|
|
|
| | Energy Star challenged over 150 Hz requirement for LED lamps | | 12 Jan 2010 | | The Energy Star criteria for Integrated LED Replacement Lamps includes a requirement for 150 Hz operation, which is now being challenged. | | The US Department of Energy (DOE) is facing a challenge to its Energy Star criteria for Integrated LED Replacement Lamps, which were published in December 2009 (see News) and are due to go into effect on August 31, 2010.
DOE is already looking at the issue and consulting with technical experts, as described below (see "DOE response to queries").
The issue concerns an item on page 7 of the criteria, and applies to all replacement lamps. It relates to LED operating frequency, which must be ≥150 Hz. Of course, mains voltage operates at a frequency of 50 or 60 Hz, depending on the country.
The document says "This performance characteristic addresses problems with visible flicker due to low frequency operation and applies to steady-state as well as dimmed operation. Dimming operation shall meet the requirement at all light output levels."
This requirement was, apparently, introduced at short notice just before the criteria were finalized. The previous draft version contained a requirement for ≥120 Hz. The 150 Hz requirement is being challenged by, among others, Lynk Labs Inc. (see press release).
Mike Miskin, president of Lynk Labs, told LEDs Magazine that "increasing the frequency requires additional drive components to be added. This increases cost but also requires additional real estate on the product. Many lamps with small form-factors may not have the real estate to allow for the additional components." Since these lamps could not be made, they would effectively be excluded from participating in the Integral LED Lamp market, which "will contradict the very mission of DOE Energy Star," argues Miskin.
Another company, Once Innovations, has filed formal requests with the DOE under the Freedom of Information Act (FOIA). The company has requested all information and documents used as the basis for changing the Energy Star requirement on operating frequency for integral LED lamps from ≥120 Hz to ≥150 Hz.
Once's press release says that the change "was apparently based on a vague suggestion in a comment from NGLIA/NEMA regarding the possibility of visual flicker problems. That comment simply suggested that the issue be looked into further."
In an earlier press release, Once said that hundreds of pages of research material showed "no evidence, other than speculative statistics, to support claims that luminous modulation over 100Hz is visible or harmful to human health."
DOE response to queries
The DOE Energy Star team is responding to queries about this issue as follows: "As stated in the criteria document, the intent of this requirement is to prevent visible flicker in qualified integral LED lamps, in support of the successful market adoption of high-efficiency light sources.
"However, since publication of the final criteria document, DOE has been contacted by a number of industry stakeholders who have raised technical and market issues related to this particular requirement.
"We are currently investigating these issues and engaging technical experts and the relevant industry standards organizations to resolve them. We will be mailing additional information on this issue to ENERGY STAR partners as soon as we are able. We anticipate resolution to take approximately 30-60 days."
| | COMMENTS | | Name: bob mueller Posted: Wed, 13 Jan 2010 20:01 |
| Don't various discharge lamps already suffer from this same "problem"? I have heard few complaints. And we look at TV images and computer screens with flicker rates under 120 Hz (with the exception of a few modern models) without many consequences. The presence of the flicker can be verified easily enough. Sometimes I see it when rapidly turning my head to change what I am observing, but under steady conditions I am fully unaware of the flicker and doubt that I am alone. | | Name: shoi_ Posted: Wed, 13 Jan 2010 21:01 |
| European fluorescent lamps flicker at 100Hz, US at 120Hz. The flicker at 100Hz is apparent to just about anyone and has been shown by Wilkins and others to be "unpleasant" (days off sick, headaches, work errors etc). http://lrt.sagepub.com/cgi/content/abstract/21/1/11 or try http://scholar.google.co.uk/scholar?hl=en&q=fluorescent flicker
Flicker from fluo lamps is I think much less obvious than flicker from LEDs, because of the decay time of the phosphors. In fact 120Hz flicker from fluo lamps is less obvious but still apparent because the 1.7millisecond difference (10ms vs 8.3ms) allows the phosphor to decay to nearly zero.
I don't have any data but my recollection is that the decay time on standard Nichia white leds was much quicker, and I know that in our designs some years ago we certainly upped the frequency to eliminate visual flicker.
I would suggest any sensible spec would actually call for a modulation in light output of less than N percent. Computer monitors (longish persistence phosphors) are acceptable at 80Hz, fluo lamps are not good enough at 120Hz.
As a jolly aside the Wilkins study, which showed how to make an unpleasant work environment, was conducted in a tax office.
The CRI of leds is bad enough already, lets not legislate for more sick buildings. | | Name: bonkers Posted: Wed, 13 Jan 2010 22:01 |
| I can understand suppliers getting upset at additional costs, but they are not considering all the factors.
Firstly, the rules are the same for all, so whatever the requirement is, nobody gets an unfair advantage.
Secondly, although "not harmful to human health" the 120Hz flicker of old-style fluorescent tubes was roundly hated by those subjected to it - many say it triggered migraines - who cares, building this annoyance into a new generation of lights will reduce uptake. We've all seen the stream of dots when you flick your gaze over the indication neon in say a multi-plug adapter. look at the money we all spent to move to 85Hz monitors - note also that monitors and fluorescents have a thermal inertia in the plasma that reduces the flicker amplitude - the minimum illumination is maybe 60% of the maximum, in one cycle. With LED's this effect is not present - and a simple bridge rectifier would result in a close to 100% on/off ratio.
A smoothing capacitor will reduce this, at additional cost, and even the primitive circuits proposed by the complainant manufacturer can be smoothed to an arbitrary degree by adding bulk smoothing capacitance. I don't know what the permissible flicker depth is - maybe 5% - at this level its cheaper to put a high frequency ballast in.
So - lets get it right as far as the human factors are concerned, the electronics cost is trivial compared to the lifetime cost save - conventional high efficiency bulbs carry high frequency ballasts whilst still costing under $2 - they flicker by only a few percent at rates over 20kHz - and are totally acceptable. | | Name: apollo Posted: Thu, 14 Jan 2010 05:01 |
| Does DOE try to exclude the AC LED from the market? Flickering relates to the light waveform and the frequency, a sine wave light of 100Hz is more comfortable than a pulsed light at 100Hz, and flickering may not be observed. | | Name: zdenko Posted: Fri, 15 Jan 2010 00:01 |
| Problem is, majority of people does not understand definition of frequency and harmonics. Magnetic ballast operated fluorescent tubes cannot be classified as a 120Hz light sources. If you measure light modulation frequency, 120Hz is fundamental frequency however in virtually every magnetic FL you will find 60Hz (and many other) fundamental subharmonic and that’s what people are perceiving as flicker. Also, magnetic FL produces subharmonic hum which is, in my opinion and opinion of several studies, more significant element of discomfort. On other side, majority of pure 120Hz LED systems do not have frequency subharmonic because they are pure linear devices.
| | Name: zdenko Posted: Fri, 15 Jan 2010 00:01 |
| Bonker,
If you are talking negative health effects, you should mention dark side of the high frequency ballasts. Did you ever measure EMI of the CFL lamps with 40KHz switching converter incorporated into parabolic reflector (majority down light fixtures)? Some of the devices I had a chance to test may easily pass as an Electronic warfare. And they do have nice large ENERGY STAR label on it.
| | Name: mr hz Posted: Fri, 15 Jan 2010 08:01 |
| If DOE really want to prevent sensitive people to "see and feel" flickering then they need to go above 400Hz. My guess is that this is something strictly political. | | Name: shoi_ Posted: Fri, 15 Jan 2010 11:01 |
| Thinking about it some more, if they allow 120Hz in the US then we will very likely get the self same lamps in Europe (ie without frequency shifting hardware) only they will naturally operate at 100Hz which will be significantly more "flickery".
You know how the "unpleasantness" and poor performance of CFLs has brought about a near revolt on both sides of the Atlantic, it would be foolish to run into the same issues with LEDs. Yes it's another cost barrier to overcome, but still all manufacturers are faced with the same problem, and probably the solution will be implemented in the form of a chip anyway. It's better to be in a market selling products customers want to buy. | | Name: led man Posted: Fri, 15 Jan 2010 19:01 |
| A 1961 study dome by DH Kelly shows human perception of flicker quits at about 90Hz, even at 100% modulation. This is well below the 100Hz or 120Hz of concern in this matter. I believe the DOE was incorrect in this matter. | | Name: augusto Posted: Sat, 16 Jan 2010 01:01 |
| Firstly, I couldn't find any medical evidences that operating frequncy of ≥100Hz make peoples harmed in the world in the situation of living under illumination not staring at monitor and TV. All test data shows that ≥80Hz is enough for normal people excluding Superman and his friends. Guess is only guess! DOE, please do not make a point foggy continuosly. If they have no evidence, put away the curent criteria of LED Operating Frequency among criterias.
Now, for our earth and USA,we must think about the efficiency and cost in using LED integral lamp to replace the current bulbs. If we forgive up AC LED in this time, we might make the earth ruined more than using the old lighting solution by using additional components like Converter;short life circle more than LED chip for DC LED and make USA consumers spend lots of money that we don't need to spend additionaly for buying new LED LAMP. | | Name: shoi_ Posted: Wed, 20 Jan 2010 13:01 |
| Led man -
I cannot trace this study by Kelly, but there is no doubt that people can easily detect the flicker in a fluo lamp. You can test your friends and neighbours tonight and I'm sure you'll get a result. Even so such effects can be insidious, ie you don't have to be able to detect something consciously to be affected by it.
Augusto-
I quoted one study exactly to the point by Wilkins - http://lrt.sagepub.com/cgi/content/abstract/21/1/11
or you can do a general search and come up with plenty others around the same point
http://scholar.google.co.uk/scholar?hl=en&q=fluorescent+flicker
Further I would contest that the takeup of CFLs is low because people dont like them. I contend takeup might well be increased if we had lamps which people actually thought worked well.
It's not true to portray LEDs as plant-saving technonlogy; now and for the next few years LEDs wont offer any efficiency (ie CO2) savings over current alternatives, nor will they have purchase cost savings. Much better to make them work properly than to make them cheap and nasty.
| | Name: hopshop Posted: Thu, 04 Feb 2010 15:02 |
| shoi_ : I believe this is the article LED MAN was refering to in 1971.
http://www.pnas.org/content/68/11/2802.full.pdf | | Name: shoi_ Posted: Sat, 06 Feb 2010 00:02 |
| I cannot find the name Kelly nor the figure 90Hz in it so I don't think that's the one. Anyway it's just untrue to say that you cannot spot a magnetically ballasted fluo lamp which flickers at 100or120Hz, from an electronically ballasted lamp which essentially doesn't flicker at all. I can, you can, anybody can.
About the proposed 150Hz lower limit. Can anyone here think of why or how one could make a lamp which worked at just above line frequency, surely it's cheaper and easier to make something working in the >35kHz region, like fluo ballasts?
I now know that the IEEE working group are likely to have something out for public consumption maybe this month, I'll post a link as and when. The group is at http://grouper.ieee.org/groups/1789/. | | Name: rickr Posted: Mon, 08 Feb 2010 17:02 |
You have to remember that this is for Energy Star listing. That program is aimed at being a quality reasurance for residential use. Also a major aim of the LED requirements is to avoid the consumer backlash CFLs suffered.
The point of 150hz is to require frequency control. So "if 150hz then why not 240hz?" When the science is reviewed/completed many are predicting a higher than 150hz requirement.
I would love to have line voltage LEDs that are very cheap and small. Some applications don't hinge on flicker. Residential incandescent replacement requires incandescent quality. | | Name: shoi_ Posted: Fri, 19 Feb 2010 22:02 |
| The IEEE first report on flicker is now available for all to see, and all are invited to comment.
IEEE PAR1789 "Recommending practices for modulating current in High Brightness LEDs for mitigating health risks to viewers"
http://grouper.ieee.org/groups/1789/comments.html
Your chance to contribute to getting this right. |
| | | |
