TheFishMan65
New member
Well I was testing one and it was working. But I think I fried it. Always check your voltage BEFORE you apply it. OOPS
DIY LED driver for reef lighting
- Thread starter der_wille_zur_macht
- Start date
TheFishMan65
New member
OnSemi's Answer
OnSemi's Answer
I got an answer from OnSemi, I like stu's much better
OnSemi's Answer
I got an answer from OnSemi, I like stu's much better
perikaruppan
Member
Well, to give another feedback on the ZXLD...
I wired up 12 CW and 12 RB's to 2 ZXLDS in one board. As mentioned earlier, i have 7 Drivers in a 4" X 4" Board. So as suggested by FishMan, going to use 3 maybe 4 in one board.
And lit them up... adjusted the voltage on the Power Supply to get A constant current at 610mA. The Voltage across the CW LEDs were 3.79 V and across the RB LEDS were 3.77V on average, with a few leds going +/- 0.02V...
And BTW, these are all EDISION LEDS with CREE Dye. And they suggest the max Vf can go upto 4.3V. Driving them @ 3.8V is optimum.
And i have got some heat like before.. but bearable on it... so running a 3" DC fan on top of it and found it to work properly.
Got some heat on the heatSink, as there are no Fans on top of it to cool ... so i think once the fans are on then that will also be taken care off!!!...
Well @ 47V input supply the total Power on the LEDS is 45.48V (3.79X12) and estimated efficiency of 96.76%!!! what do you all think about that?
And now as per DWZM idea and Spec sheet, i did a PWM Circuit with a transistor and resistor to the ZXLD. Going to use an Arduino to control the Dimming. Just got a few questions:
1. If i set the Analog out Signal Value from 0 to 255 the brightness on the LED should increase from 0 to 100% right?
2. Assuming i use Pin 9 of the Ardunio to PWM, then just connect pin 9 to the Transistor circuit of the ZXLD right?
3. Do i need to get the ground of the Arduino connected to the GND of the ZXLD Driver Circuit for the circuit to work properly? even though the DC adapter to driver the ardunio and SMPS for the ZXLD are connected to the same AC power Supply?
Will post some pics soon.
I wired up 12 CW and 12 RB's to 2 ZXLDS in one board. As mentioned earlier, i have 7 Drivers in a 4" X 4" Board. So as suggested by FishMan, going to use 3 maybe 4 in one board.
And lit them up... adjusted the voltage on the Power Supply to get A constant current at 610mA. The Voltage across the CW LEDs were 3.79 V and across the RB LEDS were 3.77V on average, with a few leds going +/- 0.02V...
And BTW, these are all EDISION LEDS with CREE Dye. And they suggest the max Vf can go upto 4.3V. Driving them @ 3.8V is optimum.
And i have got some heat like before.. but bearable on it... so running a 3" DC fan on top of it and found it to work properly.
Got some heat on the heatSink, as there are no Fans on top of it to cool ... so i think once the fans are on then that will also be taken care off!!!...
Well @ 47V input supply the total Power on the LEDS is 45.48V (3.79X12) and estimated efficiency of 96.76%!!! what do you all think about that?
And now as per DWZM idea and Spec sheet, i did a PWM Circuit with a transistor and resistor to the ZXLD. Going to use an Arduino to control the Dimming. Just got a few questions:
1. If i set the Analog out Signal Value from 0 to 255 the brightness on the LED should increase from 0 to 100% right?
2. Assuming i use Pin 9 of the Ardunio to PWM, then just connect pin 9 to the Transistor circuit of the ZXLD right?
3. Do i need to get the ground of the Arduino connected to the GND of the ZXLD Driver Circuit for the circuit to work properly? even though the DC adapter to driver the ardunio and SMPS for the ZXLD are connected to the same AC power Supply?
Will post some pics soon.
der_wille_zur_macht
Team RC
A few people have PM'd me about gerber files for the PCBs. I had put Seeed's CAM job up on the google code site, but looking at their site today, it looks like they've changed their requirements.
So, rather than posting the CAM jobs or the gerbers on the code site and trying to maintain them for different board houses and how those board houses' requirements change over time, I'd just suggest you get the latest guidelines from your board house and use those at the time you purchase boards.
For reference, here's Seeed's Fusion service order page outlining the requirements. Down near the bottom there's a link to a file that has a rules file and a cam file so you can check the design and generate gerbers with appropriate format and filenames, etc:
http://www.seeedstudio.com/depot/fusion-pcb-2-layer-5cm5cm-max-10pcs-p-513.html?cPath=64_12
So, rather than posting the CAM jobs or the gerbers on the code site and trying to maintain them for different board houses and how those board houses' requirements change over time, I'd just suggest you get the latest guidelines from your board house and use those at the time you purchase boards.
For reference, here's Seeed's Fusion service order page outlining the requirements. Down near the bottom there's a link to a file that has a rules file and a cam file so you can check the design and generate gerbers with appropriate format and filenames, etc:
http://www.seeedstudio.com/depot/fusion-pcb-2-layer-5cm5cm-max-10pcs-p-513.html?cPath=64_12
TheFishMan65
New member
I will let someone more familiar with the Arduino answer the first 2.
You will be better off connecting the grounds together. They may or may not be connected depending on how the power supplies work. If they use a transformer (I think this is the most common) to go from 120 (or whatever your voltage is) to 47 then there is no connection between the AC plug and the 47 volts. Transformers work by magnetic induction (I think that is the right term), the 120 AC produces a magnetic field in a coil another coil is interlaced and uses that magnetic field to produce 47 volts which is then converted to DC. So there is no connection between the AC and DC. Of course the PS could use something other than a transformer.
You will be better off connecting the grounds together. They may or may not be connected depending on how the power supplies work. If they use a transformer (I think this is the most common) to go from 120 (or whatever your voltage is) to 47 then there is no connection between the AC plug and the 47 volts. Transformers work by magnetic induction (I think that is the right term), the 120 AC produces a magnetic field in a coil another coil is interlaced and uses that magnetic field to produce 47 volts which is then converted to DC. So there is no connection between the AC and DC. Of course the PS could use something other than a transformer.
der_wille_zur_macht
Team RC
Very good results!
Efficiency will always be higher with a longer LED string, assuming a fixed drop across the driver circuit. You are seeing right around what the datasheet predicts.
1) In theory, yes, but the datasheet doesn't show a response curve to PWM dimming so we don't know if there's a cutoff or not.
2) Yes, pin 9 goes to the resistor which is in series with the base on the npn transistor.
3) Yes, GNDs must be connected. Connect the GND from the Arduino to the GND on the output side of the DC supply powering the LED driver.
perikaruppan
Member
So will ground the Ardunio and GND of the SMPS to make a common ground.
So lemme keep my fingers crossed and dim it and see. I was praying all this while that my LEDs should Blow!!! now the prayers should be double i suppose!!!...
so just send a value from 0 to 255 with a delay.. that should brighten it right? and from 255 to 0 with a delay to dim it. i hope i am correct
So lemme keep my fingers crossed and dim it and see. I was praying all this while that my LEDs should Blow!!! now the prayers should be double i suppose!!!...
so just send a value from 0 to 255 with a delay.. that should brighten it right? and from 255 to 0 with a delay to dim it. i hope i am correct
der_wille_zur_macht
Team RC
For a test run, you can use the example "Fading" sketch. It's under Examples>Analog>Fading in the File menu.
perikaruppan
Member
ya thats what i was about to use... thanks a ton.
der_wille_zur_macht
Team RC
Be "warned" that the example sketch isn't very smooth. It jumps in steps of 5 each iteration of the loop. Nothing wrong with that, but it looks "choppy." When I test using that sketch I change the increment in the two loops to 1, and increase the delay before and after each loop so it "parks" at 0 and 255 for a little longer.
cat4101tv-t75
perikaruppan
Member
did that... increments of 1... delay of 1000ms before and after loop and increased loop delay to 100ms instead of 30ms...
is it alright?
or shall i reduce the 255?
is it alright?
or shall i reduce the 255?
der_wille_zur_macht
Team RC
It really doesn't matter, just test it with whatever values you want - eventually you're going to want to test the whole range I'd imagine, so might as well start now. Reducing it won't be any "safer" since it's just turning the entire circuit on/off really fast - when it's on it's on, wether that's for 100% of the time or 80% - and you already know it works when it's on, so might as well test the whole range.
perikaruppan
Member
tested it ... and it is not working !!!...dont know what the problem is.
PWM - Resistor - Base of Transistor.. Emitter is grounded, and Collector is to the ADJ pin.
using a BC547 transistor.
PWM - Resistor - Base of Transistor.. Emitter is grounded, and Collector is to the ADJ pin.
using a BC547 transistor.
perikaruppan
Member
used a 1.8k resistor. Shall i byepass that and connect directly to the Base of the Transistor as shown in the Spec sheet?
TheFishMan65
New member
Don't know the circuit real well, but you can test the transistor circuit by applying 0 volts and 5 volts. At 0 the LEDs should be off, and 5 they should be on. If this works then the questions become is the PWM really sending a pattern (got a scope). And is the PWM frequency too fast or slow (also a scope). But just apply 0 or 5 volts should give some idea if your circuit is working. I am assuming the circuit is designed for 0 to 5 volts if not use whatever it is designed for,
der_wille_zur_macht
Team RC
There's an internal pullup in the ZX IC that sets the pin at +1.25v when floating, which is what you want. Pulling it up to +5v would fry the chip and/or the LEDs.
