. I have been monitoring this change for well over 6 months now and have not noticed a significant change in energy cost. This may have been intentionally misleading propaganda by the people who make the bulbs.

One important thing i discovered in hindsight, however, is that you need to buy CFLs with the right color temperature. Daylight is 6500 k but most of the lower priced bulbs are 2700-3000. I bought some of those for my five month old daughter's room. Now it looks like a parking garage. The daylight bulbs are slightly more expensive than the other CFLs but you won't go blind. I found the higher kelvin bulbs at Wally World.

Great point man...I switched out a bunch of my bulbs as they burned out with the higher K bulbs from my local "bargain club"...great deal at $13 for a six pack of 100W equivalents.

Back to the original thread about savings from switching. My electrical usage is like clockwork, I've been in this house for five years, and the bill literally fluctuates by penny's through the seasons, and compared year to year the energy used in particular months is almost identical (says something about creatures of habit).

In he past three months, I've replaced all the bulbs in my house but two or three (exterior porch light, sidewalk lamp, etc) for a total of about 15 bulbs and the savings already of about $9 per month. The bulbs should last about three or four years at the rate I use them.

In other news, Fluorescent is the way to go on savings, and the environuts can breath easy knowing that the recycling of the hazardous goodies is not a difficult process, especially for the big hitters (like factories,etc.). As a matter of fact, our bulb suppliers at my former employer even offered a replace and recycle program at no charge as long as we were purchasing our bulbs from them (they brought new and took the old for recycling). We swapped out 360 mercury overhead units for 1.8 times the equivalent lighting in fluorescent and saved almost $2K monthly on the Electric bill from day one!

If you want to see just how much your'e saving put an ammeter in line with an incandescent bulb and a C.F. there is only a minimal amount of current reduction thereby the cost of the bulb/watthour does not justify making the change until the C.F.'s last 25 yrs or more. Hold it! that increase in bulb life would justify a price increase so we're stuck in the same boat anyway.

That is correct; but it doesn't mean that that there isn't a large reduction in power - the current and voltage are not in phase. It does highlight a potential issue in areas where the peak usage of power is for lighting - the power will go down, but the wiring requirement could go up.

The reason for the possible increase in current requirements is that many compact fluorescents have a substantial 60-Hz (50-Hz in Europe) component in the light output. This can reduce the perceived* lighting levels relative to a tungsten lamp that provides identical average output level and spectrum, so people compensate by using more lamps.

In spite of all this, there should be a net benefit from a general move to using compact fluorescents

*Unfortunately, official measurements of light level take no account of this sort of effect.

This brings up another problem. By changing to flourescent bulbs, and changing the "current" phase of the load, the Power meter is no longer matched to your house load and will give false readings.

If you do a major change such as changing all the bulbs, the next step is to get the power company to re-match the power meter. (you could be paying for more than you are using)

Don't they have power factor correction in the electronics?

Yes - even in Canada

You cannot measure the current with a typical meter and get an accurate reading. Viewing a typical current waveform for a 28 watt GE Helical lamp indicates a peak current of 1.86 amps. However it is only flowing for 0.8mS out of the 8.33mS AC half-wave. This is a horrendous crest factor and if you don't know what that means it is basically why you can't accurately read the correct average current on your meter. To make things even more confusing the lamp is printed with a current draw of 0.47 amps. I calculate from my scope measurement, which was approximate and average current of 0.178 amps, which works out to 22.5 watts at my 126 VRMS line voltage. This bulb is stated to be equal to a 100 watt bulb and by my measure I'm saving in the neighborhood of 75% no matter how you cut it.

That would be true if a bulb with a very high level of 120Hz flicker and a significant 60Hz component gave the same perceivedeffect as a bulb with that had the same average light output and no flicker. In practice, if you use only 60-Hz compact fluorescents, you probably need about twice the light output for the same minimum acceptable illumination; in addition, performance at nearly all tasks (including reading average-size newsprint) is significantly degraded. The situation is even worse with the 50-Hz supplies used in Europe. I have not looked into the current "standard" measurement methods, but I strongly suspect that the problem is that they don't take adequate account of this.

Fyz

So, I guess, the true power difference can be read using a Wattmeter, or does the phase changes mess this up?

GEM

It would depend a little on the type of wattmeter as to how accurate it is. The current pulse occurs very near the crest of the voltage waveform and the power factor is indicated to be 0.9. The watt-hour meter on your house will definitely read the energy usage reasonably accurately. It is pretty hard to fool one. If it were not so, people would be building converters to steal energy by fooling the meter. You can rest assured they do everything they can to prevent this. I believe the new electronic meters are probably more accurate than the old mechanical ones since they can resolve the current and voltage waveforms to very small increments of time.

Has anyone seen LED replacements for 400 W HID light fixtures? I'm moving towards fluorescents in our plant, but may hold off until LED's can provide a better payback.

Plastic-encased fluorescent bulbs are already on the market, called ShatterShield by one manufacturer. Bulb cost gets very high when you add the protective coating, our 54 W T5s (4 foot) with plastic coating can run $30+ each. You can also buy plastic tubes that slip over the bulbs, but bulb life in our plant has been poor with those slip-on tubes.

I was recently chatting with an individual in the lighting fixture business and learned that one problem pacing high power LEDs is getting the heat out. I made a couple of suggestions that I hope will be helpful. You might want to keep that problem in mind, though, when evaluating fixtures. Failure to adequately cool the junctions could have a pretty dramatic effect on LED life. DickL

OK,

LED's are available in any color of the spectrum. You don't have to get the blue/near UV ones if you don't want them.

The government and the environmentalists would have us believe that everything man does pollutes the earth, is no good for us, causes cancer, wastes resources, etc.

While we need to be responsible for what we do, most of the substances used in these products are here already. The production and transportation of incandescent lamps nulls out against the CCFLs since everything is made somewhere else these days.

If we stop and listened to everything that environmentalists and fundamentalists whined about, they'd have us stop breathing the air because that would cause cancer!

I just got a note from our power company. The message was simple. CCFLs do save money. Unplug your TV sets when you're not using them, your cell phone charges and anything else that draws currents in standby when you're not using it (within reason).

If you really want to save energy, be proactive. Saving energy in and of itself will reduce pollution overall. You can't get rid of all of it. If you have a recycling facility for fluorescent lamps, use it. If you don't, they still save energy and the pollution created by generating the energy.

This world isn't perfect. Live within what it has to offer and protect what is within your means to protect.

LEDs available in any colour temperature!!

Are you sure ???

It took 'em long enough to get a blue LED...then white..

Surely LEDs are a quantum device so the colours will be in finite bands?

Or am I just soooo far behind?

Never did get the hang of this electrickery stuff.