I was just going to see if I could get some help on some resistance issues here. I know the formula v=ir, and I know exactly how much resistance I need for each individual LED. What I wanted to know, is does the resistor simply reduce the voltage, or does it do the voltage and current? Each LED can take 30ma. As far as circuits go, I always thought that devices that use electricity only take the current that they want. You plug in a 12volt lightbulb and 140amps are not going to just flow through it. Do I need resistors to limit the current on these leds if I buy a voltage regulator that is an appropriate voltage? They want around 3.3v so that's really easy to get since many microcontrollers use that voltage.

Thanks in advance,
Steven

 

The problem with that idea is that when an LED is forward biased it's resistance essentially drops to zero. The voltage rating tells you what it takes to turn the LED on, but once it's on it will take all the current it can get until it burns up (which is why you need a current limiter.) The resistor is there to limit the current, not to drop the voltage.

 

If wired in seris the resister will drop both the voltage to the led and the current in the entire circuit. Because the led is a much lower resistance compared to the resister you will loose most of your voltage at the resister.

 

so there is no real easy way to use one resistor with multiple leds, without the fear of one of the leds burning out and causing a domino effect?

 

LED resistance claculator

 

thanks for the link... but it still seems that either series OR parallel, you still run the risk of one led going out at some time, and causing the current to jump and destroy the others... am I right... Assuming you are using only 1 big resistor the for whole setup...

 

Sure it's a risk, but LEDs just don't burn out.

 

Maybe I should just run like 3 in series, and then run the sets of 3 in parallel... At least I wouldn't need hundreds of resistors....

Also does anyone know this... are the front turn signals on our cars basically two positives (the prongs), and then one negative, (the shaft of the bulb)? I need to know because I am going to re-wire them with my custom LED setup (as you can tell)....

This makes a difference if I should put the resistors on the anode or the cathode of the leds...

Thanks,
Steven

 

The front turn signals are full-time grounded, with seperate power sources for parking lights and turn signals. Brown wire is parking, green/blue wire is turn sig, and black is ground. You may need to go to a solid-state (Tridon heavy-duty) flasher module because the LEDs won't pull enough current to operate a regular flasher.


When I made my taillights, I had the same concern as you. I didn't want to have to use over 1000 resistors (since I used over 1000 LEDs) but I also didn't want one bad LED to burn out half the project. I ended up running parallel circuits of 6 LEDs and one resistor. That way if an LED goes it will at most take 5 others with it, and most of the light will still work (important for brake lights.) I also only had to have ~175 resistors. As it turned out, I had 3 bad LEDs right out of the box (2 open and one shorted,) and the shorted one did pull too much current through it's circuit which strained the other 5 in that series making them dimmer than the rest of the light...ended up replacing the whole set of 6.

 

Thanks for the advice... I just wish there was something like a voltage regulator that knew how much current I needed, and could adjust if an led went out... but I guess that's impossible.

Do you have any pics of your led install? I know how easy it is to use 1000 leds, I will probably end up doing all that... for now I'm doing the most important ones... the front turns, and the rear 3rd stop light...

I used almost 150 for one front turn signal...

Also, I won't have a problem with the blinking, because I have all led turn signals already, just those crappy ones from superbrightleds.com... the 12volt direct-fit (WORTHLESS!!) at least the warned me on the site before-hand... hey, they are half the price as autozone... so anyways I have a simulated load that's just spliced in on the wires before and after the turn signals...

Does anyone know how this will affect what resistors I will need for the project, or is this completely independent? Seems like it might lower the current a lot, dunno?

In the end, I will not worry at all about the blinker mechanism, because I am waiting to program and install my pic microcontroller, to run all of these features, (plus a few extra that I can easily program in )

 

I'm not sure how the sim. load will affect the voltage/current...it depends on the load itself. If you're going to a solid state controller in the near future it doesn't really matter anyway.

The only pics I have of my taillights are the "finished" pics. They're on my cardomain if you want to see. See page 3 of "my other page" in the sig.

 

right, I've seen that before, really kick *** man! The only thing I wondered was what are those chrome things? just the attaching bolts?

I'm putting these behind the stock lenses so that they will be diffused... hell I might even buy a sheet of diffusing plastic... I wish I could build a solid wall of ultra-bright leds...

 

oh yeah, so what did you do to simulate the load for your blinkers, or did you just figure out the load, and replace the relay up front?

 

The silver is the bolts that hold the blackouts on....I decided not to use the velcro with the second set since the first ones mysteriously disappeared one day.

For the blinkers, I just used a Tridon heavy-duty flasher. It'll run on lower current flow and maintain it's flash rate which makes it perfect for LEDs.

 

cool cool cool... well today I tried my blinker panel (beta 1.0) just by n riggin the wires where the orignal bulbs go up front... and they seemed great... the only problem I ran across was that I rated the resistors for the 13 or so volts that the Alternator puts out, when I started my car they were really bright, but no so impressive when it was just running off of battery... I might have to buy a 5 volt regulator after all, then buy resistors to go from 5 to 3.3...

what did you do to prevent this?

 

You should rate the resistors for about 14.5 volts, and figure about 20 mA as a 'safe' limit for the LED.
I use an 82<font face="Symbol">W</font> resistor per white LED (3.4 volt) with a 5-volt regulator, and a 150<font face="Symbol">W</font> for red.

 

hrm these are rated for 30ma... interesting...

I looked into buying some voltage regulators from radioshack today... but I don't understand exactly which resistors to use on the regulator. This is an adjustable voltage reguator, you use resistors to divide the voltage.

You have 3 legs on the regulator, one input, one output, and one voltage adjust. You place a resistor between the output and the adjust, as well as one connected to the ground, and the formula looks like this. voltage out = 1.25(1+ R2/R1) This means that I coud use any sizes of resistors, as long as the output voltage came out correctly.

So did you just use the v=ir formula to find out that you needed those resistors for the 5 volt source for your 3.4 voltage load? 5-3.4 = 2.6/.020 = 130ohm because with the regulator the 14.5 doesn't even matter anymore, right?

 

The 30 mA is generally the maximum current rating for the LED.
That's the formula I used - same as the calculator. BTW, 5 - 3.4 = 1.6 And always use the next largest value available. The 14.5 doesn't matter; I use LM7805's for regulators as they are set for 5 volts, and they have no problems operating from 18 volts. The ones you're looking at in Radio Shack are probably LM317T models.
You can pick up regulators on eBay for real cheap - 50 of the 7805's for like $10 from some people.

 

ah... well one thing I was wondering, is if I get the variable regulators, and set them for the exact voltage for the leds... what resistor should I use to limit the current on the ground leg of the led? Hopefully it won't lower the voltage at all.

And, yes, the regulator is the lm317.. or maybe the lm117 actually.. they are 2 dollars a piece... but that's not so bad since I need only 3 for now...

 

Dood, learn Ohm's law.

A LED is a diode and is non-linear, you can't look at it like a resistor. when it's forward biased, it pretty much has a fixed voltage drop across it over a wide range of current. You need to know what that forward voltage drop is, and what current you want to run them at. then just subtract the forward voltage from the supply voltage, that will give you the voltage that the resistor has to drop, then divide that value by the intended forward current, and that gives you the resistor value in ohms. and since you now know the voltage across the resistor and the current through it, you can calculate how much power it will burn up as waste heat. Get a resistor rated for TWICE the expected dissipation.

 

Here's a schematic I found on another forum that will enable you to get around 3.0 volts from an LM317T. The fellow claims it should run up to 50 LED's (in parallel) with no problems. No other resistors would be needed for white LED's.

 

tremo, the only problem with ohms law on that one... is if I get the variable regulator and set it to 3.3 volts... that makes the source 3.3, and the voltage I want 3.3... making the formula 3.3-3.3 = 0 / 0.030... which is still 0... unless I am doing something wrong with the math or being a complete idiot...

and pete... so he didn't even use any current limiting resistors on the leds themselves?

 

You do not want to drive a diode from a voltage source, you want to drive it from a current source.

Let the voltage across the diode go to whatever value it wants, you need to set the current through the diode to a fixed or controllable level.

 

I see.... is there some way to measure amperage through a dc circuit? that would be great, but the only way I know is to use ohms law, which involves the voltage and current, and I would assume that if you do the math right, and your voltage reads correctly, that you get the right current draw?

but so, if I had a 3.3 regulator, could I put a current limiting resistor on the negative terminal of the led, so that it won't lower the voltage? I was under the assumption that if you put the resistor on the postive leg, you lower voltage and current, but the negative terminal only lowers current because the power already flowed through the actual led...

 

First off you have to understand that the exact voltage drop across a diode is not a fixed exact constant. It will vary a little based on the current through the diode *and* the temperature of the diode.

Different color LEDs will have different forward voltage drops, and they will in turn be different than a standard silicon rectifier diode.

You will need to determine the approximate nominal forward voltage drop of the diodes you intend to use, it will be close enough. Then you'll need a voltage supply that's a volt or two higher than the diode drop. The voltage that will be dropped across the current setting resistor is equal to the supply voltage minus the diode voltage. The value of the resistor, in ohms, is calculated by taking the voltage drop across the resistor divided by the current you want to run through the diode. The power dissipated by the resistor is the voltage dropped across it (in volts) multiplied by the current through it (in amps).

Select a resistor that's rated at twice your calculated dissipation. Don't run the diodes at their max rated current, run them at 2/3 to 3/4 of max.

If you use a linear regulator IC to develop the supply voltage, don't forget that it will be dissipating heat as well, which will be the voltage dropped it multiplied by the current through it. You may need to heatsink the device or it may burn up.

Regarding putting a resistor in series with the LED, it doesn't matter if you put it on the positive or negative lead. A series circuit is a series circuit. Same current flows, the electrons don't give a crap. Physics.

The exact voltage output from a regulator IC will similarly vary a bit with changes in temperature, source voltage (the car's battery/alternator) and load current.

When I refer to temperature, I mean the temperature of the device itself, not how hot it is outside that day.

Now maybe you get an idea of all the things electronics engineers have to consider when designing a circuit!!!!

If you look at application notes from some of the regulator devices, I think you may see some circuits where the things are used as constant current sources.

Electronics is an interesting profession.

 

too... many... things... to... consider...

must...buy...storebought... ... something...

 

Sorry man, I didn't mean to scare you off, I was just trying to keep you from making a mistake and ending up with smoked components.

Like I said, diodes are non-linear devices. Because of that, they need to be driven by a current source, not a voltage Source. You use a resistor to convert the voltage source into a current source, OR you build a application specific current source.

With a current source drive, the current source will adjust it's output voltage to whatever is necessary in order to get a set current output. That then takes care of differences in tolerance, temperature, supply voltage, etc.....

Electronics is fun.

 

well now I've decided that I don't like the way the leds look at a lower light level... (some are not as bright)... so I guess I will just make the parking lights some small amount of leds at full brightness, and same for the tails... (like the cadillac does) I will have to think of some tasteful design..

right now I have the led's running in series, in sets of 4, and all the sets are in parallel. The annoying part about that, is when you add up the voltage for each set, its 13.2, but then you still need like 2 volts extra for compensation for the resistors, right? maybe I should brake up the board into series of 2, I would do 3, but you can only fit 4 vertical leds, and 20 horizontal leds in the front turn signal enclosure.

 

Dood, no offense, but all this LED talk is starting to sound R.I.C.E.

 

that's kinda what I was trying to avoid... but the only way to do that would be to have the lower voltage regulator, or separate set or resistors to lower the voltage and current to the leds for the tail lights, but the problem with that is some of the leds don't look the same brightness when you do that...

plus, does the cadillac look like R.I.C.E. to you? maybe it's just me, but I think it looks pretty slick...

who knows, if it looks like crap, a lesson learned is a lesson learned... I'll post pics so you guys know how hard it is to make it look good...

mabe I will just light every OTHER led in the grid, so that you can't tell... since the taillight lense will diffuse the light quite a bit. Guess I will need to do some testing here and there.