DIY LED driver for reef lighting

shall do that fishman.

stu, there is an internal pullup, i think it is 50k resistor.

I need the voltage to be within 0 to 1.25V.

when the collector is connected to the adj pin, it initially takes it to 1.25V when floating.. so i presume when the transistor is in common emmitter mode, it will operate within 0 to 1.25V based on the input to the base - which is PWM.
 
If you don't get anywhere with fishman's testing, I would suggest putting that transistor and base resistor on it's own circuit, with a voltage source, current limit resistor, and plain old normal LED such that you can verify it's behavior without "risking" the driver and HP LEDs. Putting it in a test circuit like that will verify that the Arduino is able to turn it on and off and control it with a PWM signal.

You need a base resistor to limit current, but you might want to play with values if the transistor isn't operating properly in an isolated circuit.
 
Thanks DWZM. Will use another PCB for the testing.

Actually there is nothing happening on the LEDS. They are driven at the same 100%. There no change in the Current through the LEDS - remains the same at 600ma.

But when tested the same with a Regular LED, it dims and brightens it. So there is nothing wrong with the arduino. Just the PWM Transistor Circuit is where i think the problem is.

I will first apply ZERO Volts to the Resistor to Check whether the LEDS Turn OFF.. If that doenst happen then i am in deeeeeep trouble!!!

Late night here..will post my results in 12 hours... need some sleep.
 
ZXLD1366 datasheet:

Abs. Max. values: "Adjust pin input voltage (VADJ) -0.3V to +6V"

"VADJ External control voltage range on ADJ pin for DC brightness control(d) 0.3(Min) 2.5 V(Max)"

So those two specs contradict each other.

And you need to get below 0.15 -0.2 V to turn off.

the datasheet does state that the pin should be at 1.25V & driving it to 2.5 results in 200% output current. So pulling it low should turn off the circuit.
What happens if you just short the ADJ pin to ground?

While you are driving the PWM signal put a scope on the ADJ pin and see what it is doing.

If you dont have a scope, put a DMM on it & check frequency and see if the DMM is even seeing the PWM signal.

Stu
 
Stu,
Grounded the ADJ pin to virtually give 0V. Still the LEDS were at 100% brightness. Grounding the ADJ pin is equivalent to supplying ZERO Volts right?

Dont have a scope, will use the Multimeter to check for the Frequency.
 
If you're grounding the ADJ pin and nothing changes, something's wrong with your driver.

This might sound silly, but can you check that when you ground it you're actually seeing 0v vs. some known ground on the circuit?
 
der, are you still in test mode on your design, or is it ready for 'production'? (if you were building a non-test fixture for your tank would you use this design as-is?)
 
I'd definitely consider the version that's up on google code right now to be "production ready." If I were to build an LED fixture today, I'd use it without hesitation. I tested the fully-loaded board (i.e. all three circuits running) at different voltage drops and PWM duty cycles with different mixes of Rebels and XR-Es for a solid week without a single problem.

To be fair, the version that's up there now is very slightly different than what I tested. I added a few more vias around the parts, and I fixed the missing ground on the output caps. But functionally it's pretty much identical or slightly "better" so I would definitely trust it.
 
Another Observation on the ZXLD:
Varied the Power supply +/-0.5V to get minimum heat.

I Got these results as follows:
1. 49V of SMPS power supply that are powering up 13 LEDs.. but as per the spec sheet this is too much!!! but at this there is heat, but minimal heat.. compared to 12 LEDS. these LEDs are driven at 3.62 to 3.74V with a current of 550mA. These are Edision LEDs are estimated to be driven at 4 to 4.3V and 700mA current.
In the same circuit i connected the same power of 49V to 12 LED string and found that the voltage drop across the LEDs were only 3.4V. But this IC got hot faster when compared to the IC that was driving the 13 LEDS. And these 12 LEDs were visibly less brighter when compared to that of the 13.

These are the components that i am using:

1. RSense 1W Resistor : 0.33 Ohms
2. Inductor : Resistor Type, 220uH. Looks similar to that of a resistor.
3. Capacitor : 10uF / 63V
4. Diode : 1A BAT49 : http://www.datasheetcatalog.org/datasheet/stmicroelectronics/3297.pdf

I got a reply from www.diodes.com, stating that i am using a higher duty cycle for the 13 LED String : 13LEDs X 3.67V = 47.71V @ 49V = 97.36% duty cycle whereas the 12 LED string : 12 X 3.4 = 40.8V @49V = 83% Duty cyle which is within limits!!! dont know how this will help though. But with 12 LEDs there is much heat generated....
 
perikaruppan,

Are both the 12 string and 13 string being driven at the same current? Please measure. Are they the same type LEDs. The average voltage drop across the LEDs (if they are the same and at the same current) should be the same. I don't understand why the 12 string has so much lower drop per LED. It makes sense to me that the chip (12 string) gets hotter it is dropping almost 8 volts as compared to 1.3 volts (for the 13 string). Sounds like maybe the circuits are regulating correctly.
 
all the LEDS are the same: Edision - Cool Whites. Didnt measure the current across the 12 String.. but the output was far less, as bot the 12 and 13 string LEDS were lit at the same time.

You mean to say the circuit is regulating correctly?
 
If both circuits had 550 ma then each LED should be the same brightness. Also if they are running at 550 ma the voltage drop across each LED should be about the same. From what you are saying (or I am understanding) the brightness is different and the voltage drops are different. There for I am questioning whether or not both string are at 550 ma.
 
TheFishMan65 said:
If both circuits had 550 ma then each LED should be the same brightness. Also if they are running at 550 ma the voltage drop across each LED should be about the same. From what you are saying (or I am understanding) the brightness is different and the voltage drops are different. There for I am questioning whether or not both string are at 550 ma.
Click to expand...

Agreed 100%. perikaruppan, I'd measure voltage and current on both sides of the driver (input and output) just to be sure you're not getting "strange" readings on the output side.
 
okie. Will do that.

Evil66 says that i am driving too much close to the total Vf, so he says to use 11 LEDs @ 3.7V (if that is the drop) and try running them at 48V.

Will that help? or the large voltage drop cause issues?

Also i am using a regular Inductor, that looks similar to that of a resistor. Is that a problem?
 
instead of breaking my head so much with so many tweaking issues, can might as well switch over to the CAT4101. The only problem is need to ship that to India!!!!
 
Is anyone up for a group buy on the boards and other parts? Or is there a thread with a group buy already started?
 
perikaruppan,

I think what you need to try is 2 strings of twelve and observe:
- are they the same brightness
- are the currents the same
- are the voltage drop the same
If everything is the same then everything is good. If not report back and we will try and figure out what is going on.
 
You must log in or register to reply here.

Similar threads

C
DIY Stand for Flatback Hexagon Tank
Replies
2
Views
782
Drummer98
Drummer98
T
Looking for a specific tank build - EZTube.com sent me here
Replies
6
Views
545
Reefing102
Reefing102
Emanuel92
What lighting for S. haddoni and E. quadricolor?
2
Replies
20
Views
1K
Hong Kong
Hong Kong
C
Scored a complete setup for $100
Replies
13
Views
1K
griss
griss
R
Red tufty algae - Asparagopsis - any expert advice for removal?
Replies
14
Views
2K
rallibon
R
Back
Top