Fluorescent Tubes v. LED's

[Bodach na mara]

I have been very interested in this thread and it has produced several good links (which I have bookmarked) however the introduction of the light output in lumens issue was not completely sorted.

Lumens per watt is not the whole story. You need to consider the spatial distribution from the source as well. An incandescent bulb (in a simplified ideal case) radiates equally (almost) in all directions roung a point at its centre. A short fluorescent tube does much the same. Reflector lights redirect the light into a smaller solid angle, thus an incandescent lamp emitting 400 lumens does not light a work surface 2 metres away from it as well as a reflector lamp at the same place with a beam spread of 60 degrees emitting 100 lumens.

As LEDs emit within a fairly narrow cone, even at low lumen values they provide better illumination than incandescents or flourescents of equal lumen output.

 

[FullCircle]

In the boat, we now have Ikea Warm White adjustable angle overheads, far better than original halogens, Warm White BA15S in the bulkhead shaded lighting , but cool white under cabinet galley lighting 970mm long.....

Do be careful using products designed for the home. I'm seeing this more often recently and its potentially DANGEROUS. Please ensure that you understand the implications of modifying lighting products.

IKEA lamps are designed to be used with a mains powered driver that regulates the DC output voltage to exactly that of the lamp. On a boat we don't have regulated voltage it can swing by up to 30%, depending on the particular boat battery and charger characteristics.

LED lamps are current operated devices.Tests I have done on similar lights shows that at 13.8v (typical engine charging voltage), the current flowing through the lamp under test was 50% more than when 12v was applied. Given that modern battery chargers can increase voltages to around 16v when in equalizing mode, my tests at that voltage showed that the current through the lamp had doubled and got very hot, hence my concern. The light output will vary with the current as will the heat produced by the LEDs.

Boat-owners really can unwittingly open themselves to risk. At best, using unprotected lamps results in premature lamp failure and at worst the lamps can result in fire on their boats. What would IKEA say? I'm sure we can all hazard a guess.

I just googled lots of variations of boat led lights fire hazard and not a single hit on the first pages.

Methinks thou dost protesteth too much.

 

[jfm]

I have been very interested in this thread and it has produced several good links (which I have bookmarked) however the introduction of the light output in lumens issue was not completely sorted.

Lumens per watt is not the whole story. You need to consider the spatial distribution from the source as well. An incandescent bulb (in a simplified ideal case) radiates equally (almost) in all directions roung a point at its centre. A short fluorescent tube does much the same. Reflector lights redirect the light into a smaller solid angle, thus an incandescent lamp emitting 400 lumens does not light a work surface 2 metres away from it as well as a reflector lamp at the same place with a beam spread of 60 degrees emitting 100 lumens.

As LEDs emit within a fairly narrow cone, even at low lumen values they provide better illumination than incandescents or flourescents of equal lumen output.

Exactly Wight Dawn. The specification you actually want is the so-called "fixture lumens" of the whole device, not the lumens of the light source

 

[Jabs]

I just googled lots of variations of boat led lights fire hazard and not a single hit on the first pages.

Methinks thou dost protesteth too much.

I don't have any particular axe to grind here. I have worked with LED's for other lighting applications and can confirm that it is quite unusual to just run an LED from a straight voltage, let alone one that can vary over such a wide range, even just 14.6 to 12 is very significant.

I think the light fittings we are talking about retrofitting were probably designed for 15w incandescent bulbs so are probably quite heat resistant.

Be that as it may the issue is wider. If we replace an incandescent with an LED and it is designed to run at, say, 200mA through a constant current device and we place is across a raw dc source the current can increase exponentially with voltage. I have a LED Flood Light that draws 700mA at 12v. This increases to 2100mA at 14 volts. I didn't go further.

Should the LED fail in short circuit we are relying on the fuses in the lighting circuit to protect the installation.

So, when you change to LED are you going to derate your fuses??

Just a thought!

Tony.

 

[Adrian Jones]

I just googled lots of variations of boat led lights fire hazard and not a single hit on the first pages.

Methinks thou dost protesteth too much.

Look, in good faith, I try to give readers free advice. You are perfectly entitled to accept that advice or put forward an alternative point of view and maybe even enter into a discussion. If you wish to disregard that advice, that's OK with me.

As a professional engineer, I feel its my responsibility to highlight potential dangers to anyone, when I see an important issue where danger might arise. The fact is, when non technical people hack up LED lighting products and use them for purposes they were not designed for its blindingly obvious that danger can arise. There are many people who read these threads who are non technical.

Unfortunately you 'imply' to other boat owners that it's quite safe to bodge household LED lighting fittings and use components on their boats. My tests on a number of domestic LED lighting products highlighted and I repeat, at 13.8 volts current flowing through the lamp under test increased by 50%. At 16 volts the current doubled. There really is an issue! I'm sure other engineers will confirm the possibility of potential overheating when LEDs have double the normal working current flowing through them.

Your time 'Googling' may have actually been better spent researching the electrical engineering principles that apply to lighting LEDs. You may then have been able to understand the point I was making rather than posting a comment seemingly intended just to spoil.

 

[FullCircle]

Look, in good faith, I try to give readers free advice. You are perfectly entitled to accept that advice or put forward an alternative point of view and maybe even enter into a discussion. If you wish to disregard that advice, that's OK with me.

As a professional engineer, I feel its my responsibility to highlight potential dangers to anyone, when I see an important issue where danger might arise. The fact is, when non technical people hack up LED lighting products and use them for purposes they were not designed for its blindingly obvious that danger can arise. There are many people who read these threads who are non technical.

Unfortunately you 'imply' to other boat owners that it's quite safe to bodge household LED lighting fittings and use components on their boats. My tests on a number of domestic LED lighting products highlighted and I repeat, at 13.8 volts current flowing through the lamp under test increased by 50%. At 16 volts the current doubled. There really is an issue! I'm sure other engineers will confirm the possibility of potential overheating when LEDs have double the normal working current flowing through them.

Your time 'Googling' may have actually been better spent researching the electrical engineering principles that apply to lighting LEDs. You may then have been able to understand the point I was making rather than posting a comment seemingly intended just to spoil.

Adrian, I am a professional engineer.
I don't imply anything, you have propounded a theory that it is a fire hazard. Well, prove it.

I don't go on hearsay or other less than specific theories or even tests.

Data. Present data. Where is the thermal incident reports, where are your test methods, what temperatures did you get, what were the failure modes, or did they just go pop before bursting into flames.
My boat runs a max of 14.4v measured by a non calibrated instrument.
So, what is your established threshold of safety voltage?
Why aren't you cleverly marketing a 12v modulated power supply so people can buy all the 99p ebay bulbs and use them safely.
You may be missing a trick here.
Unless of course you can't find any actual fires caused by temporarily pumping extra voltage down the pipe, in which case, it is another piece of equipment we don't really need.

You 'imply' that you have exceeded safe use voltage, but 50% over what is a small current is still a small current.



As I said, Methinks thou dost protesteth too much.

Oh, and I am in league with Machurley22. I know him as a friend, and as management of a significant Swedish retailer.

 

[Adrian Jones]

Adrian, I am a professional engineer.
I don't imply anything, you have propounded a theory that it is a fire hazard. Well, prove it.

I don't go on hearsay or other less than specific theories or even tests.

Data. Present data. Where is the thermal incident reports, where are your test methods, what temperatures did you get, what were the failure modes, or did they just go pop before bursting into flames.
My boat runs a max of 14.4v measured by a non calibrated instrument.
So, what is your established threshold of safety voltage?
Why aren't you cleverly marketing a 12v modulated power supply so people can buy all the 99p ebay bulbs and use them safely.
You may be missing a trick here.
Unless of course you can't find any actual fires caused by temporarily pumping extra voltage down the pipe, in which case, it is another piece of equipment we don't really need.

You 'imply' that you have exceeded safe use voltage, but 50% over what is a small current is still a small current.



As I said, Methinks thou dost protesteth too much.

Oh, and I am in league with Machurley22. I know him as a friend, and as management of a significant Swedish retailer.

Swedish retailer.....IKEA by any chance?

I was simply trying to give general advice to others and had not realised that it was you who actually used the bodged IKEA LED lights.. I have just read back and realised that.

As you are a professional engineer and think it was a good idea to fit the lights you now have the perfect opportunity to share with us the current readings you actually got when testing your lights before and after installation. It will be very interesting to see what the variation in current demand was between 12v and 14.4v.

You mention small currents, indeed LED lamp currents are small in normal use, but your not using the product in normal use, that's the whole point of this discussion. You are not using the supplied mains fed LED driver which is designed to control the working current. You have no limiting device other than the inbuilt tiny resistors that are sized for the fixed current provided by the control unit. When the voltage increases the increase in current will not be linear it will be pretty much exponential.

LEDs, as we both know, being professional engineers, are subject to thermal runaway if exposed to voltages positively outside of their design parameters. For the benefit of the non technical readers, those small currents can turn into rather larger currents during thermal runaway. Thermal runaway is usually prevented by either a fixed voltage source, the mains driver unit or where voltage varies a Buck-Boost electronic PWM circuit buit into the lamp. Excess current can result in the simple failure of a lamp, with a blip and if your lucky the particular lamp isolates itself by the fault burning clear and other lamps remain on. However what if the fault does not clear? We would now be relying on the protective device, the circuit fuse or circuit breaker to operate and cut off fault current. Again that's fine if you can locate the spare fuse in the middle of the night, when these things tend to happen! You of course may find that you can't replace the fuse because the fault is still and you blow the second fuse. You then have to check every lamp and prat about removing all the lamps you sweated and cursed when installing, until you find the fault. When I'm sailing with my wife and this were to occur I'd have to explain how I saved money..............I don't really want to go there!

The third scenario is one where after thermal runaway causes just enough current to flow to cause areas of the lamp PCB to glow red hot. A plausible reason and I have seen this, is current tracking that results in a carbon build up on a SMD PCB. This hot glowing PCB is a perfect heat source to ignite any adjacent or touching materials. PVC cabin headers are quite vulnerable in this scenario.

The fact is, we don't actually know what the electrical characteristics are of sub components taken out of light fittings and there is no way of easily finding out. When we use equipment in situations they not designed for you are simply taking chances.

Finally back to my tests I was not doing a scientific experiment, I was testing some lamps for a customer who had asked for my views on a couple of cheap lights he had bought for his boat. I tested the lamps using my workshop calibrated DC power supply. The apparatus simultaneously measures DC voltage and current. It was actually a couple of items, one was a ceiling fitting sold by a high street chain. The other was a an LED rigid strip light, I think sold by Aldi. Both items had a grey box attached to the fixed wiring which contained the mains transformer and LED driver circuitry. I put aside the grey boxes and did tests at the voltages I mentioned earlier and observed the rapid increase in current. The Aldi one got pretty much too hot to hold at 16V when I observed that the current flowing through the lamp had doubled from what it was at 12v. I disconnected it before it failed, sorry I cant provide you with the actual temperature readings. I actually think my recollection of my tests were quite precise.

Lets now see what your tests revealed.

 

[FullCircle]

You see, you haven't referred any significant thermal events, and have failed to present any data other than your own pratting about.
I use that word, as you insist on using bodge.

I presume that you don't CE mark your own manufactured product?
No I thought not, you get your supply from a source I bet you haven't ever visited, and probably don't really care if they manufacture to international standards like QS9000 or TS16949. You just buy it in.
Have you got any accreditation?

 

[KAL]

This is going down like a LED balloon...

 

[northwind]

Why aren't you cleverly marketing a 12v modulated power supply so people can buy all the 99p ebay bulbs and use them safely.

You just need to add a voltage regulator... Dead easy and cheap to do...

http://uk.rs-online.com/web/p/products/516-4834/

 

[pyrojames]

You see, you haven't referred any significant thermal events, and have failed to present any data other than your own pratting about.
I use that word, as you insist on using bodge.

I presume that you don't CE mark your own manufactured product?
No I thought not, you get your supply from a source I bet you haven't ever visited, and probably don't really care if they manufacture to international standards like QS9000 or TS16949. You just buy it in.
Have you got any accreditation?

I eagerly await all these fires that I will have to visit now that unregulated LEDs are flooding the market!

 

[Adrian Jones]

You see, you haven't referred any significant thermal events, and have failed to present any data other than your own pratting about.
I use that word, as you insist on using bodge.

I presume that you don't CE mark your own manufactured product?
No I thought not, you get your supply from a source I bet you haven't ever visited, and probably don't really care if they manufacture to international standards like QS9000 or TS16949. You just buy it in.
Have you got any accreditation?

I'm sorry I really can be bothered to trade insults and respond to your rather rude and incoherent rambling. If you want to risk your boat, just get on with it. Just remember to keep those fire extinguishers in date.

I'm sure most forum members reading this exchange will judge the merits of the issue.

Good sailing.

 

[FullCircle]

I'm sorry I really can be bothered to trade insults and respond to your rather rude and incoherent rambling. If you want to risk your boat, just get on with it. Just remember to keep those fire extinguishers in date.

I'm sure most forum members reading this exchange will judge the merits of the issue.

Good sailing.

You started the insults Adrian, happy to descend to your level.

You haven't proved anything here, and still have not produced data, or established any real world incidents.
So I am hardly incoherent, hm?

You didn't even pick up on the business opportunity to offer inline voltage stabilisers to the marine fraternity.
Some people just aren't very thankful.

My fire extinguishers meet EC regulations, and are in date. Thank you.

 

[LadyInBed]

My electronics is pretty much out of date these days.
Do LED's tend to go open circuit or short out?
I was under the impression that diodes tend to go open cct.
Please note that I said tend and don't infer that I mean always

 

[ytd]

The G4 LEDs I got off ebay had a circuit board on the back with about 20 components including a small IC. I presume this was a current limiting circuit. Our voltage does get up to 16V when the smart regulator equalises and no bulb failures after a year of operation. I had a spare G4 LED and replaced 12V AC transformer supplied halogen at home with it. Lasted about 2 months. Later I read somewhere that LEDs designed for 12V DC operation don't like 12V AC.

 

[Sybaris]

You see, you haven't referred any significant thermal events, and have failed to present any data other than your own pratting about.

Adrian is pointing something out which makes perfect sense and in the process he is helping out forum members who are looking for information. As a response he gets attacked by FullCircle! Totally weird.

Cheers,
Per

 

[Conachair]

You just need to add a voltage regulator... Dead easy and cheap to do...

http://uk.rs-online.com/web/p/products/516-4834/

Maybe not a bad halfway house but you loose 2v along the way. There might be chips with less drop out voltage around. And doesn't allow for the change in forward voltage in the led as the temp rises.
But not expensive and might stop you blowing a load of cheap ebay lamps when you forget to turn them off equalising the batteries.

 

[Adrian Jones]

The G4 LEDs I got off ebay had a circuit board on the back with about 20 components including a small IC. I presume this was a current limiting circuit. Our voltage does get up to 16V when the smart regulator equalises and no bulb failures after a year of operation. I had a spare G4 LED and replaced 12V AC transformer supplied halogen at home with it. Lasted about 2 months. Later I read somewhere that LEDs designed for 12V DC operation don't like 12V AC.

It sounds as if the lamp you purchased had all of the necessary constant- current controls in place. I'm interested to know if you were aware that it was protected when you bought it from Ebay and whether it was inexpensive. Often similar lamps are sold that only have resistors in place to control working current. These tend to be the 99p type. With constant-current controls it was therefore perfectly able to to control current across a range of applied voltage whilst maintaining steady light output.

I'm actually pleased that you have independently confirmed that your boat voltage does indeed rise to 16V. This is now quite common on modern good quality battery chargers and does need to be considered when purchasing LED lamps.

The issue of using DC LED lamps on domestic halogen lamp transformers is a thorny one. Often halogen lamps use a power supply which is in fact an electronic transformer. These utilize high-power switching semiconductors to interrupt the mains voltage at high frequency and create a square waveform. This waveform is then fed into a small wound transformer and converted to 12v the out put of which is still a square wave at 20-30-KHz. These are the ones most likely to give problems with LED bulbs because the electronics have been designed to work properly across a certain range of power loads. Typically they are designed for loads between 20 and 60 watts, which means that the electronic transformer will only operate within its specification when connected to a load which is somewhere within that range.

LED bulbs draw so little current that they often won't reach the minimum wattage necessary, consequently the electronic transformer can't be depended on to operate within its specification. It is possible to add a string of resistor controlled LED 12V lamps to a standard halogen transformer with one lamp being a normal 20W halogen. However when the halogen lamp fails the LEDs wont work! There is also the issue of different quality of light given out when mixing light type. When you use a constant current transformer protected LED the square wave is rectified, as rectifiers are in place to avoid polarity problems (on quality lamps!). You then have KHz pulses hitting the on board PWM electronics. I have seen some lamps work perfectly OK whilst others are seen to noticeably flicker. Overall, I have concluded that use of electronic halogen transformers results in a reduced service life for 12v LED protected lamps. I would suggest using a mains fed LED driver unit which are available at pretty much the same price of their halogen equivalents.

In principle I suggest that it's never a good idea to operate any electronic device outside of its specification.

Hope that helps.

 

[jordanbasset]

I have to disagree with you: warm white lamps are the spawn of the devil! Why would anybody want a light that imitates a candle, rather than the sun?

Finally some one else who agrees with me, was beginning to think me and my wife must have strange eyesight. Have found cool white lights excellent, especially for reading and no problems for ambient light.
But you pays your money and takes your choice

 

[Adrian Jones]

Finally some one else who agrees with me, was beginning to think me and my wife must have strange eyesight. Have found cool white lights excellent, especially for reading and no problems for ambient light.
But you pays your money and takes your choice

Cool white LED lights are actually brighter than the warm white variety, hence your ability to read more easily.

Manufacturers claim a warm white replacement 6W LED tube lamp for a 12in fluorescent fitting outputs 420 lumens whilst a cool white lamp of the same rating emits 550 lumens. This is all to do with the phosphor coating on the LED, a warm white one loses some light due to the colour conversion process.