SpacedCowboy
New member
I've attached the Eagle project (with the part in it), in case that helps
Simon
DIY LED driver for reef lighting
- Thread starter der_wille_zur_macht
- Start date
I've attached the Eagle project (with the part in it), in case that helps
Simon
SpacedCowboy
New member
You're right, the forward voltage is higher at 800mA. I don't know how I missed that... From the graph on page 15 of the datasheet, I'd say it was more like 3.45v, not 3.15v. Given that 6 x 3.45 is 20.7, which is precisely the input voltage, I think we have a 100% efficient driver [grin].
Ok, so TANSTAAFL... and I guess the driver temperature will affect things, so all I can really say at this point is that it's actually pretty darn efficient.
Simon
TheFishMan65
New member
Thanks, you also explained why my number where off. I assumed a Cree LED.
SpacedCowboy
New member
One last update...
One last update...
... because I'm sure people are getting sick of this now [grin], but I left the LEDs on overnight, and apart from a slight increase in current-draw (0.82A yesterday, 0.84A this morning) which I put down to temperature effects, it's working just like it did yesterday
The driver is still coolish-warm to the touch as well - also a good sign
Simon.
One last update...
... because I'm sure people are getting sick of this now [grin], but I left the LEDs on overnight, and apart from a slight increase in current-draw (0.82A yesterday, 0.84A this morning) which I put down to temperature effects, it's working just like it did yesterday
The driver is still coolish-warm to the touch as well - also a good sign
Simon.
der_wille_zur_macht
Team RC
TheFishMan65
New member
Based on SpacedCowBoys test I am switching from the XLD1366 to the CAT4101. I am wondering what I should try and design in. Does anyone (besides me) worry about PWM dimming adversly affecting corals? Has anyone tried PWMs with Coral?
I am going to add a 10k potentiometer ins series with the sense resistor. I don't currently have anyother way to dim them and I figure for a dollar it is a pretty reasonable thing to add. If I decide I don't need them I can always jumper them out. Is there any reason that anyone can see to make these pots be able to be controlled by the arduino (when I one day add this in)?
Seems like an arduino should have no trouble driving the PWM with a 200K input resistor. Doesn't have a 40ma supply? So I think I will have 8 drivers on a board with inout for two separate PWMs. I know DWZM did three, but becuase of a space limitation if I am correct. Makes sense to me to al least have an even number half white and half royal blue.
I am considering adding a LM7805 for the 5 volt supply.
I am going to add a 10k potentiometer ins series with the sense resistor. I don't currently have anyother way to dim them and I figure for a dollar it is a pretty reasonable thing to add. If I decide I don't need them I can always jumper them out. Is there any reason that anyone can see to make these pots be able to be controlled by the arduino (when I one day add this in)?
Seems like an arduino should have no trouble driving the PWM with a 200K input resistor. Doesn't have a 40ma supply? So I think I will have 8 drivers on a board with inout for two separate PWMs. I know DWZM did three, but becuase of a space limitation if I am correct. Makes sense to me to al least have an even number half white and half royal blue.
I am considering adding a LM7805 for the 5 volt supply.
kcress
New member
I'm going to say it won't matter a bit. My reasoning is that the coral is using the light in a photochemical application where photons in plus some chemicals equal reaction.
Take your eye for example. The photons hit your rods or cones and consume chemicals in the process. Those chemicals then need to be replenished which takes some amount of time. If the light is OFF briefly that replenishment period is still occurring, setting up for the next flash. I doubt they can even detect the flashes. We can't and our eye's are faster than their photo chemical responses - that aren't trying to capture images.
But this is all just the theory I'd operate under until proven false. :jester:
TheFishMan65
New member
I sort of agree, but if you go back to post 1069 (page 43) in the DIY LEDs:
http://www.reefcentral.com/forums/showpost.php?p=16381996&postcount=1069
You can read what I am talking about. Quick summary photosyntheis is happening in the nano and pico seconds which is a lot of time compared to the PWM frequency. DWZM found an article about tomatoes that help support that the behavior is the same. The part missing was if you are too bright (so bright the coral get's too may photons and bleaches), and dim with PWM can you still do harm? Will short burst of too bright be the same as too bright all the time? There was no answer to the question.
I may just be looking for problems - that why I asked about experience. All the LED setup I have read about don't mention dimming by PWM.
http://www.reefcentral.com/forums/showpost.php?p=16381996&postcount=1069
You can read what I am talking about. Quick summary photosyntheis is happening in the nano and pico seconds which is a lot of time compared to the PWM frequency. DWZM found an article about tomatoes that help support that the behavior is the same. The part missing was if you are too bright (so bright the coral get's too may photons and bleaches), and dim with PWM can you still do harm? Will short burst of too bright be the same as too bright all the time? There was no answer to the question.
I may just be looking for problems - that why I asked about experience. All the LED setup I have read about don't mention dimming by PWM.
terahz
1x10^12 Hz
Hmm, I don't think that's how it works. It is true that PWM is a burst of on and offs but the end "device" sees an average, doesn't it? Think about PWMing a fan. You don't have a fan that's spinning on high and then turning off very quickly, but a fan that sees a lower voltage thus spinning at lower RPM. I always thought that's what happens with LEDs, not quick bursts of too bright and then no light, but in fact just LEDs that are ran at lower current.
My LEDs are PWMed at 60% all the time (on a 1A drivers) and I don't see any problems in terms of bleaching. In fact my corals have been coloring up nicely recently.
perikaruppan
Member
What kind of power supply do you all use to drive these drivers? DWZM mentioned about mjpa power supply but they are all SMPS power supplies. And i heard that they do have a larger ripple when compared to that of linear Power supplies.
I was thinking of making a linear power supply for 30V/5A using a LM338 to drive a 8 LED series X 5 Parallel = 40 LEDS using one Power Supply.
similar to this :
will eliminate 0.02ohms at the output, take the output of 32V along the 2200Capacitor and make the input as a 15-0-15 transformer.
Thinking of using the ZXLD 1366 as i have already procured 16 nos of them. apart from the costing perspective, is there anything wrong with the ZXLD?
I was thinking of making a linear power supply for 30V/5A using a LM338 to drive a 8 LED series X 5 Parallel = 40 LEDS using one Power Supply.
similar to this :
will eliminate 0.02ohms at the output, take the output of 32V along the 2200Capacitor and make the input as a 15-0-15 transformer.
Thinking of using the ZXLD 1366 as i have already procured 16 nos of them. apart from the costing perspective, is there anything wrong with the ZXLD?
TheFishMan65
New member
terahz,
The PWMs I am familiar with turn on and off. It sort of depends on the circuit. The fan is probably being turned on and off the mementum of spinning up and down just makes seem to be a lower voltage and spinning at a slower rate. The ZXLD1366 used an inductor to try and regulate the current, but there was still some ripple. But at least I know someone is using it. A scope would tell you for sure. But if there is no inductor in you circuit then it is probably on/off.
Couple of thoughts I had over night. kcress I have heard (and could probably find reports) the epileptic can have trouble with fluorescent lights, but they don't affect most people. So as long as we don't have epileptic fish or corals or a similiar behavior we are OK. No disrespect of an epileptics you are just the only examples I know of.
My other thought was that we should (and I probably will) use PWM at 100% for daylight settings. The means that the short burst are guaranteed not to be bright enough to damage anything since they are on at 100% later in the "day" and are not hurting anything. Thoughts anyone.
kcress, we will continue to learn more, but it is the questions that get asked that will get us there.
Based on what I have read so far I am going to go with PWM (0 - 100%) dimming, but use a resistor to set my maximum current. If I have to I can redesign the driver not to use PWM. The CAT4101 makes it cheap enough that if I have to I can, and depending on the redesign I maybe able to use most of the current design.
perikaruppan
I far as I am concerned the ZXLD1366 is still good - it is my fall back. I got 4 (do you need more) to start with, and drew up a schematic. It is still cheaper for me to switch to the CAT4101.
The PWMs I am familiar with turn on and off. It sort of depends on the circuit. The fan is probably being turned on and off the mementum of spinning up and down just makes seem to be a lower voltage and spinning at a slower rate. The ZXLD1366 used an inductor to try and regulate the current, but there was still some ripple. But at least I know someone is using it. A scope would tell you for sure. But if there is no inductor in you circuit then it is probably on/off.
Couple of thoughts I had over night. kcress I have heard (and could probably find reports) the epileptic can have trouble with fluorescent lights, but they don't affect most people. So as long as we don't have epileptic fish or corals or a similiar behavior we are OK. No disrespect of an epileptics you are just the only examples I know of.
My other thought was that we should (and I probably will) use PWM at 100% for daylight settings. The means that the short burst are guaranteed not to be bright enough to damage anything since they are on at 100% later in the "day" and are not hurting anything. Thoughts anyone.
kcress, we will continue to learn more, but it is the questions that get asked that will get us there.
Based on what I have read so far I am going to go with PWM (0 - 100%) dimming, but use a resistor to set my maximum current. If I have to I can redesign the driver not to use PWM. The CAT4101 makes it cheap enough that if I have to I can, and depending on the redesign I maybe able to use most of the current design.
perikaruppan
I far as I am concerned the ZXLD1366 is still good - it is my fall back. I got 4 (do you need more) to start with, and drew up a schematic. It is still cheaper for me to switch to the CAT4101.
der_wille_zur_macht
Team RC
Simon, some more questions about your CAT4101 experiment:
1) Any chance you've tried to dim it? What are you feeding the EN/PWM pin?
2) How are your lead lengths to/from the PCB? Thoughts about how extending them might cause problems? I know I've seen in several LED driver datasheets that additional capacitance is required across the LEDs and/or input if lead lengths are long. Thoughts?
3) Thoughts on heat dissipation? If I follow my standard practice of cramming as many as I can "fit" on to a 5cm square PCB, that means I'm working with a little over 3 square inches. The datasheet uses 1 square inch (double sided) for heatsinking when calculating power dissipation. That means I can "fit" three on this size PCB in terms of thermal limitations. But if it's not really getting THAT warm, maybe 4 wouldn't be a stretch?
4) Any chance you've cranked up the voltage to see what happens when the IC is dropping more than the bare minimum? I know that's not best practice to run it like that, but I'm wondering what it'll do.
Thanks again for being the prototyper for this chip!
1) Any chance you've tried to dim it? What are you feeding the EN/PWM pin?
2) How are your lead lengths to/from the PCB? Thoughts about how extending them might cause problems? I know I've seen in several LED driver datasheets that additional capacitance is required across the LEDs and/or input if lead lengths are long. Thoughts?
3) Thoughts on heat dissipation? If I follow my standard practice of cramming as many as I can "fit" on to a 5cm square PCB, that means I'm working with a little over 3 square inches. The datasheet uses 1 square inch (double sided) for heatsinking when calculating power dissipation. That means I can "fit" three on this size PCB in terms of thermal limitations. But if it's not really getting THAT warm, maybe 4 wouldn't be a stretch?
4) Any chance you've cranked up the voltage to see what happens when the IC is dropping more than the bare minimum? I know that's not best practice to run it like that, but I'm wondering what it'll do.
Thanks again for being the prototyper for this chip!
der_wille_zur_macht
Team RC
The drivers with inductors are switching regulators; the inductor is there so the design will be constant conduction under the normal operating frequency of the internal switch. Since most of these switching regulators have a VERY high internal frequency compared to the frequency we'd likely PWM dim them at, our PWM dimming efforts are basically turning them completely on and off each cycle. Some of these ICs have scope readouts in their datasheet that show this.
By contrast, some switching regs (the NCP series I started with and the ZXLD series) also allow you to muck with analog dimming by altering the reference voltage - this will result in a true "analog" dimming in that the average LED current will just adjust up and down, it won't switch the chip on and off at a low enough frequency to pass the on/off pulsing through to the LEDs.
Any linear regulator (like the CAT4101) will likely be dimming on/off with the PWM signal. The datasheet for the CAT4101 shows this - there's a graph showing LED current compared to PWM signal, and while there's a tiny bit of ramp up/down it is actually shutting the current off each cycle.
Agreed, that's basically my approach. Design with a sense resistor that sets the current at the max I want for full brightness in the middle of the day.
Above you mentioned putting a pot on the Rsense pin of the CAT4101 - I'd be careful about the pot you selected if you still decide to go that route. Seems like pots are often very "noisy" and I'd hate to have a spike from the pot send the LED current above the max the LEDs can handle. When you're mucking with the sense resistor, the only "failsafe" you have is the max current shutdown for the IC - vs. when you're mucking with a pin that's designed for dimming, the sense resistor will still catch anything out of range - if that makes sense.
TheFishMan65
New member
dwzm,
I should have clarified. I will use a pot in series with a 549 ohm resistor (if I remember this is the value for 1 amp) so that I can not go higher than 1 amp drive. Or are you saying that pots are nosier over time - taday it measure 500 and tomorrow 600? From the usage I have seen they don't do that, but are only noisy while adjusting them - the 549 ohm resistor should handle overdriving the LEDs.
I should have clarified. I will use a pot in series with a 549 ohm resistor (if I remember this is the value for 1 amp) so that I can not go higher than 1 amp drive. Or are you saying that pots are nosier over time - taday it measure 500 and tomorrow 600? From the usage I have seen they don't do that, but are only noisy while adjusting them - the 549 ohm resistor should handle overdriving the LEDs.
der_wille_zur_macht
Team RC
FishMan,
"Noisy" in the sense that if you watch resistance on one while you turn it, they'll often occasionally spike up or down while you're actually turning it - not in the sense that it'll drift over time or anything. This is just something I've observed using them for different applications. If using a pot, my approach would probably be to shut the array down, adjust to the value I wanted (checking with a multimeter), and start it back up. Maybe it's just me, but messing with an adjustable Rsense while the driver is running makes me a little nervous. Not really a big deal, just thinking out loud.
csa dad, what specifically did you want cost info on? Building one of these drivers, or building an entire LED array for your tank?
"Noisy" in the sense that if you watch resistance on one while you turn it, they'll often occasionally spike up or down while you're actually turning it - not in the sense that it'll drift over time or anything. This is just something I've observed using them for different applications. If using a pot, my approach would probably be to shut the array down, adjust to the value I wanted (checking with a multimeter), and start it back up. Maybe it's just me, but messing with an adjustable Rsense while the driver is running makes me a little nervous. Not really a big deal, just thinking out loud.
csa dad, what specifically did you want cost info on? Building one of these drivers, or building an entire LED array for your tank?
