LED light bulbs: should CHOICE keep testing them?

Home Improvement
22

Regarding RF interference from LED lights: You’re quite right that RFI from lighting can be a problem for TVs, radios etc. We have conducted informal RFI testing in the past (and found no particular problems) but a more thorough test would be required to identify any problems, and this would require specialised lab equipment, lab space and expertise, none of which we currently have. Accurate assessment of RFI and EMF issues is a complex business. In addition, problems can be specific to particular locations (such as areas with weak reception) and to particular configurations of wiring and cabling within the home.

I agree that cheaper LED lights are more likely to cause RFI. A friend of mine had significant RFI issues in her home,mainly TV reception, caused by LED downlights. If you switched on the hall lights for example, several channels turned to snow. We found the lights were unlabelled and quite possibly non-compliant with local standards. The installer (who had sub-contracted the lighting installation to someone else) agreed and replaced them with name-brand products, Osram I think, and the RFI issues went away immediately.

5 Likes
23

Grahroll,

Thanks for your reply. The real problem lies with the Government letting all sorts of switch mode devices into the country. There use to be a time when manufacturers had to submit samples of anything electrical / electronics which were tested for RFI. That was the only way a manufacturer could obtain a C-tick for their product(s). Now it has become self regulating and testing is no longer required leaving it wide open. That is why we see a lot of cheap and nasty pieces of electronics (including LED devices), switch mode power supplies, inverters etc.

The above does not apply to all products. There fortunately are manufacturers out there producing fine electronic devices which do not produce RFI.

Also, all Hams have studied electronics in order to obtain their Ham licence and we are quite capable of solving most issues re. this subject. But why should we have to? El-cheapo manufacturers need to be pulled in line and a testing report from choice re. LED lamps showing RFI results will be a step in the right direction.
Gerard

6 Likes
24

I support Choice keeping the players honest. While I’d like to see periodic bulb testing to continue, I’m thinking more from the standpoint of meeting specified brightness, power efficiency, & brand reliability.
My experience is that a proportion of LED bulbs fail within a year. That number has decreased fast in last couple of years.
A major factor was how well the dimmable globes work in existing dimmers, which generally pre-date the LED technology. Some work fine, others flicker at some dimmer settings - usually full on. I’ve had the same problem with a recent new LED compatible dimmer. Interestingly, my first LED purchase many years ago was an expensive Philips LED (the ones that look yellow when off). Still going well, works perfectly in all the dimmers. It’s only problem is a brief switch on lag.

2 Likes
25

I think you’re right about directing your resources elsewhere.
I recently changed a lot of my lighting to LED and have only seen one significant issue. The LED lamps are connected to a dimmer a unit rated for this type of load. After about 30 minutes they begin flickering then settle down for many minutes. I get the same thing on two different circuits one with led general light bulbs and one an LED smd fitting so I suspect that the dimmers are at fault. Perhaps there is some interference between the drivers and the dimmer or perhaps its the dimmers being sensitive to mains noise??? Not the end of the world but annoying.

1 Like
26

SMD = Surface Mounted Diode (lot of individual chips)
as opposed to the other technology
COB = Chip On Board (one large LED - generally yellow)

1 Like
27

LEDs continue to be evolving rapidly and could spend forever testing.
While there is no ‘standard’ for testing the life and lumen output of LED, most manufacturers refer to ‘Approved Methods’ although there is not requirement to abide - as there is with an IEC Standard to which Manufacturers of CFL, Linear Fluroescent and High Intensity Discharge lamps are tested to.

LM79 - The Approved Method Electrical & Photometric Measurements Of Solid State Lighting Product
LM80- The Approved Method Measuring Lumen Maintenance of LED Light Sources
TM21 - Lumen Degredation Life Time Estimation Method For LED Light Sources
When Choice was testing LED ‘lamps’ than should have tested under the same conditions as the manufacturers do which is the information used to make their lumen output and life claims.
Any other testing criteria (turning on and off etc over certain time) makes any results irrelevant as a comparison.

When testing fixtures need to ensure:
Junction temperature of the chips are with in spec - too hot will reduce lumen output and life.
Fixture complies with EMC requirements
If remote driver both fixture and driver must carry EMC
Driver and chip compatibitity with dimmers (leading edge, trailing edge and universal)

2 Likes
28

That is not precisely what COB is, but visually that is what they appear to be. The ‘C’ is chips, not chip although they are commonly called chip on board.

1 Like
29

While there appears to be some consensus that testing of Led general bulbs can be discontinued, I do not find the same reliabilty in led down lights being either gu10 (240type) or mr16 (12v type).
I would be interested in seeing some testing on these.
While I’m aware that LEDs should not run for extended periods, I have family members that need better training and need to turn them off more religiously. Can we see testing on what bulbs survive better if occasionly left on for extended periods?

1 Like
30

For the many who posted in favour of testing downlights, but might not have read the OP.

2 Likes
31

I think it reasonable to stop testing the globes and concentrate on the downlights too.

32

I also have concerns about the actual useful light output from an LED in terms of the cast area of illumination. By that I mean how evenly the light is being cast onto a target area. Many LED lights, especially little ones like torch-lights show a concentrated beam in the middle and much weaker illumination outside that core. This badly balanced cast creates too much contrast and makes the light less useful than a conventional light bulb and a CFL.

1 Like
33

I also agree with your premise of no longer testing general use LED lights.

I must add though that our bedroom lights, which are only turned on for short periods (no kids living at home and studying!) and the old incandescent globes just won’t give up! We have not changed a globe in these lights for many years, and don’t intend changing them whilst they still work. We have installed LED lights in our kitchen/dining area and use the low energy compact globes in any light that is left on for more than a few minutes.

1 Like
34

We lost that battle decades ago - and with the ever increasing commercial use of digital with it’s associated error correction/etc, the government seems to care even less than it’s original zero care factor years ago. People like us using analogue modes are becoming rarer and rarer. Touch lamps, ethernet over powerline to name a couple of bigger killers, but then there’s the garden variety interference from commonplace items - HF interference everywhere …

And with the way the Witless Destitute of Australia is self destructing - hams are nearly extinct I’d reckon … sadly.

a VK5 exiled to VK8 …

2 Likes
35

Well, yes and no.
I had the pleasure of activating one nature reserve and one national park today during the VKFF activation weekend and had over 100 contacts on 40m and 20m combined.

HF is a live as well.

Gerard

2 Likes
36

You sound quite turned off by the experience, but I wouldn’t want to be left totally in the dark. (I realise there are readers who take a dim view of puns, and should also apologise if I have LED this discussion astray - there’s always one bright spark…)

On a more serious note (B♭), I must admit to some brief puzzlement when reading those test scores. The worst-performing bulb has a “failure score” of 0%, which the review goes on to clarify means most of the samples failed. Normally, one would expect a failure score of 100% to be A Bad Thing - but not here. Is this perhaps an area where Choice can consider the way it presents some of the results? That is, instead of a failure score would readers better understand a failure rate? Maybe a ‘lasted the test’ score?

In the same review, it refers to “Efficiency at 13,800 hours” in lm/W. This is presumably the correct terminology, and there is a pop-up information box to tell me a little more about what it means, but could this kind of information be presented as a proportion of claimed efficiency - or initial efficiency - or… some other way that means I don’t have to read the manual?

Thirdly (I’m sorry - I just keep noticing these. I’m not looking for them!), in the title of each reviewed globe you state its wattage. Example: “12W wattage”. Please feel free to tell me what the W stands for if not watt. If watt, then why ‘wattage’? (Boy, those two sentences were fun to read back.) There appears to be some redundancy here.

Finally (still on the same review - this time on the top performer), am I correct in understanding that this was brighter than claimed by the manufacturer? That is (from the review):

Claimed light output (lm) 1055
Measured light output after 100 hours (lm) 1160
Measured light output after 13,800 hours (lm) 1102

So at both measured points, it was brighter than claimed? Does Choice consider this a good thing, or a bad thing?

This of course would be a problem, although I am not familiar with the story regarding incandescent globes - it is more something I have heard said of nylon stockings. Source?

The other thing I would say about potential racketeering/cartels is that I suspect there are far too many LED manufacturers now for this to occur; they’re not like some of the more complex electronics, where only a few companies have the capability.

This is something printer manufacturers have been trying to do with that most expensive substance on Earth - ink - and so far the backlash has been loud enough to stop it. I will leave my view of ‘free’ ‘trade’ ‘agreements’ (yes, that formatting is deliberate) for another time.

And don’t forget that the movie industry has been relying upon encryption since the early days of VHS vs. Betamax and on into ‘regional encoding’ for DVDs and BluRay; hasn’t really helped it much either. The encryption has to be performed locally, and so can and will be broken - regardless of outdated laws. The same would be true of light bulbs.

2 Likes
37

I agree that it is a relative waste of resources and space to be testing LED lights that are standard replacements for incandescent globes in edison screw or bayonet socket light fittings. These now sell for around $5 a piece. With an expected life of 20,000 to 50,000 hours, it will take you a long time to check their durability. We are way past the 1,000 hours (two months @ 24 hours per day) mark and into the 2 - 5 years of testing to check and you need five or so of each to get a reasonable statistical sample. I suspect hosting an anecdotal customer comments/review site for particular brands would be fine as a forum. For example, I have found the Aldi globes generally to be good with no failures and reasonable dimming range. I’ve bought small edison screw candles up to 12W 1000 lumen globes in warm white all with no problem.

When it comes to downlights, 12V LED 4W or 5W seem to be the standard fit out but there are plenty of variations and it is quite confusing. From experience, I find the transformers are the problem - there are plenty of cheap transformers that are hard wired to a circuit and these may well fail. It then isn’t a simple job of replacing a globe but a need to call the electrician. Unfortunately, these are invisible to most people as they are in the ceiling. The transformers are cheap but the labour isn’t! The downlight tglobes are already really cheap and inexpensive to replace with most people who feel comfortable up a ladder able to swap them out.

Architects and customers specify the type of light they want but the bit where the real risk and cost is involved is usually left to the electrician who picks the cheapest unit from their wholesaler who gets them from an importer who buys in bulk from a Chinese manufacturer. Unlike electric cable which is certified to an Australian standard, these drivers/transformers are not subject to any real control/validation. They could run overly hot (real risk there) or not survive long (real cost there).

I’d much rather you looked at the driver/transformers and gave them a real hammering with lots of on/off cycles and dimming and other shock treatment (pun intended) such as a slight surge in voltage to emulate real world scenarios. As these are supposedly built to a universal standard, they should be capable of taking from 90V to 270V and 50 - 60 Hz +/- 20% without faltering.

3 Likes
38

Reputable electricians really should install downlight drivers (they are voltage converters, not transformers) by first installing socket outlets, and then connecting the drivers to them, rather than just hardwiring the drivers to the mains wiring. The electrician call-out fee is always much more than any additional hardware such as a simple 3-pin batten socket.

Our electrician did this when we had our kitchen and bathroom renovated 10 years ago. When we changed the halogen downlights to LEDs, I could replace the inappropriate 20-50W drivers with 0-20W ones myself. Similarly, if one ever fails in future, I can replace it myself without calling out an electrician.
If you ever want to replace a downlight with the LED type that has the driver built-in to the fitting, then this will also be a piece of cake.

Of course, you need to be comfortable with accessing your ceiling space (your electrician can advise whether it is a safe place to be - electrically speaking - and address any issue if it isn’t), and as always, YMMV. :grin:

2 Likes
39

You are indeed correct. It is usually easier to use a ladder to change a light or transformer/driver connected to a plug than to climb into the roof void and walk across the ceiling rafters - too much plaster to put my foot through!

Having seen new builds of apartments, these usually do not have 3-pin batten sockets but are hard wired, presumably to save money.

I always insist on batten sockets for my lights if they are being changed over for any reason. I am in a 120 year old home which was rewired in 1997. I didn’t think of it then but I am about 85% upgraded with batten sockets now.

1 Like
40

I feel that commonly used domestic LED “bulbs” have matured and are now of an acceptable reliability. Certainly far better than incandescent or CFL or halogen bulbs.

So I suggest it is time to close down the test and use the time for other activities.

Thank you for the thorough testing you’ve done.

3 Likes
41

I still have a compact fluoro bulb bought in the 1980s that still works! It is no longer plugged in as we have all LEDs these days.
Compact fluoro bulbs were notorious for causing radio interference (AM radio, FM radio) - they operated at 4800 Hz rather than the 50 Hz of conventional fluorescent tubes. I’ve found LEDs to be much better (less interference, less heat, less energy consumption)

1 Like