DIY LED driver for reef lighting

and try it at 48V?

This is what i observed when i increase the Input V, as i keep increasing the V, the current across the LEDs increase until a point where it DROPS suddenly. But the V is pretty high, for the 13 LED string the V is around 52.5V and V across the LEDS is 3.78V (AVG) and there is much heat.

Will try the 12 LED string and revert asap.
 
This seems odd (but maybe because it is 13 LEDs). As the input Voltage increase the current should go up and at some voltage near the voltage dropped across the LEDs the current should stabilize. The the same experiment with 12 as it sounds like you intend to do.
 
perikaruppan said:
as i keep increasing the V, the current across the LEDs increase until a point where it DROPS suddenly.
Click to expand...

Does it never actually stay constant? Over the input voltage range the chip is able to regulate, the output current should be NEARLY constant (few percent variation.)

If you're changing input voltage over a wide range and never have a period of several volts where the output current is constant, then something major is wrong with the circuit, and it's not regulating output current correctly (or at all).

This period of constant output current should begin a volt or two above what the LEDs are dropping at the target output current, and it should end at the point where the IC is over any of it's limits (temperature, input voltage). It should be at least several volts wide, and the current should be somewhat near the target operating current.
 
Output current max's upto 0.65A (As per specs @Rs, Ia =0.61A).
The voltage variation i have is only between 40 to 52V and assuming the Vf =3.7, should start the system around 45V (44.4 Votage across LEDS) and up till 49 and will give the reading of the current.. if not stable then i need to re-look at what is wrong then.
 
Well i conclude that there is something wrong in my circuit. 12 LEDS, 2 Strings.

these are the readings:
45v- 0.35A, LED @ 3.4V
46v- 0.45A, LED @ 3.5V
47v- 0.55A, LED @ 3.62V
48v- 0.60A, LED @ 3.69V
49V-0.65A, LED @ 3.72V
50v-0.67A, LED @ 3.74V
and anything above that, it dims to 0.3A. So clearly it is not regulating!!!

I have already mentioned the components used. so kindly help. Otherwise will go with CAT4101.
 
I agree, it's not correctly regulating the current, the rise in current you're seeing is proportional to what you'd expect for those rises in voltage. It's almost as if your circuit is just acting like a (constant-value) resistance. My very first NCP3066 prototype behaved like this because I had the wrong timing cap, but the ZXLD doesn't have an external timing cap so we can't blame that.

I can't find your list of components but I only looked back a few pages - can you post a link or repost it? Any chance you can get your hands on a scope or anything besides a multimeter? Hard to know what's wrong at this point.

Can you post a photo or sketch of your circuit layout? I'm wondering if there's interference on the Rsense traces and it's not correctly sensing current in the string.
 
You have me confounded. I expect you have already double checked all your values and your circuit lay out. I never got around to building anything with this chip so other than than I can't offer much more advice for finding the problem. Any access to a scope? Maybe looking at the voltage wave form of the sense resistor will tell what is going on. You also mentioned your inductor was sort of non standard can you try a more standard inductor?
 
Link to the diode isn't working. Have datasheets for the inductor and capacitor?

Any chance you have spare parts sitting around in similar values - you could try swapping things around one at a time to see if there was any difference.
 
How 'bout specs on the cap you're using? Is it ceramic or electrolytic?

You're over the rated continuous current stated in the diode's datasheet but I'd expect that to cause the part to break, not the sort of operational problems you're having.

I don't think the case style for the inductor matters if the specifications are all correct. It strikes me t hat 220uH is much larger than the datasheet calls for, but again I wouldn't expect that to cause this sort of problem.

How 'bout a photo of the circuit layout?
 
when time is bad, all is bad!! my cam's battery is dead.
Shall replace and revert back as early as possible and thanks a ton for all your support.
 
Well I took the big jump. I put on the other seven circuits for my board and they all work :). Like DWZM I had a slight layout problem. The footprint for the potentiometers is a little off (luckily to big a print). I was able to bend the pins out a little and it goes in just fine.
 
I haven't tried that yet. I don't have enough LEDs to run all 8 at once. I can try 4, but need to track down a power supply. Maybe tomorrow.

Also I think I mentioned frying a CAT4101. I was worried about replacing them if the tab is soldered, but it is not so bad. Heat the tab and bend it up, then lay the soldering iron across all five pins and off it comes. No braid to clean the board so that is for tomorrow too.
 
DWZM and all CAT4101 folks,

I have been thinking about the heat issue. I think I it solved it at least for me. Each of my circuits has a potentiometer. I originally put these in for manual temporary dimming (new coral that was not acclimated). It turns out that this probably won't work, because it would drop too much voltage across the CAT4101 and produce a lot of heat and maybe go into thermal shutdown. But what it will allow me to do is balance my strings. If I have a CAT4101 that is getting a little hot then I can decrease the potentiometer to allow that string to have more current and there for more voltage drop across the LEDs and less across the CAT4101 (other LEDs issues, but assume their handled). I know some one is going to say, but then the strings are running at different current and producing differ lumens.

We already have that problem to some degree. I designed my circuit for 900 ma using a 1% 604 ohm resistor. That means that it can go fro 598-610. Extrapolating that gives a current of 889 to 911 ma. This does not give us much change in forward voltage (to fine for the data sheet graph). Let's say .05 volts well when done across 6 LEDs that is a string to string difference of .3 volts. Then let's look at the CAT 4101 the current accuracy is for a 350 ma set point is 315-385 that's 10%. If that same 10% is true at 1000ma I am getting real glad I picked a max of 900 for my circuit. Ten percent current change (from the CREE data sheet) in the 900 range might be 0.1 volts or for 6 LEDs 0.6 volts. So already there is almost 1volt possible variation in the voltage across LED strings. From what I have read this is kind of the upper limit of what we want to drop on the CAT4101. Now throw LED variation and binning becomes critical.

So here is my plan. Pick a desired current probably 700 ma, using the CREE data sheet find out the voltage drop (looks like about 3.5), which will give me a voltage drop across 6 LEDs of 21 volts. I am then going to set my power supply to 21.75 (this may change 21.6?), and then adjust each potentiometer until the voltage across each string is 21 volts. This has the added benefit of optimizing the efficiency of all the CAT4101s.
 
Here is my next prototype, going to let it run for a few days and then see if I can make 12 of these.
IMG_0646%20(small).jpg

The PCB is home made so I dont have plated through holes so I added some loops of wire soldered on both sides to act as a heat sink. Next I have to learn how to make solder paste stencil for the star boards.
 
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