try for the warranty replacement/repair, just don't volunteer any information about how it was used just that it stopped working. overheating can kill them if there isn't adequate ventilation over time.
Meanwell LDD driver: for those who want to dim to 0 using Arduino
- Thread starter pwreef
- Start date
try for the warranty replacement/repair, just don't volunteer any information about how it was used just that it stopped working. overheating can kill them if there isn't adequate ventilation over time.
n2585722
Well-known member
sfsuphysics
Active member
Oh definitely. although I did notice there's a bit of corrosion over the fan intake which is where it may get me in trouble. And I looked under the fan, and there's more there too. But then again that could be normal simply because that's where humid air in general could exit.
heatdissipation
New member
I finally figured out my problem. After getting frustrated last night I have up and today I started from square one. The first led in my string was bad. After hooking up to the next one in line they lit up. The rest of my leds won't be here till later this week so I will finish it then. This is a pretty big build and as soon as I have it done I am going to start a build thread.
I'm having a problem getting PWM working on my ldd-700hw. It works fine at full brightness, but connecting the PWM pin to my arduino results in no change.
I'm using a hydra-reef for the controller with slightly modified stock firmware (from terahz). It powers my meanwell ELN without issue. I'm using pin 10 for the ELN (with shield) and now pin 9 for the LDD.
I can measure the voltage decreasing when I lower the PWM value (around .745 v at 20% duty), but I have no way of actually measuring it other than with a multimeter.
I have it wired using a 1k resistor on pin 9, and arduino ground tied to Vout - pin on the LDD. No 10k resistor between Vout- and PWM as of yet.
Is there anything else I need to do in the Arduino or wire it differently to get it to work?
I'm using a hydra-reef for the controller with slightly modified stock firmware (from terahz). It powers my meanwell ELN without issue. I'm using pin 10 for the ELN (with shield) and now pin 9 for the LDD.
I can measure the voltage decreasing when I lower the PWM value (around .745 v at 20% duty), but I have no way of actually measuring it other than with a multimeter.
I have it wired using a 1k resistor on pin 9, and arduino ground tied to Vout - pin on the LDD. No 10k resistor between Vout- and PWM as of yet.
Is there anything else I need to do in the Arduino or wire it differently to get it to work?
heatdissipation
New member
I might be wrong but I think you need to tie the arduino ground to the vin- not the vout- Someone correct me if I am wrong. But that's how I just got mine working literally 5 minutes ago
Nice work for your first go.
A couple of observations:
1) You will need to make the traces for DCIN+ much larger. I would make a Signal BUS DCIN+ using polygons.
EDIT: You could also simply rename your bottom ground polygon from GND to DCIN+
Assuming you were using LDD-1000H's.....
If you had all 6 LDD's running at 100% you need to supply up to 6 amps plus ~20% headroom. 7.2 AMPS!
Using the calculator linked below using 2oz^ft copper boards your traces need to be 90 mils yours are 16 mils.
2) Likewise you should check the datasheet for your DCIN screw terminal and ensure it can handle the current.
3) You will also need to increase the trace width of all of your Outputs LED+1 / -1 thru LED+4 / -4. At full power they need to handle 1 amp plus ~20% headroom.
4) Invoking Ratsnest when you first load a board file to show pours is normal for Eagle with all operating systems. It makes sense to me as I do 99.9% of my work without the polygons filled. Easier on my eyes.
You also might want to make one of your GND polygon perimeters larger than the other. It makes it much easier to choose one or the other. You can make them larger than the perimeter of the board. Eagle will only fill the board never outside.
There are some good Trace Width calculators on the net.
Here is one I have used. http://circuitcalculator.com/wordpress/?p=25/
Cheers,
-BB
Last edited:
Yup that was it. Somehow I screwed it up, but amazingly didn't kill my arduino or drivers
Thanks!
the current you'll be running also needs to factor in. 3.3A @ 48V will be fine if running 700mA LDDs however if using the 1000mA version with full strings you'll potentiall be drawing over 179watts which exceeds that PSU.
@700mA driver current you'd be pulling 125watts which will pretty nearly max out the PSU when considering that the LDDs will consume some power as well.
if using less than 14 per string then you need to figure out the projected total watts being used by the leds at full power and use that number to size the PSU, keeping a 15-20% overhead is a good practice with the PSU selection as if they are running at max they have better chance of failing due to heat build up, etc.
was the bad one a new LED? it happens but not real often. Did you perhaps hook it up backwards? been there, done that
heatdissipation
New member
Ya. It was a new led. It wasn't hooked up backward. Just a dud. It happens. I'm just glad it wasn't anything more.
summoner12
New member
Human Battery Man! :lol2:
joshlawless
New member
Thanks for all your helpful input. I was working up an alternative layout based on your previous post -- here's what I've got so far:
Now I understand why there were multiple DCIN+ and GND connections spread over multiple screw terminals in O2Surplus and RRasco's 3-, 4- and 5-up versions. I've added another two screw-terminal block, and increased the trace sizes for the LED outputs, and built a hefty bus for the DCIN+.
I've also tried to improve the labeling somewhat.
V1.1 is attached.
Last edited:
Thanks for all your helpful input. I was working up an alternative layout based on your previous post -- here's what I've got so far:
Now I understand why there were multiple DCIN+ and GND connections spread over multiple screw terminals in O2Surplus and RRasco's 3-, 4- and 5-up versions. I've added another two screw-terminal block, and increased the trace sizes for the LED outputs, and built a hefty bus for the DCIN+.
I've also tried to improve the labeling somewhat.
V1.1 is attached.
That's looking good! Those traces are plenty "hefty" and will work fine.
joshlawless
New member
Super ignorant question, but have I got those screw terminals attached the correct way (so that the wires come from off the board)? I'm not sure which side is which from looking at the design in Eagle -- there's a little triangular arrow opposite pin 1 that points in towards the PCB in the current design. Did I get that right?
Thanks!
