|Elliott Sound Products||Solid State Lighting & Temperature|
Solid-state lighting (SSL) generally refers to LED (light emitting diode) light bulbs or fittings, but compact fluorescent lamps (CFLs) also contain electronics. While the light emitter may not be 'solid state' the drive circuitry most certainly is, and it's subject to a limited temperature range. Semiconductors (ICs, transistors and LEDs) all have an absolute upper temperature limit of 125-150°C, but for normal operation is it expected that the temperature should always be well below the upper limit. As the temperature increases, the device life is reduced, typically halving the normal life expectancy for each 10°C temperature increase. There are also minimum operating temperatures for most semiconductors and some other parts.
At the time of writing, misguided governments all over the world are banning incandescent lamps, either by direct legislation or stealth. I have no problem at all with the idea of more efficient lighting systems, and my own house and workshop uses at least 85% LED lighting, with a few CFLs, two T5 electronically ballasted fluorescent lamps, and one solitary incandescent lamp. I haven't counted the lights inside the microwave or gas ovens - the choices for those are non-existent. Incandescent lamps are the only things that will work in high temperature environments.
What is not considered by politicians, bureaucrats and their 'advisors' is that all forms of electronic lighting can only operate over a limited temperature range. If it's too hot or too cold, then electronic lighting systems either work very poorly or in some cases will fail prematurely. Especially at elevated temperatures, it's not only the semiconductors that will suffer and fail, but capacitors that are used as essential parts of the circuit are also vulnerable.
Most people involved in electronics are well aware of the upper temperature limit of most components - not just semiconductors. Electrolytic capacitors are used extensively in all forms of electronic lighting, because they provide high capacity at low cost. Electrolytic caps are commonly available with maximum temperature ratings of 85°C and 105°C, and less commonly at 125°C. The typical rated life at maximum temperature may only be perhaps 2,000 hours - well below the claims made for the lighting product itself (typically between 25,000 hours and 50,000 hours).
The only way to get electrolytic capacitors to operate for 50,000 hours is to ensure that ...
For example, a 450V, 105°C capacitor operated at 320V DC and with its temperature maintained below 65°C has a fighting chance of reaching 50,000 hours. Increase either the voltage or temperature, and the expected life will be reduced. It might sound easy enough in theory, but in practice it can be almost impossible to keep the temperature low enough, because the average householder is unaware that no electronic lighting system can be enclosed without airflow. Indeed, many lighting professionals will be unaware too, because it's not what they are used to. Lights are no longer just simple globes with a filament inside - they also contain a significant amount of electronic circuitry.
This is one of the main reasons that CFL and LED lighting products were de-rated from the original (often outlandish) claims made. I recall claims that CFLs would last for 20,000 or even 30,000 hours when they first became popular, but few claim more than around 16,000 hours now, and even that is often highly optimistic. Likewise, LED lights were sometimes claimed to last for 'up to' 100,000 hours. Really? Most now claim more realistic figures of around 30,000 hours - which is still a very long time.
High temperature operation can be tested easily, and I have performed just such tests many times. Even a 10W CFL in a fairly large (3 litre) sealed enclosed fitting will produce a temperature rise of about 35°C, and larger lamps or smaller fittings will make that a great deal worse. With the lamp's electronics at around 60°C we can expect a long life, but at high ambient temperatures life expectancy will be reduced.
With many different lighting products analysed for failures, I have found that electrolytic capacitors are usually the first component to die. In particular, low value high voltage types (e.g. 1µF, 400V) are the most unreliable, having racked up hundreds of failures when the lights are used 24/7. Many will fail well before the warranty expires. LEDs are also vulnerable, especially when the manufacturer has little experience with thermal management and the LEDs run too hot because of heatsinks that are too small or inadequate thermal conductivity between the LEDs and the heatsink.
This type of problem will continue until retro-fit 'bulbs' are finally a thing of the past, and luminaires are purpose-designed with integral LEDs and dedicated power supplies. While there are many such products available, they are not mainstream. It doesn't help that many 'home improvement' stores continue to sell light fittings that are not suited for any lamp that uses electronics (LEDs or CFLs).
CFLs are renowned for having very poor light output at low temperatures, with some flatly refusing to even light at all if the temperature is below -20°C or so. Even if they do manage to turn on, the light output will be very low until the tube warms up enough to allow free mercury vapour to create a respectable amount of light. Some are even pretty poor started from normal room temperature, and take a few minutes to produce their claimed light output.
In early 2014, many places in northern America (including Canada and especially the US mid-west) have seen exceptionally low temperatures, with claims of -50°C filtering through the news media. More commonly, there are reports of night time temperatures of -25°C over wide regions. In places such as Siberia, -50°C is common in some parts. Antarctica is generally considered the coldest place on Earth, with temperatures as low as -90°C claimed. Extremely low temperatures guarantee that CFLs will be all but useless, but (and perhaps surprisingly) even LED lighting can be affected.
Electrolytic capacitors are used in all CFLs and almost all LED fixtures and bulbs, and the electrolyte is perfectly capable of freezing. When that happens, the capacitor loses much of its rated capacitance, increases its internal resistance, and can easily cause the power supply to malfunction. Electrolytic capacitors are commonly rated for minimum temperatures of -25°C, -40°C or -55°C, but that does not necessarily mean that they will still work properly at those temperatures, only that they will not be irreparably damaged. Looking through some data sheets indicated that one manufacturer rates their caps to -40°C, but only provides impedance figures down to -10°C. This implies that there is no guarantee that the caps will function normally at the lowest rated temperature.
Exceptionally low temperatures can cause other issues as well. Most electronic equipment has a maximum allowable humidity, always stated as 'non-condensing'. Unless the electronics are hermetically sealed, it's almost inevitable that some outside air will enter the enclosure. As it's heated it absorbs moisture. When the lamp (or other gear) is turned off, if enough moisture has built up in the warm air inside it will condense on the electronics. If this happens, failure is not far away. The US military seems to be aware of this issue , but the reference primarily concentrates on high temperatures. Elsewhere, condensation gets hardly a mention.
Semiconductor devices usually also have a minimum operating temperature, and for many 'consumer grade' parts that might only be down to 0°C. While the storage temperature range might be from -65 to +150°C, the operating temperature of most consumer grade parts is from 0 to 70°C. Characteristics are not guaranteed at any temperature outside that range. In particular, at low temperatures, the gain of most semiconductor devices falls, and the turn-on voltage for silicon junctions increases (~2mV/°C). For a transistor that is expected to turn on at 0.65V (at 25°C), the circuit might not function if that's increased to (say) 0.73V (at -40°C), especially since it has lower gain than normal. LEDs are also semiconductors, and if cold enough, the power supply (assuming it works) may be unable to supply a high enough voltage to make the LED junctions conduct. I don't know if this is a problem, and I doubt that it has been looked by too many manufacturers.
Unfortunately, it is extremely difficult for most of us to even test whether the circuit will work at -50°C or not. I certainly can't, and I suspect that few manufacturers would be able to do so either. That's well outside the range we expect from normal refrigeration systems, and specialised systems are needed if you want to be able to test at temperatures below around -18°C (typical of domestic freezers). Even many industrial low temperature freezers can only get to -45°C or so, limited by the characteristics of common refrigerants .
Despite some fairly serious searching, there is not very much information available for electronics equipment operating at sub-zero temperatures. Most studies look at the highest likely operating temperature, but seem to gloss over the fact that there will be equipment that's expected to work at -20°C or less. If you look at the specifications for commercial-off-the-shelf (COTS) power supplies from major manufacturers, not many will claim that they are suitable for sub-zero operation. There are exceptions of course - many of the supplies from Meanwell are rated for -30 to +70°C operation.
Needless to say, an incandescent lamp doesn't care what the temperature might be - they function fine from sub-arctic to oven temperatures quite happily. There are no electronics involved, and the filament is in a (partial) vacuum, completely sealed from the outside world. Since the filament typically operates at around 3,000°C, a few degrees here or there is of no consequence. There will be a bit of extra thermal shock when the light is turned on from (very) cold that will reduce the life a little, but most of the time an incandescent lamp will simply provide light - and heat of course. It will not misbehave in mysterious ways, regardless of the temperature.
No special precautions are needed to account for extremely low temperatures, and thousands (millions?) are in daily use in freezers. Outdoors, they will function happily whether the temperature is -50°C or +50°C. They are certainly far from perfect, mainly due to poor overall efficiency, but they will work where electronic 'equivalents' won't - or at least the alternatives that are considered equivalent by politicians and bureaucrats.
What they completely fail to understand is that each type of light has its uses, and no one light source is suitable for all applications. Blanket bans based on one criterion only are ill-conceived, and mainly serve to annoy a proportion of the populace. Sometimes, there is simply no other sensible choice, and it's very irritating if that is taken away by people with little or no knowledge of electronics, and who lack any real understanding of lighting in general.
The life of incandescent lamps of all kinds can be extended dramatically by using a modern trailing edge or universal dimmer, because they have a soft-start function where power is applied relatively slowly. This minimises thermal shock and stress on the filament, so it is less likely to fail at switch-on. Despite all the claims you will see about dimmable CFL and LED lights, they are often quite unstable when used with standard household dimmers. For those who want to know more about dimming, see Light Dimmers and Dimmers & LEDs. The latter also applies to CFLs, which have similar problems.
Traditional 'wall-plate' dimmers were designed for incandescent lighting, and can only ever work properly with incandescent lights. That they work at all with electronic lighting products is a minor miracle, and as noted above in some cases the only way 'dimmable' electronic lighting will dim in a reasonably sensible manner is to use an incandescent lamp in parallel. Halogen lamps are still incandescent, but need less power for much the same light output. For example, a '60W equivalent' halogen lamp uses around 43W, and looks virtually identical to the traditional 60W lamp. Naturally, they are more expensive than the bulbs they replace.
Yes, I'm annoyed, and have been for some time now. I don't use incandescent lamps any more, apart from the one ('high efficiency' halogen type) I use to make a dimmer behave itself on an allegedly 'dimmable' CFL. Fortunately for most of us in Australia, we live in a temperate climate, and it's uncommon for the temperature to drop much below zero, even in the middle of winter. The CFLs I have work just fine for the places where they are used - in well ventilated fittings of course. We do get some very hot days during summer (over 40°C a few days each year on average), but outdoor lighting is never needed in the middle of the day so high temperatures aren't a major problem.
Not so fortunate are those who live in places that get extremely cold in winter. Some, probably most, LED lights will work fine, but CFLs will generally be close to useless for at least the first 5 minutes. Some may refuse to start at all if the temperature is low enough. It seems that almost no-one has properly recognised this as a real issue, which does come as a surprise. Government decisions need to be made based on facts and reality, rather than knee-jerk reactions to a perceived 'problem'. Mostly, it's not a problem at all. It would also help if governments would admit that they made a mistake, but don't hold your breath waiting for that to happen .
Even where they are still available, no-one should really be using standard incandescent lamps for most applications. Although they are cheap to buy, they are comparatively very expensive to run. Indoor locations where light is needed over extended periods will benefit from the use of CFLs, or better still, LEDs. Replacing fluorescent tubes with LED tubes is now affordable, and you'll get more light with lower power consumption. Where lights are not switched on and off a lot, CFLs are relatively cheap and work fairly well - many early versions were not considered acceptable by many users. Sometimes, there is simply no useable alternative to a tungsten filament lamp. The new ones, which are still legal in most countries, might be a little different from what you were used to. Most have a small halogen bulb inside a 'traditional' bulb, but they will work anywhere, just like they always did.
As electricity prices continue to rise, people will seek more efficient forms of lighting by themselves in order to save money. If the need arises, they also need to be able to use a product that works, even if it is inefficient. We have every right to be very annoyed when ill-advised governments make unilateral decisions that make life just that little bit harder for us. It's even more galling when we are told that "it's for our own good". I have enough incandescent lamps saved up to last me until well past my expiry date, and I expect that many other people also have their cache - just in case. There are reports all over the Interweb of people stocking up on incandescent lamps before they all vanish from the store's shelves due to the latest government 'initiative' in Outer Mongolia or wherever.
It's interesting to note that the Australian 'Minimum Energy Performance Standards' (MEPS) for incandescent lamps seems to have stalled. Mains voltage halogen lamps were supposedly being phased out as of 2011, yet they are still readily available. Although there is no official recognition of the fact, it looks like the regulators have accepted that in some cases there are exactly zero alternatives to tungsten filament lamps. It's to be hoped that regulators elsewhere finally wake up to reality and continue to allow the sale of incandescent lamps for situations where nothing else will work.
|Copyright Notice. This material, including but not limited to all text and diagrams, is the intellectual property of Rod Elliott, and is Copyright © 2014. Reproduction or re-publication by any means whatsoever, whether electronic, mechanical or electro-mechanical, is strictly prohibited under International Copyright laws. The author (Rod Elliott) grants the reader the right to use this information for personal use only, and further allows that one (1) copy may be made for reference. Commercial use in whole or in part is prohibited without express written authorisation from Rod Elliott.|