finally here... my DIY LEDS have been tested and works !!!!

[genocidex]

Well here it finally is my small scale LEDs. I decided to run this by a constant current source I made myself and I designed it myself (with the aid of my teacher and text book to make sure I wasn’t making wrong assumptions). I would like to point out that my camera on my phone stink so sorry. Any ways…
All these leds are 1w leds with nothing more than the star board for heat sinks. They seem to get warm to the touch after a half an hour so before I actually install them in my tanks I will be looking for a better heat sink.
Heres the basic form of the circuit:
IMAG0144_zpsb18ad7ec.jpg

The parts list is as followed:
Q1 : IRF640 MOSFET (should be fine for almost any leds application)
Q2: 2N5088 NPN tranistor
R1: 22K ohm 1/2W resistor (if you are using more than 9V source I would step that up a little)
R2: .47 ohm 2W + 4.7 ohm 1W resistors (to make it a better fit for me.) in parallel. *
Z1: 1N4733 zener diode ( regulates the voltage at 5.1V).
*R2 in this set up is a current setting resistor. To make a value specific for your application you need a formula. R = .58/I. I is the current that you want your leds to be driven. I needed about .43 ohms that’s why I used the 2 resistors. The reason they are such high wattage is because all the current in the circuit is flowing threw them P= I*I*R. that will give you the minimum power rating for that resistor. But like filters its always best to double it. my circuit was running 1.4A so P= 1.4*1.4*.43 = .85W -> 2W resistor. My smaller resistor needed to be 2w but my larger one could be easily half a watt if I wanted it.
How this works? Well to simplify it a bit, Q1 regulates the current (using R2 to set it) by droping voltage. Q2 is there to monitor Q1. Since Q1 might change the currents with temperature (causing a bad failure), Q2 was installed to stop that. When Q1 gets hot it lets more current threw, when Q2 sees this it slows down Q1. Pritty neat huh?
R1 isnt too special. It is just there to help Z1 by dropping voltage ( in this case 3.9V).
Z1, when put into this circuit the way it is, will always keep a max voltage there of 5.1V. its just the way Z1 works.

Now how to dim the circuit?

we need to take a special property of LEDS into account for this. they are either on, or off. its bad but good at the same time. since leds dont need warm up times like florescent lights. so how in the world can i dim them? well LEDS have a response time of about 8ms some of the fastest light response. if we were to pulse DC into the ciruit the LEDS will turn on super quick and off again before they are fully on. a good frequency to do this in is about 500 Hz ( 2ms time) and vary the high and low. the longer the high cycle the brighter the LEDS. but we can use a micro controller to do the dirty work for us. Arduino makes it super easy by giving a command that is meant for just this. AWESOME! so heres how it looks:

IMAG01441_zps36916c94.jpg

the micro controller (Arduino uno R3 in my case) is being used as the voltage to switch the mosfet on and off. not much else has changed. but theres a problem i dont like, you cant keep the lights on without having the micro controller on and having a code to keep the output high. so i made a change here:
IMAG0145_zps6ee9e512.jpg

when the micro controller is off *that is VERY important* you can switch SW1 on and the leds will say on till the switch is pulled off. if the micro controller is on when you do this you will just cause a short circuit and the leds will stay off. i dont recommend shorting the circuit on purpose, but i didnt blow any thing up....

[genocidex]

what is you want to dim your leds and not spend 35$ on a micro controller? well you can always build a pulse wave modulator out of a 555 timer IC. and do the same thing, just not programmable. you will need to turn a knob to make it dim. a small trade off i guess. below is a basic 555 timer circuit and my alterations for better fitting (parts list will follow at the end):
IMAG01461_zpsdd1178e1.jpg
thats the basic one, but there are improvements. the low voltage will always be as long or longer duration than the high voltage. because
F = 1/ ( LN(2) * R1 * 2R2 *C1) thats the forumla to find frequency. if you want to know why ill tell you later lol.

IMAG0146_zps84683662.jpg
see the new diode? well that makes you be able to have a duty cycle of less than 50%. what i mean is the high voltage can stay on longer than the low.
the new diode take out 1 R2 from the equation by shoriting it one way. so the new frequency formula would be this
F = 1/(LN(2) * R1* R2 * C1). this is good...

heres the actual circuit with parts list:
IMAG0147_zps44b94f2b.jpg

that will give you close enough to 500Hz to work good.

here is the 2 circuits combined:
IMAG0148-1_zpscb17121b.jpg

[genocidex]

so, now that i bored you half to death with the theory behind everything here is the results.

leds off:
IMAG0141_zps61bde56e.jpg

on:
IMAG0142_zpsaf1d5cb6.jpg

my empty 29g with LEDS:
IMAG0150_zpsb821ca39.jpg

my empty 29 with 17W t8:


those last 2 dont give my leds justice, they really look great IMO

and lastly the actual circuit i built:
IMAG0143_zps5192d9dd.jpg

note the heat sink, that is on the mosfet. i have these badly configured for power consumption. since the mosfet is forced to disipate nearly 2V i decided a heat sink would be a good call. dosent get too warm so i think its doing a good job.


to come in this thread would be other pics when i add them to a bigger heat sink and actually install them in my tanks. also future builds would be:
150g
10g shrimp tank with 8/12 3W
29g not planted.
a new 55g (not sure plants or not)
and a 10g low light betta tank.....
ooohhhh


oh and a new note, i am tackling a DIY par meter, the results wont be perfect until i find a real par meter to calibrate it to, but they will be good enough, im guessing +-5 par... which isnt bad at all.... and with that i will give my par readings for the leds.....

hope you all enjoyed the read.
if any one wants help building a led system like that i can defiantly help by either giving you advice or selling you a regulator...

[Longshot]

That's really cool. Of coarse it's written in a foreign language but I don't need to understand it to appreciate the effort that you put into it. I can see that you are passionate about electronics, I commend you for your success. I wish I had an inkling about what was going on, it would be really fun to build this kind of stuff. Good Job.

[genocidex]

well heres my newest addition to my project. here you can see the leds getting brighter and an lcd screen that displays how bright the leds are in a percentage. obviously i only recorded a small portion.

[Longshot]

What will you be mounting all that into? They look good tho, I wish you much good luck with them!

[genocidex]

when i make the final project they will be mounted into a canopy ontop of the aquarium with the lcd screen mounted into the front. i plan on having my arduino monitor (and hopefully control) temp, control the lighting, and a few other things im trying out before i even mention lol