110v Lights to 12v / Removing Transformer

[belizesailor]

Originally Posted by sarafina

...
The transformer steps the power down to 12v, so why do I need to change the bulbs? ...

You don't.

[sarafina]

Oh, This is just getting TOO good!

So the one issue that keeps being raised is running the lights while the batteries are charging.

But we would only swutch the lights onto the 12v line when we were not at the dock. And thus would not be charging the battery banks.

So I can have dual power lights all in one installation.

Why isn't everyone doing this?

[Paperbird]

What's the model name of the Ikea goose neck lights? We've been looking for some Ikea LED reading lights for the aft bunk. Given our experience with the Inreda overhead lights, I'm game for adapting lkea reading lights as well.

Thanks!

[erie bound]

Just remember that transformers can be fed in reverse. If you don't make sure your 120v source is switched off, when you are on battery power, you might see 120v @ the prongs of the plug. Depends on internal wiring of transformer and location you tie into 12V. I.e. before or after any switches.

[Ex-Calif]

1 - I am not convinced the "12v" LED can't handle charging voltages up to 14.6 or whatever.

I have installed LED lights for guys and didn't even consider the issue. Genny, shore power, alternator or batts didn't seem to fuss them. YMMV

Sara - While the switch will likely work, why the complexity? If you go only 12v and you are at shore presumably your battery charger is plugged in and you have an endless supply of 12v?

Forgetting to switch etc. etc.

[HopCar]

I'm with Ex-Calif, I don't think LEDs are all that voltage sensitive. There are plenty of LED running lights that have to be on while the engine is running and charging the batteries. They seem to last pretty good.

Halogen bulbs may be much more sensitive. We sell a heck of a lot of replacement halogen bulbs.

Tellie, when I take a shot at you, you'll know it!

[Dsanduril]

I don't think LEDs are all that voltage sensitive. There are plenty of LED running lights that have to be on while the engine is running and charging the batteries. They seem to last pretty good.

If they are good bulb assemblies then they have built-in current limiting circuits. These are an absolute necessity when LED bulbs are used over varying voltages. You only have two options; maintain the voltage exactly at design levels, or include a current-limiting circuit in the bulb. For a decent discussion, see here:

LED Magazine

In particular:

too much current and voltage can damage the light-emitting junction of the LED diode

a small change in forward voltage produces a disproportionately large change in forward current

As LEDs heat up, the forward voltage drops (Figure 2) and the current passing through the LED increases. The increased current generates additional heating of the junction. If nothing limits the current, the junction will fail due to the heat. This phenomenon is referred to as thermal runaway.

Good LED assemblies have current limiting circuitry. Cheap LED assemblies do not. If you look at the specs for some of the various LED running lights, you will see that several tout themselves as better than the others because of their current limiting circuits. Given their prices I suspect the IKEA lights either use nothing or a resistor as their current limiting circuit, neither of which is truly current limiting and both of which will be susceptible to voltage changes. Since they are designed to be run by the IKEA power supply, that is not a design flaw. As long as the two are matched all is good. Put them in the marine power environment where voltages can vary by 20% (11.8 to 14.4) and they may not do so well.

Having said all of that, I am curious about experimentation. If the IKEA light out of the box has a 20,000 hour life, and running it at 14.4V reduces that to 15,000 will I ever know? If, on the other hand, it reduces it to 10 or 50 or 100 then I probably will.

[Delancey]

I've attached a photo of an Ikea LED I installed in a common Perko dome light as an experiment, anyone happen to know if the small black rectangular shaped components are resistors?

They are much smaller than what I know a resistor to look like. I am currently charging my batteries off of shore power and have yet to see how they perform while charging from my engine.

With regards to the service life of LED's, it is my understanding that 'white' light is produced by using a blue LED with a phosphor coating similar to a fluorescent light.

This coating will degrade over time resulting in a color shift from 'warm white' to a 'blue' light. Something to think about when considering the service life of these lights, if you are sensitive to color you may find yourself replacing them sooner than expected.

I believe running them hot (higher voltage) is bad for LED's in general but is also particularly bad for the phosphor coating used in the white ones.

Smoking Bulb - YouTube

[Dsanduril]

anyone happen to know if the small black rectangular shaped components are resistors?

Looks like it, probably in what the industry calls an "0603 package", although could be one of the other 0xxx packages depending on exact physical size.



With a resistor as the current limiter the current will follow E=IR, which means a 20% rise in voltage during charging will result in a 20% rise in current through the LEDs.

Will be interested to see if that results in any measurable change in life.

[Wotname]

Originally Posted by erie bound

Just remember that transformers can be fed in reverse. If you don't make sure your 120v source is switched off, when you are on battery power, you might see 120v @ the prongs of the plug. Depends on internal wiring of transformer and location you tie into 12V. I.e. before or after any switches.

While I haven't studied this thread in great detail, I am sure you will find that the 12 V source will be a DC source. The transformer would only "transform" this to 120 V if the 12 V was AC.

[SoonerSailor]

Those small black rectangles are likely resistors put in place to prevent gross current imbalances to the groups of LEDs placed in parallel to each other. I suspect each group is 3 LEDs in series with a current limiting resistor, then the 6 groups placed in parallel.

Life at higher than 12 volts will depend on how hard they are pushing the LEDs at 12 V input, and the temperatures they run at. At the price, seems like it is worth the experiment.

[Tellie]

Again, two years running IKEA lights and haven't replaced one yet. Maybe I'm just lucky,...if I am, it's about time.

[Dsanduril]

I suspect each group is 3 LEDs in series with a current limiting resistor, then the 6 groups placed in parallel.

Exactly. It looks like the LEDs are 0603 LEDs with a forward current rating of 5mA and a nominal forward voltage of 3.0V. With 6 parallel strings that works out to 30mA total, at 12Volts that works out to 3.6W, which is slightly greater than IKEA's 3.0W label.

If you work backwards there are three LEDs in series, which means 9V required to drive them. The resistor has to drop 12V to 9V at 5mA => 3V drop =>3V/0.005A = 600Ohm resistor.

If you now apply 14.4V you have to play around with the I/V curve and the resistance to get the new figures.



From the I/V curve it looks like you get to 8mA at about 3.15V, so 14.4-3*3.15 = 4.95V drop across resistor, => 8.25mA current (close enough, since we don't know exact model #'s, etc). Therefore, running the lamp at 14.4V will cause each LED to run at ~8mA and 3.15V instead of the design 5mA and 3.0V. The total power dissipated by the LED will be 3.15*8 = 25mW, rather than the design level of 3*5 = 15mW. In other words the LED will be dissipating 25/15 = 167% of it's design power.

Typical values for these LEDs are 20mA max current and 70mW max dissipation, so it would appear that they are not in danger of instantaneous failure, but how much this kind of use will affect overall life is not really published in the specs. Will be interested to hear the results of the experiment.

Mind you, all of this is to save about $3 on the manufacturing end, that's the difference between the cost of a real current limiting circuit and the resistors that are used in the less expensive variety.

[SoonerSailor]

One nice thing about current limiting resistors is that they produce no RFI. True constant current circuits have to be very thoughtfully designed to be radio quiet, and they typically aren't.

[Dsanduril]

true constant current circuits have to be very thoughtfully designed to be radio quiet

Great point. At $10-20 a light these things are really pretty cheap, so if running them part of their life at +/- 150% of their rated capacity reduces their life from 20,000 hours (IKEA spec) down to 5,000 or 10,000 then maybe it all doesn't matter.

two years running IKEA lights and haven't replaced one yet

is a good sign, assume that over that time voltage has been all over the place. Sara's question is probably a good one, "why isn't everyone doing this?"

I'm tempted to see if I can find time to hook one up at home this weekend and run it 24/7 at 15V to see what happens as a worst case test. If I can find the time... So many more worthy boat projects to do, but this sounds fun.