DIY LED driver for reef lighting
- Thread starter der_wille_zur_macht
- Start date
der_wille_zur_macht
Team RC
I'm guessing it's the drivers in the 17's that are overheating? You could add an LED to each of those strings, sure - you may have to turn the power supply up a few tenths to keep it at least .5v above the output.
Or you could get a different DC supply and rearrange which strings are on which driver so the strings dropping 17v are all on the same PCBs...
Or you could get a different DC supply and rearrange which strings are on which driver so the strings dropping 17v are all on the same PCBs...
TheFishMan65
New member
Not as fancy and a little wasteful of electricity, but you could add some power resistor to balance the strings as well. It won't work if you are changing the current since the voltage drop will also change. But depending on what you have laying around this may be a temporary solution so you can get them running.
der_wille_zur_macht
Team RC
Or even just a huge fat diode or two... Basically you'd still be burning off the same amount of power as heat, but you're diving up the job of burning it off between the IC and whatever you add, instead of letting the IC burn it all off and overheat itself.
zachtos
Active member
Say you want to make a typical 400W MH replacement nowadays. IE, 1000 PAR at first 6" depth, and 300-500PAR at bottom around 24" depth. I'm starting to feel 48 LED 1:1 ratio XPG coolwhite to XPE royal blue all run at 1A and 40 degree optics? Will this cover a good 24" x 24" square with what I'm asking above? I'm thinking equal spacing on a 6" x 12" piece of aluminum will work. Or is it closer to 36 LED's? I'm just getting ready to build a prototype unit to get it ready to scale up to 4 to 5 fixtures to replicate 4 to 5 - 400W MH.
Last edited:
der_wille_zur_macht
Team RC
Those guesses are probably in the right ballpark. With the exact specs you described, probably 36 LEDs.
What sort of color are you looking for? Mounting height? Optics?
Might get more results for your question in a standalone thread vs. buried here in the DIY driver thread.
What sort of color are you looking for? Mounting height? Optics?
Might get more results for your question in a standalone thread vs. buried here in the DIY driver thread.
zachtos
Active member
Meant to say for 1000 par top and 500 par bottom, LED = square inches/12... Not sure if that is accurate. Mount height/optic/spacing does not matter as long as we can get the designated PAR at the most efficient wattage. I can print up a modified version of these drivers no prob, but not for the prototype. I prefer a 12 to 14K color w/ best PAR. I can also do an ATMEGA controller w/ LCD etc no problem. I've been looking for a thread that has exactly what I'm looking for, but very few have REAL PAR readings with a DIY like I said above (I have a meter but many dont).
cat4101 and pwm ?
cat4101 and pwm ?
I have a question one the cat 4101. If there is no voltage going into the pwm pin then the leds shouldn't light right? So why when i take a wire and connect it to the pwm pin and touch the other end of the wire with my bare hands the leds light up? is this a grounding issue? everything seems grounded right power supply,cat4101 and arduin all grounded together.
Seems like when i do that im acting like a ground for the pin but i thought the pwm pin was a input not a output ill get my voltmeter out tonight and hook it to the pin and a ground see if i get any reading.
cat4101 and pwm ?
I have a question one the cat 4101. If there is no voltage going into the pwm pin then the leds shouldn't light right? So why when i take a wire and connect it to the pwm pin and touch the other end of the wire with my bare hands the leds light up? is this a grounding issue? everything seems grounded right power supply,cat4101 and arduin all grounded together.
Seems like when i do that im acting like a ground for the pin but i thought the pwm pin was a input not a output ill get my voltmeter out tonight and hook it to the pin and a ground see if i get any reading.
So Terahz, what ever happened to the idea of JUMPER SELECTABLE? Do you still plan on adding this or did you dump the idea?
Gorgok
New member
Interestingly enough i did the same test/play when i first had forgotten my GND to the controller and was able to dim and light the LEDs with just pressure applied to the PWM wire. It is way cool...
What i assume is that the PWM pin is just sensitive enough to pick up the electricity you generate to move muscles, and when squeezing more or less the duty cycle changes to the muscles, IE PWM. Either that or you are picking up stray voltage from somewhere and conducting it to the wire... I like the first explanation better though.
terahz
1x10^12 Hz
That's still the plan.
My guess is stray voltage ... in electronics lab in HS we used to the oScopes with a probe on our finger, and it was amazing what you pick up. Try waving your hand around near stuff. You might dim with a gesture :spin2:
== John ==
terahz
1x10^12 Hz
My LM3409HV v0.5 PCBs arrived today. I've included the additional small jumper to disconnect the IADJ pin from the analog circuit for "jumper only" configuration with everything populated.
Final v0.5 that went to the PCB house looks like this:
And here are the 3 revisions that I ended up ordering to get to here (please excuse the rosin covered board, too sleepy to clean it):
I have not done the same load test as with the previous revision yet, but did a quick test with a new PFET and 8 XPGs. I'm happy to say that the FQD17P06TM FET did not even feel like it was on with the 8 LEDs shining at almost 1A.
Efficiency seems fairly good too. The board was powered by 31.985V and was drawing 770mA. Voltage drop for the 8 XPGs was 24.68V and final maximum current was 932mA.
With the resistors I have selected(100K and 50 Ohms) for setting the cutoff voltage, the comparator circuit worked like a charm, turning off the driver at ~4mV. As a reference, if you use Arduino PWM to generate an analog voltage with an RC filter, PWM of 1 is about 20mV, so you don't loose any resolution.
I'll probably put together the v0.5 files tomorrow and post them.
Enjoy.
P.S. The v0.5 PCB fits nicely in a small Altoids tin
Final v0.5 that went to the PCB house looks like this:
And here are the 3 revisions that I ended up ordering to get to here (please excuse the rosin covered board, too sleepy to clean it):
I have not done the same load test as with the previous revision yet, but did a quick test with a new PFET and 8 XPGs. I'm happy to say that the FQD17P06TM FET did not even feel like it was on with the 8 LEDs shining at almost 1A.
Efficiency seems fairly good too. The board was powered by 31.985V and was drawing 770mA. Voltage drop for the 8 XPGs was 24.68V and final maximum current was 932mA.
With the resistors I have selected(100K and 50 Ohms) for setting the cutoff voltage, the comparator circuit worked like a charm, turning off the driver at ~4mV. As a reference, if you use Arduino PWM to generate an analog voltage with an RC filter, PWM of 1 is about 20mV, so you don't loose any resolution.
I'll probably put together the v0.5 files tomorrow and post them.
Enjoy.
P.S. The v0.5 PCB fits nicely in a small Altoids tin
Hi Terahz,
I'm a newbie in this SW hobby have been reading your LM3904 build for couple weeks now. I'm planning to go with led for my new tanks and have a few questions:
- If I'm running 14 XPE leds with Vf=46.2V what is the minimum Vin I have to have?
- What would be the rough estimate cost per board (PCB & components)?
- Analog dimming is it 0-5V?
Thanks a lots for you help.
Tron
I'm a newbie in this SW hobby have been reading your LM3904 build for couple weeks now. I'm planning to go with led for my new tanks and have a few questions:
- If I'm running 14 XPE leds with Vf=46.2V what is the minimum Vin I have to have?
- What would be the rough estimate cost per board (PCB & components)?
- Analog dimming is it 0-5V?
Thanks a lots for you help.
Tron
terahz
1x10^12 Hz
Just finished building my driver box with whatever I had laying around. The nice thing about analog dimming is that you can always put a pot in the middle of the signal and have manual override.
Wiring done:
All stuffed in. Binding posts for 48V input, DB-9 for dimming signal and output is a single ethernet cable:
I'll probably be replacing my CATs tomorrow or next week.
Wiring done:
All stuffed in. Binding posts for 48V input, DB-9 for dimming signal and output is a single ethernet cable:
I'll probably be replacing my CATs tomorrow or next week.
kcress
New member
Well... under-drive anyway.
new2SW
New member
So I built one of the 8 driver Cat4101 boards last night and could not get it to fire up. I haven't tested all the drivers on the board; however, I have one question as I proceed. Do I have to have a PWM signal in order for the circuit to work? Fishman had told me that I could just jump 5v to the signal for now, does this mean just supply 5v from a seperate source other than what's already supplied to the Vin from the onboard 5v? I'm building up a test string from som XR-Es I have laying around so I don't fry my actual tank LEDs.
