From Grumpy Granddad on unthreaded, probably worthy of a discussion thread.

Hi,

For some years now I have been using low energy fluorescent light bulbs as has every else. I have found that the majority of low energy bulbs fitted in our house are used in a short duty cycle, on for a few minutes then off, cycle repeated many times / twenty four hours. The low energy bulbs over our bathroom mirror are an example, only on for a few minutes whilst taking a shower or using the toilet, then off untill the next call of nature. This causes very early failure, (within a few months). I also have a 5W bulb in a security fitting outside the back door which is left on all night, this has not failed in 5 years!

I think the Chinese bulb makers do not take enough care with their design to protect the electronics from inrush current / voltage.

So just over a year ago, I decided to take the plunge into LED bulbs. We were redecorating our bedroom and decided to fit two bed head led lights. They are each 3W fitted in a stainless bullet shaped head, with a flexible stalk to the led head from a wall plate containing the constant current sealed module. These lights nearly broke the bank at £75 each! They give an excellenent narrow beam of light so I and my wife can read easily and without disturbing the other if they wish to sleep.

However just out of warranty, one lamp starting buzzing in the power module and has now become intermittent, light flickering at low rate. I have just ordered a new module @ £14 to repair it.

This is in a device that is claimed to have a 100,000 hour mtbf! It cannot have run for more than 500 hours!

Once again the encapsulated power module is made in China, CE mark == China engineered! If I could get it apart I am certain I would find that the design has been optimised for cheapness and not reliability.

When our beloved masters forced the use of CCFL and LED lamps on us, I wonder did they ever do properly conducted tests on a large sample of manufacturers product, with realistic duty cycles, to obtain a true MTBF.

Is there a published specification for these lights that they are tested against before sale, or is it another free for all jamboree? (I recently bought some Ever Ready branded low energy buls, I was very disappointed to find they were also made in China! )

I am completely in the dark about this

cheers

GG

Jul 30, 2014 at 9:53 AM | Unregistered CommenterGrumpy Granddad

Problems include reliability, high cost, light quality, actual power savings v claimed, after life disposal are amongst problems with low energy bulbs. From my own experience the cheap Chinese LED and CCFL bulbs are a bit variable in quality, higher quality are expensive and in low use applications I guess the pay back time would be measured in decades.

LAst summer I have replaced our halogen GU10 spotlights (40-50W for the most part) with LED. The MTBF for the halogens wasn't that long and they are expensive. I used mid-price range LEDs and so far had no failures. We have 8 in our living kitchen area and they are on for about 12 hours per day in winter I have used 5W LED replacing the halogens and the light level is down a bit and a different colour but perfectly acceptable. I also have a couple on some stairs running about 14 hours day everyday I initially put in the low cost variety but had a failure after about 1000 hours so have upgraded. Having the lights on all day is as much psychological as anything, having a dark corner in the hallway depresses me.

I think that the weak points in these bulbs for reliability are electrolytic capacitors and transformers. Both have in built problems which can lead to Early Life Failure (ELF). The classic bathtub reliability curve for components has, in my experience a shorter background failure rate and that is at a higher level than for most other components in all applications. I suspect in very low usage applications the failure rate will be quite high.

Other things which you might not consider in terms of saving the earth are pollution in the manufacturing and after life disposal. The one I hadn't really thought about until recently was Power Factor (PF), which means that if the PF isn't 1.0 there is more load than it says on the bulb. It is explained quite well in this document. CFL Power Factor

I'm hoping that there'll be other contributions for an interesting and informative discussion.

Hi

Thank you for creating the discussion thread.

Some more thoughts.

MTBF of electronic components

I have been involved in electronics design all my life, I spent many years in the labs at BAC, working on missile electronics then Concorde stuff, I then moved to DEC (Digital Equip Co), where for some years I worked in CSSE, a group that amoungst many things, tried to ensure that a product was serviceable, and had an ecconomic MTBF. After all if we we selling a production run of 100,000 widgets, we did not want to have to repair or replace more than the number that gave us a good profit bottom line. Too high (good) a MTBF cost too much in design and produce, too low an MTBF ate into the warranty budget and field service contract revenues. I have spent many many hours in the test labs determining the base MTBF of the representative samples, which was then extrapolated statistically to the production numbers.

As you say electronic componennts and also mechanical components, (motors switches etc) have the well known "bathtub curve" with a high initial failure rate and then a long low failure rate during operational life and then a rising rate at end of life, (wear out).

I did some sub contract work whilst at BAC for NASA, on the old Apollo space program. The yanks had a special category of reliability for the Apollo program "manned space flight". In this EVERY individual component had a history card and was individually tested, burnt in, serialised and certified, an attempt at a zero defect program. I remember approaching a UK resistror manufacturer, who when he understood the process wanted £12 for a 1/8 w thin film resistor, usually purchased for 1p. (Before the days of smd)

Back to today's consumer electronics, including lights. The goal of all the manufacturers is to make the product as cheaply as possible , although keeping the failure rate below the point at which the consumers are upset at the failure rate, i.e. "Its just one of those facts of life", against being so pi**ed off they refuse to buy any more devices from that manufacturer. Unfortunately a fly by night manufacturer doesn't give a toss. He will make a number of poor quality devices, sell them off, make a pile, then move on and reopen under a new name.

I have a friend who ran an UL certified lab, he spent most of his time qualifying electronic prodcts, for proper CE certification, it took a lot of time and a lot of ££££'s. Do any of these Chinese lights have proper certification or are they all self certified?

I have looked at some of the led systems on sale at the moment, they seem to be mainly two classes:-

1. Where a string of led units are powered from a separate constant current power unit, and 2. the type I purchased, when each LED lamp is fed from a small integral power module. The type with the integral module seems to run very hot, there is no ventilation, the white encapsulation on the power modules of my LED's is toasted to a nice shade of brown. I have no experience of the common power unit type, but if it is like the 12v power units for the 12v halogen bulb systems, they might be better constructed and ventilated., giving you a better MTBF.

CCFl low energy lamps almost always have the electronic CC supply built into the bulb body. There are a few industrial units that have replaceable tube units.

One of the characteristics of both CCFL and led lamps is that they need a current limiting impedance to limit the operating current to non distructive values. The old domestic batten fluorescent lamps had a big choke mounted in the fitting. Nowadays this is an electronic constant current module. Unfortunately these devices are transformerless i.e. no 50Hz iron transformers, but they may have HF ferrite transformers, using electronic regulation. Some of the Chinese circuits I have seen have used very inadequate capacitors in the circuit. The problem is the ERF (equivalent series restance), which ALL capacitors exhibit when presented with an ac ripple voltage. The ERF causes the capacitor to heat and in the worst case of old wet electrolytics boil its electrolyte and blow the cap apart. ERF is still a problem even with solid tant SMD devices.

Capacitors that can withstand ac ripple are expensive and are NOT normally used by Far Eastern designers. One of the most obvious examples of this was the multiple failures of capacitors in the power supply of the early Sheevaplug ARM computer. Pictures and description can be seen in the old Sheevaplug forum, if it is still running.

I think it is another symptom of the EC and the Briitish Government legislating that we have to use these devices without the slightest technical knowlege of the pitfalls. Just like their dash for renewable power sources.

cheers

GG

Grumpy Grandad
We have a similar background, I've worked for most companies in the now nearly defunct British Electronics industry. My first job was in Electrolytic Capacitor manufacture, which was an in at the deep end experience in many ways. The bulk of the workforce were coal miners wifes and daughters, quite an experience for a country boy (Teuchter). I've worked in both military and civilian industry. The ESR issue was an everyday problem, in fact the whole manufacturing process was a headache. I worked there barely 2 years before the place went into terminal decline. 99.999% pure Aluminium foil, Etching, forming, measuring, winding, impregnating,ageing, canning, ageing, sealing then ageing. With sample testing at each stage and often 100% test at the end. For HV variants the failure rate was impressively high. If there was an alternative to electrolytics then they would have been replaced long ago, Only the Anode of an electrolytic has a formed layer, which is created at a voltage appropriate to the application, of oxide to keep the amount of foil used to a minimum, so any issue with the application can be catastrophic as hydrogen is a by product of the forming process.

As you say consumer products are always a compromise between reliability and cost. I think that sometimes the bean counters concentrate too much on the cost side of things. Reliability used to be king, VW and Honda in motor cars for instance. But much of today's gadgets have a life expectancy of about 18 months before they are swapped for the latest model.

I would suspect, without any proof, that the very cheap eBay type bulbs haven't even been self certified, just had the appropriate marking printed on them. I also think that the plant where they were made is probably an H&S and environmental nightmare.

I prefer the LED from LEDs to CFLs, but incandescent is best. I'm not really up on the latest PS technologies as I spent the last 20 years working in IT. I've only used the Type 2 in your categories, straight replacement for incandescents. The ones I have purchased run cooler than those they replaced so over heating isn't a particular problem. I have noticed a couple of things, the ones in the kitchen seem to have quite a long after glow, which I put down to capacitor discharge but haven't confirmed. Second I have had to replace one of the wall switches for these lights in the kitchen, it was only a couple of years old. Again I'm not sure if the bulbs caused the issue or if it was an ELF switch.

Has research been done into all the issues associated with these new technologies? A couple of years ago some research suggested that LEDs could damage the retina, and also I think caused bleaching when used to illuminate paintings and photos. I don't think there is any definitive evidence either way, but after the MMR scare I would hope that some research was being carried out by the likes of Philips, but somehow doubt it.

Because of the "hassle factor" in recycling failed low energy bulbs I'm sure a fair number end up in landfill or whatever is going to replace it.

Firstly, I would say go for the big brand versions of these bulbs. They WILL be more expensive but will have been tested for a proper performance.

Secondly it was pointed out years ago that the low energy bulbs work best when they are NOT switched on and off frequently. Not only do they take time to give full light (obviously dangerous on stairs) but their efficiency relies on a constant light that is not interrupted by being switched on and off . Frequent switching on and off will cause premature failure

We still retain the old fashioned light bulbs in places where short term illumination is required.

In our kitchen we have a set of Led's which are perfectly acceptable.

My next task is to bite the bullet and replace the 8 pieces of 35 or 50 watt halogen bulbs in our living room with Led's. These need to be dimmable which puts the price up and there will be a notable loss in light quality. In winter that is important.

So, horses for courses as regards light bulbs. Some bulbs will be better than others in specific situations and the light quality aspect can not be ignored.

tonyb

@Jul 31, 2014 at 10:53 AM | Registered Commenter Martin A
Perhaps he meant 'wet electrolytics', which did indeed fail frequently.

I know of no modern electronic equipment that does not use electrolytic capacitors.