PBO, LED lights project, July. Electronic engineers please!

M

MASH

N/A
Joined
8 Mar 2004
Messages
2,188
Visit site
This great article shows how to make LED lights at huge savings in battery power.

Can anyone provide formulae for calculating value of resistor required in such circuits cos I can't figure it out from the numbers given by using V+IR.

How critical is the current draw?

Does it affect the brightness?

Can anyone explain the tech spec of an LED as defined by IFmx, VFtyp, VFmx and VRmx, or point the way to a site that does, cos I can't find one!
 
Well, the LED you use should have a prescribed wattage on its spec. sheet, and power = amps x volts, so divide by 12 to find the current you need for optimum brightness. Then use Ohm's law (Volts=Amps x Ohms) to work out the resistance:
e.g. a 1W LED will need 83mA, which means you fit a 150 Ohm resistor. Make sure the resistor is rated to 1W as well.

Without looking them up, I'd say IFmx is maximum forward (i.e. normal) current draw, VFtyp is typical forward voltage drop, VFmx is max. forward voltage drop and VRmx is maximum reverse voltage drop.
 
First the meanings of the variables. IF is the Forward Current and it has a specified maximum value IFmax. VF is the Forward Voltage drop and will have a typical or maximum value. VRmax is the maximum allowable reverse voltage, which is the voltage that, if applied the wrong way, will destroy the LED. It will usually be much higher than any voltage available on-board. Now VF is usually in the region of 1 - 2 volts, so the current needs to be controlled to a safe level, which is where the resistor comes in.

To calculate the size of the series resistor, you need to know your supply voltage, say 12 Volts, and the required current, IF, usually 20 mA. For an accurate answer, you also need to know the VFtyp which is the voltage drop when the LED is conduction the current. The resistor is in series with the LED to limit the current, so not bothering with a value for VF gives you a safe result where the LED current is a bit low, and the light is a bit below maximum. It also gives you a safety margin for the battery voltage being high when it is on charge.

Using the values above, for a single LED, the safe value of the resistance is V/IF = 12 / .02 = 600 ohms. If the value of VF is 1.5 volts, the actual voltage drop across the resistor has to be 10.5 Volts and the resistance should be 10.5 / .02 = 525 ohms. The final twist is that resistors are not available in all values and the nearest available value ABOVE the calculated value should be used. That would be either 680 or 560 ohms and I would use the 680 ohms one.

The power handling of the resistor also needs to be considered. A 680 ohm resistor carrying a current of 20 mA will have a theoretical voltage drop of
680 x .02 = 13.6 Volts (yes, I know that is impossible but bear with me.)
The power dissipation required is 13.6 x .02 = 0.26 W and you will probably get away with using a 1/4 watt resistor, but a 1/2 watt would be safer.

This has been a quick and dirty way of arriving at a specification for a 1/2 watt, 680 ohm resistor. Of course if a series network of LEDs are user, you do need to know the total VF and work using the correct maths.
 
Forgive me, as I know not what is watt!

If the boat has a 12v supply! Why can't I buy a 12v LED from the Maplin Catalogue, choosing the wattage and be done with it?
 
I suggest that if you are asking that sort of question then you are likely to find the process frustrating - unless one of your objectives is to learn more about semi-conductors in the process.

If you want to fit LED lights to your boat then I suggest you have a look at ultra leds and look at their ready-made 12V bulbs - you might even find one to fit your existing light fittings.

This whole area is much more complex than appears at first sight - partly because LEDs are semi-conductors, not resistive devices, so Ohms law does not apply and partly because the voltage input on a boat can vary between about 11.8V and 14.8V

However the basic answer to your question is that it is the current through the LED that is critical (not the voltage).
 
The sort of LEDs used for lighting all drop about 1.5-2.5V, but they are current devices, not voltage devices. If you want the easy life go to the Ultraled site I posted above - they will have what you are looking for,
 
The LEDs you describe have an integral resistor and their quoted dissipation is 105 Watts. The brightness of LEDs is usually said to be (in Watts) about ten times that of an incandescent bulb, so to replace a 10W bicolour light you'll need ten of these.

To be honest, unless it is something you want to do for fun, it would be hardly more expensive to buy something from these folk or similar, and fit it to your housing.
 
I have not seen the article but using resistors is not the way to go.

LEDs are normally driven off switch mode power supplies. Tiny ones. The reason is that they suffer from the same problems as a bulb. The power on a boat varies between 11v and 14.8v normally. So when the resistor is selected the maximum voltage must be used. This means that at normal voltages the LEDs will be a lot dimmer. If you only select 12v then the forward current would be too high when the engine was running.

The affect is worse if you series LEDs together to eliminate the huge (relative) power losses in the resistors. The derating using passive resistors is even worse. An active current source is required.

There are many driver chips been designed and used in phones etc. for powering LEDs correctly. These will have some power loss, but the chances of blowing all the LED are a lot less.

After derating and power loss through resisters LEDs and bulbs are not that far apart. The same power supply techniques can be applied to bulbs so that they do not have to be designed to survive large voltage fluctuations and hence be more efficient.

Using resistors the loss of power in the resistor is about 5 times that used by the LED. How much more efficient are the LEDs they are proposing?
 
Guys, I think you have made your point! Reading your explanations on how to work things out is very interesting I applaude your ability and wish I did not have this mental block when trying to follow maths and transformation of formulea! What I wanted to indicate is! I bet 80% of the fraternity following the Forum struggle to cope with that type of detail in working out calculations and would settle for an LED in a packet off the shelve at B&Q if they sold them!
 
As so often happens we have a language problem. The LEDs come as a tiny bulb 3/16 or1/4 inch diameter with 2 wires sticking out the back. Now people are selling a whole light fitting with LEDs and current drivers and people are calling them LEDs.( I would prefer to have them called LED powered lights) To add to the confusion you can buy a 12v LED for idicating power is on, on a switch board.
Yes LED arrays can be far more efficient and cope with varying voltage supply using a current controlling silicon chip (integrated circuit)
A resistor and essentially an incandescent light bulb has a fixed resistance so that more voltage applied to it gives a linear increase
in current and linear increase in watts dissipated which is light o/p for a lamp. So it is conveneient to rate the lamp in wattage at a given supply voltage. It could just as easily be described by it s current at the rated voltage.Light bulbs have a way of showing low resistance when the filament is cold wich increases as it gets hot so the resistance is best calculated or described when it is hot ie from voltage and current but at the rated voltage.
A resistor on the other hand exhibits the same resistance up until it melts from overload. So a resistor is described by its actual resistance and its max power dissipation ability but never by its voltage. So for a series resistor for a LED device it is best to use a higher wattage resistor smply because it is bigger and more robust. Being physically bigger it will not get to such a high temperature while dissipating a given power as a smaller resistor.

A LED (the basic device) has a queer relationship of votlage and current. If you connect a variable voltage across an LED no current flows until the forward voltage is reached typ. 1.5 volts. At that point a large current will flow with a tiny increase in voltage such that it will be destyroyed very quickly. At the specified current the LED will glow nicely but is dim at much less current and gone at any higher current. A resistor in series automatially limits the current by effectively dropping the voltage available to the LED. ie the more current the LED passes the lower the voltage so it comes to a static current just what we want.
Ohms law is applied to the series resistor value relating to the voltage dropped across the resistor (supply less the forward v of the LED) and the current required in the series circuit.(LED current)

Only in the case of high powered LED (devices) are they described rated in watts and then that doesn't relate to the power used frrom our ships battery but rather the heating and hopefully the light output. Of course LED lamp replacements or light fittings are rated in watts from the battery.
So it is all very complex not aided by name confusion. I suggest that experimenting AFAIK with current control chips is a bit tricky for the boaty amateur electrician. But using resitors is easy. ..olewill
 
A watt is the unit of power disipation or energy conversion. It is calculated in D.C. electrical circuits as Voltage drop across the device multiplied by Current through it. If this is hard to understand, as has been suggested elsewhere in the thread, I suggest that designing your own electronic circuits is not a good idea. I say this as someone wiith a degree in physics who did not really understand electricity or semiconductors until I had destroyed a few. Yes, they do go off with a bang if you mistreat them!

Why not buy a 12 volt version? Good question. Again, it has been pointed out that these would be better named as LED "devices" as there are not only LEDs in them. You need to look at the specification sheet however to make sure that they can withstand the overvoltage that can be found in boat electrical systems. Again, if you don't understand how to read the spec, then it is really not a good idea to experiment unless you do it under controlled conditions.

I use LEDs as indicator lights, but do not just install them on the boat. I build a test circuit in the lab and measure the current and voltage drop at various supply voltages up to the maximum that I expect they will be exposed to. If you get it wrong, it will not work, and could even start a fire.
 
Re:A passing thought

Reading this and other threads about electrickery there is enough disinformation published in this forum to electrocute several people and start innumerable fires.

If advice on this forum is followed and the result is a really good fire destroying a boat and perhaps several of its neighbours - is there a question of liability?

Can I sue PBO or do people who give well meaning but erroneous advice have some liability.

More important how does one tell the difference

Or do we just trust to luck and follow whoever sounds most convincing, perhaps who deals with dissent in the most aggressive and abusive manner.

Me?

I'm sticking to bulbs.
 
It all sounds very complex and I'm wondering if it's worth the bother.

What are the economics of using a single stabilised 12-volt supply "box" to feed all the lighting circuits so that when filament bulbs fail I can replace them with LED "bulbs" and know that I'll get the best out of them safely? Does such a box exist or would one have to make it? I can imagine using an inverter and then a 240AC to 12DC PSU (giving 'approximately' 12vDC to controlled 12vDC) but that seems a bit of a sledge hammer.

Given that that works, should I then have two so that I'm not left in the dark (sic) if the "box" fails?
 
You have only brushed the surface

Once you understand exactly how LEDs work you can drive them above their rated max current using a few discrete components. For example: You should be able to design a masthead red/green/white to fit into an Aquasignal outer with 2 mile visibility (4300 mCd) which runs at less than 0.6W /forums/images/graemlins/smile.gif (The total component cost from RS Components and a couple of other suppliers will not set you back more than £17 inc VAT for the whole caboodle.)
 
Ths is Yaker, the bloke again who is struggling with LED circuits. You will have read my input for 12V LED simplicity. I bow to the guys who tell me voltage in a running engine can blow things. As a parting last comment I wish to put forward another well used application of LEDs in clusters. I would like to know how the HGV trailer manufacturers overcome the voltage problem?
In the HGV trailer world, quite a few manufacturers have beeen producing trailers over the last few years with front running lights incoporating LED light clusters, which meet goverment 'Lighting Regulations,' instead of the bayonet bulb fitting. HGVs towing trailers are regulary slave started when the batteries are run down, thus there must be some pretty low and high charging peaks be they volts or amps! I have yet to inspect one for its annual MOT test with a failed LED cluster. Can anyone tell me how manufacturers have got around this? I will make one comment. I bet the clusters are quite expensive, and will never be able to compete with the Maplins and RJ componets parts costing probably a pound! Thanks again.
 
Re: You have only brushed the surface

Hi,

First, I build and sell lights, so take this with however large grain of salt you choose to.

You can make a credible LED light using a dropping resistor, so long as you are willing to accept that you will either not get the desired intensity at low battery levels or you will decrease the expected life of the product. Many of the lights that are flogged off (at stupid prices) to the automotive market do just this.

To compute the value of the resistor that is optimum, you should really measure the actual forward voltage drop of each LED, esp. if you intend to use a high output LED. As the drop in these devices have a typical of 3-4 volts (as opposed to 1.8-2.2 for the non-high output LEDs) , this is outside the range of the diode check function in most multimeters. The actual measuring is because in a single lot of 10k devices, the distribution of the drop does not follow a normal (z or t) distribution.

If you dont want to do that, and really as a rule of thumb, take
(14.4-3.5)/0.01 for a single LED or (14.4-10.5)/.01 (for a series string of LEDs) for the value of the resistor. The 14.4 is for max normal voltage seen on your system, you just have to ensure that the light is turned off before an equaliation charge. The .01 is because the high intensity LEDs use a phosphor reflector to gain the white light. Subjecting the phosphor to the max current will cause it to degenerate prematurely. If you dont care, use .02 as a rule of thumb.

The light output will, of course, vary with input voltage. Built as stated, you will get about 1k-2k hours from the light.

I find it unlikely to have an explosion or fire if the device is connected to a properly sized breaker or fuse.

As pointed out above in this thread, there is a lot more to this than meets the eye (excuse the pun) to produce a viable product for the long haul. It is also unlikely that you can buy high quality LED devices (we get a zillion emails a day from China for the cheap LEDs, tnx but no tnx) in small quantities for less than you can buy a completed light from a reputable mfg.

But it is a fun project (with a resistor) and folks tell me that you have a lot of rainy days in the UK to do it!

Best regards,
Michael
 
[ QUOTE ]
I would like to know how the HGV trailer manufacturers overcome the voltage problem?


[/ QUOTE ]

Do you have one of those little bricks, generally black, that convert your 12v on the boat to the voltage required by you computer.

Or one of those nifty lighter socket things that charge your phone. They convert anything from 10v to 16v to exactly what your phone needs.

Well imagine a really, really, small one that is made with the sole purpose of driving a set of LEDs. Now hide it in the bayonet fitting and a low power tail LED light is easy.

As for cost, the packaging, distributors, transport, middle men, reps, fitting agents, import duty, marketing, and general supply chain groupies take a lot more money that the few cents needed for the components.
 
It only costs a few pence to create the regulated voltage you need to get constant light from the LEDs - one approach is a zener diode, which cost about 8p, A better approach would be a 3 terminal regulator for about 30p (and those are Maplin prices - wholesale is probably a fraction of that)
 
You must log in or register to reply here.

Share this page

Top