 | Home > Technical > LED Headlight operation |
 
|
|
|
| sako243 Member Since: 08 Jul 2014 Location: Wales Posts: 1218   
|
Rather than go off-topic on this thread I thought it wiser to start another one.
Please note that this comes from my experience as an electronics engineer with reasonable experience in automotive electronics rather than any particular expertise with LED headlights, most of that side of things is educated guesswork... Fundamentally LEDs are current controlled devices. Put simply unlike conventional bulbs their brightness is a function of current not voltage. However they do have what is referred to as a "forward voltage" (Vf), this is effectively the minimum voltage required to turn them on. It is a property of the type of LED and is dependent mainly on the colour of the LED, it doesn't matter whether its 100mW or 10W if they're both red then they'll have the same forward voltage. These are typically of the order or 1.5-2.5V depending on colour of light emitted. It's not quite as straightforward as that because Vf is dependent on other factors such as temperature and current flowing through the device. The brightness of the LED is governed by the current flowing through the device, in trivial circuits when you're learning electronics and likely playing with microcontrollers you will see a current limit resistor in series with the LED. This will be there to control the amount of current flowing into the LED, without it you can burn out the LED by letting it draw too much current. Now from an efficiency point of view a current limit resistor is a poor choice because you just waste the energy as heat. Given that all forward voltages are of the order of 2V for a single LED we can do some simple maths 
Assume (not unrealistically) that the LED in a typical headlight is 20W (ignoring losses we assume that 20W goes into the production of light). Let us also assume that we see 12V at the headlight connector and the Vf of the LED is 2V. If we were to simply connect the LED across the battery this means that the LED drops 2V but there is still a 10V discrepancy, to try and rectify this (assuming nothing else connected) this means that more current must flow through the LED to try and eliminate this discrepancy. Given a lead acid battery can deliver 100s (or in the case of Defender batteries more like 1000A) very quickly you would get an attempt to put 100s of amps through the LED and then you're pretty much guaranteed to blow it. Or in the case of a Defender you're more likely to melt the wiring harness first... To prevent this we need something to "drop" the 10V, the simplest device would be a resistor. So we now drop 10V across the newly installed current limit resistor. The LED is drawing 20W nominally so at 2V requires 10A, to limit current to this we can then use Ohms Law (V=IR). So 10V across the resistor at 10A means a 1 Ohm resistor. Simple enough to source, but... The power dissipated in the resistor is given by P=(I^2)R, so P=(10*10)*1=100W being used to be at up the resistor. So our 20W headlight is now actually requiring 120W to function 
Since the killer here is current we can put multiple LEDs in series, this will stack up forward voltages but because they are in series the same current is required. Lets put 5 in series. Now we have 5x2V=10V being dropped across the LED array, leaving us with 2V across the resistor. Resulting in a 0.2 Ohm resistor. This now dissipates 20W, much better but still not great. As an aside if we limited the current to 5A then we would affect the light output of the LED. In reality using a simple resistor as you've seen is far from ideal, in particular temperature will affect lots of things such as the resistance of the resistor, the relationship between current through the LED vs light output etc. So as the temperature of the components change the brightness of the headlights will be affected (trivially it would likely be lower for higher temperatures). So solve this you almost always in any non trivial circuit use a current source (will output a fixed current irrespective of other factors) or in this context it would be referred to as an LED driver. This basically an active circuit that constantly monitors the output current and regulates it to a fixed value, more advanced ones can compensate for temperature to maintain constant light intensity. These will likely be integrated into the headlight unit themselves. They will take the input voltage and regulate it to produce the correct output, for so simplicity / cost of manufacture this will almost only be a buck converter, I.e. It can only chop the voltage to a lower level and not "boost" it to a higher level. There are plenty of buck-boost supplies available but there's simply no point in this scenario. According to JWs datasheet it requires 9-32V input and 3A. Going back to the thread that prompted this if the voltage seen by the headlight unit drops below 9V then although the LED will only likely require a couple of volts to operate the LED driver may not so you'll find it cutting out. This will drop the load on the battery and likely cause the voltage to rise again, cutting in the headlight which starts the process all over again. Hence the flickering. So in summary uprated headlight looms may not be needed for LED headlights but it won't hurt to have them fitted, ultimately it'll make things a bit more efficient since there's less loss than in the thin cables as fitted by default. Have a Merry Christmas, time for some sleep now the puppies and Lyra have settled down. Ed 82 Hotspur Sandringham 6x6 95 Defender 110 300Tdi |
||
25th Dec 2016 5:15am |
|
| LR90XS2011 Member Since: 05 Apr 2011 Location: bickenhill Posts: 3641  
|
interesting that explains why some have had flickering LEDs when voltage drops DEFENDER 90 TDCI XS,
I hope everyone is well and your land rovers make you happy |
||
|
25th Dec 2016 6:14am |
|
| mark_d Member Since: 07 Sep 2013 Location: Northern Ireland Posts: 266 
|
I thought they would use a DC to DC (Buck/Boost) converter to regulate the voltage fed to the LEDs. It is an efficient way to give a very wide input voltage range without wasting energy.
This uses a principal similar to pulse width modulation at a fast switching frequency to convert one voltage into another. If the input voltage drops below what the buck converter can support, you would get flickering because of the high frequency switching nature of the power supply. http://defender90xs.blogspot.co.uk/ |
||
|
25th Dec 2016 8:31am |
|
| sako243 Member Since: 08 Jul 2014 Location: Wales Posts: 1218
|
Given LEDs have a low forward voltage and you can probably stomach one individual LED taking out a chain of them the most efficient manner is to chain a series of them together to get you just under the nominal system voltage. There's almost no point in making it a buck-boost supply as it will be less efficient than a well tuned buck supply. They tend to be rather inefficient at the point they switch over from buck to boost operation.
The flickering that you see in slow motion on camera is a by product of switching the full supply of the LED, it's often more efficient since LEDs start so quickly compared to the human eye to simply turn them on and off faster than humans can see rather than smooth the supply. That I think is what you meant when you referred to high frequency switching. The flickering I was referring to, which is more noticeable is a by product of crappy designs and/or operating them out of spec. Basically the buck converter shuts down in an undervoltage situation (perhaps caused by additional loading), this causes the LED to go off and therefore reduce the load on the system supply. This means the battery can recover and the voltage returns to a higher level. This causes the converter to start up and the whole sequence starts over. You can mitigate this through a tweaked design, my suspicion is that, particularly the cheaper units, aren't designed or built to the same spec Up on order to save cost. Conversely higher end systems might have been designed to be much more efficient and therefore require tighter control of the system voltage. Ed 82 Hotspur Sandringham 6x6 95 Defender 110 300Tdi |
||
|
25th Dec 2016 10:26am |
|
| Cupboard Member Since: 21 Mar 2014 Location: Suffolk Posts: 2971 
|
It should be noted that a lot of non headlight LED lights aren't regulated and do just use currently limiting resistors, so have a much tighter voltage tolerance of something like 10-15V instead of 9-30V
|
||
|
25th Dec 2016 10:21pm |
|
| sako243 Member Since: 08 Jul 2014 Location: Wales Posts: 1218
|
That tolerance will only be to output the same brightness. The resistor will drop the necessary voltage across it, but a higher voltage will limit the absolute current less, so you run the risk of over driving the LED Ed
82 Hotspur Sandringham 6x6 95 Defender 110 300Tdi |
||
|
25th Dec 2016 11:07pm |
|
|
|
| All times are GMT |
< Previous Topic | Next Topic > |
Posting Rules
|
Site Copyright © 2006-2024 Futuranet Ltd & Martin Lewis
