12V LED downlight group test

Thought about limiting the voltage of the circuit? A buck boost dc converter costs less than a bottle of beer from eBay.
I have indeed. The usual way to do this is to limit the current passing through the diodes, with a "constant current" LED driver. Another option is to stabilise the voltage, as you suggest, either at the supply end or at each LED light. If you do this at the supply end and have a long chain of lights you may find that they go dimmer towards the end of the chain. Individual buck/boost converters is the best way to do it, and indeed that's what most "automotive" LED modules use - often cast into the same piece of resin that contains the LED itself. Another benefit of using a boost capable DC/DC converter is that the lights will continue to work even at very low battery voltages.

There is a third option however, and that's to use a "Pulse Width Modulated" voltage to drive the LEDs. PWM simply means that the light is switched on and off very quickly (typically hundreds or thousands of times per second), and by changing the ratio between how long it is on vs. how long it is off (called "duty cycle") you can control how bright the light appears. I say "appears", because it's really turning on at full brightness, only for such brief periods that we perceive the light as being dimmer. This is the method I have intended to use, mainly because it requires very few components and is easy to build, and should work with most, if not all 12V LED lights - not just those designed for automotive use. Each light will be equipped with a small PCB containing a microcontroller and a transistor and not much else; the microcrontroller (a tiny single chip computer) generates a PWM signal based on input received from the light switch (short press for on/off, long press to dim up/down), and this is used to rapidly turn the transistor on and off, which in turn supplies power to the LED.

Now this is where it gets interesting. Although I can reduce the amount of energy delivered to the LEDs by lowering the duty cycle, I wanted to make sure that the lights I had chosen would not be destroyed - or set the boat on fire - when run at a 100% (or 99%) duty cycle at an elevated voltage. That's why I did the "stress test" at 14.5V, which seemed to disqualify the ELD lights I wanted to use. BUT. I just tested them again with a 14.5V supply that was PWM:ed at 1kHZ with a 99% duty cycle and they don't get hot at all. Even covering the back of the light with a folded up cloth, to simulate it being installed in the ceiling, I couldn't get the temperature up over 40 degrees. This seems pretty counter intuitive to me, since a 99% duty cycle means the LEDs are consuming 99% as much energy as they did when fed a steady DC voltage, and I don't have a good explanation for why there is such a big difference in temperature. It could be the voltage drop introduced by the transistor, but in any case it's very welcome, as it means I can safely go ahead with the ELD lights.

Interesting and useful. Thanks for going to the trouble of reporting your findings.:encouragement:

My pleasure!
 
Last edited:
Would be grateful if you could providd a link to one on ebay so i know the correct type for this application. Thanks.

Hard to say what would be suitable without knowing more about your intended use, but here area a few examples:

http://www.ebay.co.uk/itm/1-pc-Conv...-18V-Uout-12VDC-Iout-1-25A-180g-/172492170796
http://www.ebay.co.uk/itm/DROK-DC-B...y-80W-12V-24V-Voltage-Regulator-/222438161818
http://www.ebay.co.uk/itm/25W-DC-DC-converter-9-2-18V-12V-2-1A-enclosed-/112197139452
http://www.smgeurope.com/Marine/product-details.cfm?p=3436&s=136
http://www.powerstream.com/DC3611.htm

Basically, any DC/DC converter that accepts a ~9V to ~18V input, and either has a fixed 12V output, or one which can be adjusted (to 12V) will work, but they come in different sizes depending on how much current you need. Quality also varies a fair bit. It's worth looking for one that has a high efficiency - anything 90% and up is good.

Edit: I should add, perhaps, that adjusting the voltage is not a good way to dim LEDs - if you want to be able to adjust the brightness you need either a constant current or a PWM dimmer (or one that does both!). Either type should be able to operate on typical battery voltages (11.5V to 14.5V), so would negate the need for voltage stabilisation.
 
Last edited:
There is a third option however, and that's to use a "Pulse Width Modulated" voltage to drive the LEDs. PWM simply means that the light is switched on and off very quickly (typically hundreds or thousands of times per second), and by changing the ratio between how long it is on vs. how long it is off (called "duty cycle") you can control how bright the light appears. I say "appears", because it's really turning on at full brightness, only for such brief periods that we perceive the light as being dimmer. This is the method I have intended to use, mainly because it requires very few components and is easy to build, and should work with most, if not all 12V LED lights - not just those designed for automotive use.

You need to be a bit careful with pure pwm as it can make a mess on VHF/HF radio, the ebay buck & buck/boost are basically smoothed PWM anyway, I've found some of the cheaper ones need a capacitor on the in and out to filter a bit more noise out.
 
You need to be a bit careful with pure pwm as it can make a mess on VHF/HF radio, the ebay buck & buck/boost are basically smoothed PWM anyway, I've found some of the cheaper ones need a capacitor on the in and out to filter a bit more noise out.

There should be no noise when doing pure PWM of a resistive load. I think the DC/DC converter is the culprit if you're having noise issues.
 
There should be no noise when doing pure PWM of a resistive load. I think the DC/DC converter is the culprit if you're having noise issues.

I used to get interference on my VHF radio from my solar panel PWM regulator until I added a large electrolytic capacitor to the output.

I do get some interference on my TV (beat frequenct lines on the screen) when connected to a satellite decoder driven by a constant 12Vcd buck boost regulation from E-Bay.
 
I used to get interference on my VHF radio from my solar panel PWM regulator until I added a large electrolytic capacitor to the output.
Odd; your battery is a huge capacitor - I don't see what difference adding a few uF would do. Unless maybe you have really long cables from your PWM regulator to the battery bank, and they pass close to your VHF or its antenna cable.

I do get some interference on my TV (beat frequenct lines on the screen) when connected to a satellite decoder driven by a constant 12Vcd buck boost regulation from E-Bay.

Unsurprising, given the quality of a lot of the stuff that's sold on eBay. Here you may see an improvement by adding a filter cap.
 
Top