DIY LED driver for reef lighting

der_wille_zur_macht said:
Eagerly awaiting results!
Click to expand...

Well, the results were ... mixed :)

The circuit worked fine, but only briefly (4 mins or so) then it switched off - just as I was getting ready to take a photo in fact :).

I think the sense resistor burnt out (which is odd, because the current through it ought to be miniscule at 1.2v and 768 ohms) - I couldn't get a reading on the multimeter for any resistance through it once the circuit had switched itself off, though.

I'm going to put together a circuit with slightly beefier components and try again. I'll keep y'all posted...

Simon
 
Interesting.

Did you measure the resistance on the sense resistor in circuit? Could it be possible there was a problem elsewhere (in the IC itself? Or a solder bridge?) that was causing the weird reading?

Still promising that it worked briefly. Any sense of how warm it was running? How many LEDs and what supply voltage?
 
Interesting.

Did you measure the resistance on the sense resistor in circuit? Could it be possible there was a problem elsewhere (in the IC itself? Or a solder bridge?) that was causing the weird reading?

Still promising that it worked briefly. Any sense of how warm it was running? How many LEDs and what supply voltage?
 
der_wille_zur_macht said:
Interesting.

Did you measure the resistance on the sense resistor in circuit? Could it be possible there was a problem elsewhere (in the IC itself? Or a solder bridge?) that was causing the weird reading?
Click to expand...

It's possible - I was measuring it in-circuit, but I got good readings (within a couple of percent) in-circuit prior to the shutdown, so I figured the change was relevant ...
der_wille_zur_macht said:
Still promising that it worked briefly. Any sense of how warm it was running? How many LEDs and what supply voltage?
Click to expand...
The device didn't even get warm - I had my finger on-top and it was cool to the touch. When it's in front of you, it completely dwarfs the 0603 capacitors/resistors around it - it's huge by comparison...

I was running my test-rig of 6 Endor rebel stars (Vf=3.15v, so I was feeding 19.4v through it from the bench psu).

Simon.
 
Well, I made a new PCB for the different parts but I still couldn't get anything out of the circuit - I think the chip must have fried, not the resistor; the resistor measured correctly out-of-circuit... I suppose I've been spoilt with the (relatively bulletproof) ZXLD1366...

For once I didn't order multiple chips, so I'll have to wait until digikey deliver some more... Guess you'll have to wait too :(

Simon
 
ok I just did my prototype with the CAT4101. I don't have high power LED. so I did use 44 regular leds (22 strings of 2) so I use about 660mA.

The prototype is running good so far (on since 2 hours). Did tests with the PWM of my pic16F877 board and everythings looks good.

Next step: try to found some LEDs ......
 
Mea culpa

Mea culpa

So, it turned out that the reason the circuit wasn't working was a grounding issue - doing a homebrew board without plated-through vias, I'd missed that part of my ground plane just wasn't...

Anyway, thinking there was something wrong with the PCB and redoing it with non-SMT components was a blessing in disguise - this thing is totally home-solderable. See the pic of the PCB. Everything on there was hand-soldered. Even the ground-tab that is nominally underneath the chip has a protruding edge, so I put some thermal paste on the underside (where I had the vias), and just soldered the protruding tip of the tab to the ground plane.

And, then it worked ... It's been on for an hour or so and the driver is still cool to the touch...

Simon
 
SpacedCowboy; Way to go.

I usually just solder those with a large iron. Heat that tab and feed in the solder at a high rate.
 
Has a working prototype board been made available yet?

I have been reading along but can't seem to recall a working board yet.

If there is, will it be made available to download in Eagle or one of the other CAD formats to be sent off and a PCB made?

I ask because my last original PS for my Solaris has died. I will make a new DIY cable and use an off the shelve PS as I have done in the past, but I want to start working away from the Solaris LED Boards and this setup.

Thanks
 
SpacedCowboy said:
And, then it worked ... It's been on for an hour or so and the driver is still cool to the touch...

Simon
Click to expand...

Woohoo! I'm going to do up a board design with several of them on one PCB. I see the sense resistor is available in 1206 or 1210, and the caps in sizes just a little smaller than that, which shouldn't be too bad.

stephenrwright said:
I found this 24V 6.5A power supply for $19.00
http://webtronics.stores.yahoo.net/pospu1524vsi.html
Anyone ever used Webtronics before? The name sounds familiar but its been years since I played with electronics. For $19 I will give it a shot.
Click to expand...

Worth a try. Might want to do "generic" questions like this in one of the big LED threads. Most people use the MPJA 6.5A or 8.3A supplies which are only $15.

shackscs said:
Has a working prototype board been made available yet?
Click to expand...

Several prototypes have been made:

1) My original NCP3066 "dual driver" design works just fine, and the EAGLE project files for two versions are up on the google code site linked a few pages back.
2) A few people have gotten the ZXLD1366 working, but it's a hair expensive.
3) As is obvious at this point, Simon has a prototype of the CAT4101 working right now.

When I get around to actually ordering the PCBs for my big tank, I'm going to order a bunch of extras so people can get in on those (at cost) if they want. Otherwise, you're pretty much on your own. :)
 
I was running the voltage at 19.4v (these are Endor rebel stars with a forward voltage of 3.15v. 6 x 3.15v + 0.5v = 19.4v); the thing is that I was only getting 0.4A being pulled from the PSU - in fact you can see the reading on the PSU in the photo in my earlier post ...

So, given that the driver was running cool and that the LEDs weren't *that* bright, coupled with the fact that I trust my bench PSU, I thought I'd up the voltage until the current stayed constant - that happened at 20.7v, with a constant current of 0.82A. That current is pretty much in-line with what I'd expect with a 680R resistor (the datasheet says it ought to be 0.8A).

So, at least with these LEDs, it looks as though the efficiency isn't quite as good as we were thinking ...
<table><tr><td align=right>Total voltage being fed to the circuit = </td><td>20.7</td></tr><tr><td align =right>Total drop voltage over the leds = </td><td>6 x 3.15v = 18.9v</td></tr><tr><td align =right>Voltage consumed by non-LED parts of circuit = </td><td>(20.7 - 18.9) = 1.8v</td></tr><tr><td align =right>Power dissipated by non-LED parts of circuit = </td><td>1.8 x 0.82 = 1.476W</td></tr><tr><td align =right>Power dissipated by LEDs = </td><td>(6 x 3.15 x 0.82) = 15.498W</td></tr><tr><td align =right>Efficiency rating = </td><td>15.498 / (15.498 + 1.476) ≈ 91% efficient</td></tr>
</table>
This assumes the LEDs drop what they say they do, of course. 91% is also still reasonably good when measuring in the real world rather than reading datasheets, at least IMHO. DWZM might disagree [grin].

After 30 mins or so running at this rate, the driver is noticeably warmer to the touch as well, but not hot - I can comfortably hold my finger on the chip, and it only feels warmer-than-my-finger.

Simon
 
did connect yesterday a power resistor (instead leds (I don't have yet) to load 930mA). Did diy a heatsink for the CAT (TO220 heatsink) and after 1 hours, it was jusr little warm ... looks good so far. Did burn the buffer between my uP and the CAT. Did put a 5.1V zener on the output of the buffer ... Did not catch a overvoltage on the scope but for now that's working good like I said after 1 hour ....
 
SpacedCowboy,

Did you measure the voltage across the LEDs with a meter? Looking at the datasheet I would beleive 3.15 volts at 0.4 amps, but maybe a little more. At 0.8 amps it looks closer to 3.55 (or more). So 6x3.55 is 21.3 volts. Which is more than you are supplying - so obviously my numbers don't work. But I do wonder where the 3.15 came from. The data sheet says the voltage should be a little higher with more current.

Thanks
 
Last edited:
You must log in or register to reply here.

Similar threads

C
DIY Stand for Flatback Hexagon Tank
Replies
2
Views
763
Drummer98
Drummer98
T
Looking for a specific tank build - EZTube.com sent me here
Replies
6
Views
537
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