/diy/ - Do It Yourself

This is a very dangerous design. Rule #1 when dealing with LEDs is: Use a series resistor.

You could do, for example, a chain of 5 diodes, each with 120 ohms in series. gives you about 40mA per LED. Then add more chains in parallel, each one just like the first, until you get enough illumination.

keep this in mind also: unless you're wearing a night vision device, you're not going to anything with the IR lighting

>>368053
What will happen without the resistor? Will they just burn out?

Do I just need the resistor to reduce the current to the other components?

>>368055
*not going to see anything

>>368053
So something like this? But wouldn't the reduced current reduce the amount of light each LED emits? I'm aiming for maximum brightness.

> But wouldn't the reduced current reduce the amount of light each LED emits?

yes, that's the whole point: to keep current to a safe level so your LEDs will last a few years rather than dying after 5 seconds of maximum brightness. Rule #1 of engineering is Be Conservative in your Design. For obvious reasons, if you have a 1000 pound weight, you design a platform for it that can hold 1500, not 1001.

>>368078

you need the series resistor, like you show, one per series string of LEDs. you're close OP....

the math is very simple, but it requires knowing about the inherent volage drop across each LED, Ohm's Law, etc. it's not hard, if you're going to do this sort of thing, it's worth the effort to figure it out.

in the mean time, you can use online LED calculators to play 'what if' with values.

http://ledcalculator.net/

you cannot exceed the maximum rating of each LED. Wel you can... but they'll die, sooner or later (or instantly, with no resistor). it's best to run below ABSOLUTE MAXIMUM RATING (if you have a data sheet) or 30mA for most ordinary diodes. not sure where you got 60 mA, but if the data sheet says that's a good operating current go with it. Note that MAXIMUM means shorter life...

>>368442
Just on my math, using the second diagram I posted above, wouldn't a 75ohm resistor give me 60mA of current across each branch of the parallel circuit, meaning that each LED gets their maximum rated current?

I think I'll go for 150ohm resistors to reduce the current a bit to add longevity to each LED.

If I wire two 150ohm resistors in parallel, do they act as a 75ohm resistor or am I horribly mistaken?

>>368506
using 6 LED's in series:
12V - (6*1.5v) = 3v = voltage drop across resister

60mA: never run anything at its max rated value. 80% is a good idea. use 60mA*80%= 48mA

resister:
3V
48mA
R=V/A=3/.048=62 ohms

so, 6 LED's in series with a 62 ohm resistor draws 48mA


How many can you run?
power supply rated 1.2 Amps
80% of this is 0.96A

.96A/.048A = 20 (series circuits) for a grand total of 120 LED's possible

>>368506
>If I wire two 150ohm resistors in parallel, do they act as a 75ohm resistor
yes
3 in parallel is 50 ohms.ect...

>>368521
also: use 68 ohms, a standard value.

>>368521
Thanks, learning loads over here.

>>368521
So let me get this straight, if I put two 160ohm resistors in parallel across each series branch they effectively act as a 75ohm resistor.

Using those, each branch will draw exactly 40mA of current (A=V/R=3/75=0.04), so each LED will receive 40mA of current. And this will be fairly bright for the LED and is below capacity so they will last for ages.

I'm using 150ohm resistors because I got given 100 of them so it's convenient they work out roughly right (assuming my math is correct).

Thanks for all your help guys, now I need to buy way more LEDs because it seems my power supply can handle a massive number of them. Learned a lot here today.

>>368542
*150ohm, not 160ohm

My project:

I'm building a home security system to watch my yard at night. We've had some prowlers come into our yard at night (as evidenced by random boot prints, unidentifiable grimy hand prints on walls that none of us made, strange sounds, makeshift beds by our wood pile, etc). Rather than fork out $500 for a decent night-time security system, I thought I'd build my own.

I've already dismembered about 5 USB webcams and removed their IR filters. The best one is a 720p I got off a friend who thought it was broken for $5. It's quite sensitive to infrared light so it should be perfect. The other cameras could come in handy if I decide to mount multiple cameras. The camera is hooked up to a server running iSpy so it records whenever it detects motion at night.

It's passable in the dark, but I want to build an spotlight that can illuminate a reasonable portion of the yard to improve the motion detection capabilities and maybe improve the image quality enough to make out identifying features of an intruder. I've already been using a standard LED spotlight which is very small, not very powerful and very obvious, so to add a bit of secrecy to it I want to use infrared LEDs, but I'm trying to wire it on the cheap. You guys have been a huge help so far.

The reason you need a resister is that components aren't precision. OP's pic works fine in theory but consider:

If the LED's only drop 1.4 volts, then that leaves .8V unaccounted for.

the current would be .8V/0 ohms = infinite current
(the voltage across an LED actually varies slightly with current so the voltage would actually reach 1.5V but current would be unpredictable)

>>368555
Thanks, I couldn't work out why exactly I needed them but that makes a whole heap of sense.

>>368542
you got it

Are are aware infrarred light is invisible to the naked eye, OP?

In case you want to illuminate it for a security camera it's alright, they can see infrarred. But if you want to have light that's visible to you, you'll need regular white LEDs (or whatever other color you want)

>>368562
That's the point, hopefully intruders won't notice the system until I have their face saved on my hard drive.

I'm using 850nm IR LEDs which do give off a faintly visible pink glow, but it isn't bright enough to be immediately noticeable.

Another noob question:

Lets say I have 6 LEDs wired in series with a 9v battery. The calculators tell me I should throw in a 1ohm resistor, obviously to account for variance but as the voltage over the resistor would theoretically be 0 doesn't the equation above mean there'd be no current drawn? (0V/1R=0A)

I'm assuming because the resistor is just there to cover me in the event the LEDs aren't precise, the current over this resistor doesn't matter much at all, but don't all components end up with the same current going through them?

6 LEDs running from a 9V battery is the same situation as 8 running from a 12V battery: you have no safety margin. you can do it, but it's dangerous. Use 5 diodes and a resistor, or 4 diodes and a larger resistor. the bigger the resistor, the more leeway you have as the battery changes voltage. they usually start at 9.6V and drop to about 8.0 before becoming useless. In your example with 6 LEDs, an 8V battery will deliver almost no current to the LEDs, whereas with 4 or 5 LEDs, they will still light at around 80% of max luminosity.

>>368901
You need more resistance.
The calculators are throwing in a resister cause it's needed, but 1 ohm is not enough.

The voltage and current relationship in a diode is not linear.

This chart shows that a slight increase in voltage creates a large increase in current when in the forward "conductive" range. A resister puts the current curve more towards a linear one.

So, say your voltage goes up 10%. (not uncommon)
Resister - current goes up 10%
Diode - current goes up a lot more than 10%

>>368914
>>368917
Thanks, I've got it now.

>>368914
Just to double check, 5LEDs connected in series to a 9V battery with the equivalent of a 30ohm resistor in series with them will receive about 50mA of current. If the battery starts at 9.6V, this will potentially fry the LEDs because they will receive 70mA at that voltage.

So a 40ohm resistor would be good because at 9.6v it will deliver 52.5mA to the LEDs, at its lowest it will deliver 12.5mA and at 9V it will deliver 37.5, all of which are decent values for operating the LED.

>>369027

"By Jove, I think she's got it!"
- Prof Higgins, My Fair Lady

Why are your photos all fuckeyed, OP

>>369564
iPad

This has been a very educational thread for me also.

I'm more of a noob than OP here, would you guys recommend any books on the subject of amateur electronics. Maybe stuff that has simple project to get me started?

U should run the LEDs in parallel or their internal resistances will make the last LED dim and also if one goes for whatever reason they all go.

my 2cents

>>369710

erm, exactly backwards... run in SERIES, as shown, and the CURRENT will be the same in all LEDs, even if there are small VOLTAGE drop differences on each LED.

>>369710
These are in parallel already. Having one to a branch seems like overkill, would make the wiring much uglier on a board and would waste all of my resistors.

This way, if 1 blows 5 more of them go out, but the other 100 stay on.

>>369890

the way you have it arranged now is optimum; LEDs in series with a single resistor, then multiple series strings in parallel.

it's also the most power-efficient. it scales nicely, eg. you can simply parallel more series-strings as you need them. the total current consumption must be less than your power supply can deliver of course.