The driver will adjust accordingly to the voltage required.



The frequency is stated a little further down the datasheet, but it's 100Hz-3KHz. The circuit I have built runs at about 300Hz and works fine.

That should be easy enough. Just use my attiny13v, a transistor and voila, done

Exactly.

[quote name='evilc66' date='Jul 23 2009, 08:06 AM' post='2341548']
The driver will adjust accordingly to the voltage required.


EVIL or anyone else who knows: I just ordered two ELN-60 48D (dimming). You said above the driver will adjust voltage. so lets say i hooked up in series:

8 (CREEXRE-W107, White Cree Q5 XR-E Star) at 3.3vf = 26.4 vf
-so are you saying that the ELN-60 48Vold driver would adjust voltage down to 26.4v?

then if i hooked up in series
14 (same led as above) at 3.3vf = 46.2 vf
- the driver would adjust to 46.2.. and since they are in series, would the total amps still be 1000mA?

please help here, a bit confused. I really want to hook up 24 cool whites and 24 royal blues to 2 drivers, how would you recommend i do this with these driver...

Yeah it's pretty much as you described. The driver will always provide 1000mA (after you've tuned it down from the default 1300mA), and adjust voltage as needed to provide the constant current, up to max 48v.

Works the same in parallel, if you had 2 series of 6 LEDs in parallel (22v-ish), then each of them would be getting half the current it would be getting if it is was just one series.

Thanks Keli,

Very helpful. i also read about a current mirror, so if an LED went out in a string, the other leds in parallel would not get overloaded with current. DO you know about this as well?

Was playing with my drivers today, seems the P model can be analog dimmed to a point.

What I did was hooked a 9v battery and a external Pot to the dimming circuit. Adjusted the internal Pot to 1000ma when the external is all the way up. With a 10K Pot the LED's will dim down to 210ma and with a 100K Pot they will dim down to 140ma then drop off about half of the pot range.

Anybody else try this, have any thoughts on it.

Yes. We have had some disussion on this topic.


I had tried that when I got my "P". I got about the same results that you did. It wasn't very controllable, but it did work to a degree. Maybe it was just my driver. This was the first one that was hand built by Meanwell, so it could have been faulty.

Not just your driver, same results here. Seems dimmable to maybe 50% before it just shuts off by varying voltage

Ok. My original driver was not working with the pwm cicuit, so it was possible that the analog created problems too.

hey evil, is there is driver, meanwell perhaps, that would drive more LEDs than 13? i'm tying to help out some friends design a big LED array. thanks!

Drivers like the ELN-60-27 can run a lot of LED, but they have to be in parallel strings. The ELN-60-48 runs the most in series right now.

i just got the meanwell p style from the group buy (thanks evil) i am still a little confused is it possible to run this without anything connected to the dimming wires will the leds light at all

thanks,mike

Nope, I'm pretty sure you need 10v to connect the control lines to get light. Nothing connected means 0% light. (As opposed to the buckpucks where 0% power on control = 100% light, and 100% power on control = 0% light)

The P version is not designed to dim using analog input, and I think the color temp will change by analog dimming. PWM will keep the color temp constant while reducing the intensity. Not sure if the D type keeps constant color temp when dimming.

Doesn't work that way. The driver is just interpreting the analog voltage as a duty cycle based on voltage averages. You can't change the color temp of an LED unless you force it to run at certain extreme voltages that aren't safe. This is not capable with a driver like this where the voltage is controlled based on demand. The dimming circuit does not directly control the output like a linear driver would, so the input method will not affect how the output works.

thx evil for clarifying. I forgot you can treat the analog input as 100% duty cycle. Maybe that's the reason you won't be able to do 18%-100% dimming (according to the spec). Still surprised at that 50% dimming, 100% duty cycle should give u max output all the time though. I have the latest Arduino PCB, planning to try that on the P.

Can any electronic gurus explain the example schematics at the end of this doc:
http://www.meanwelldirect.co.uk/public/ran...7%5Cr1107_3.pdf

In particular why the parallel strings of LEDs seem to be interconnected in a grid layout (does this stop thermal runaway?) and also why one configuration has resistors built inline with the strings (is this to allow for a variable number of strings?).

Not too sure how Arduino would do the full dimming since it has 5Volt max ref for dimming and ELN-60-48P are 0-10Volt for full dimming needs. You may only be able to dim your LEDs via Ardunio halfway not 100%. I could be wrong but this is what I been told by a senior member on Ardunio forum. I am sure with few changes it would be possible to run Arduino with ELN-P versions.

It's not to prevent thermal runaway, but to prevent additional failure if an LED dies. It only works in strings of three or more, but it helps to equalize the change in load over a greater area instead of a dead LED taking a string out and affecting the remaining strings. It's not foolproof (doesn't work so well when two LEDs die), but it works. Only downside is the massive increase in wiring.



You would need to switch the pwm input with a transistor connected to a 10v source.

Many thanks Evil. Are you able to explain how it does this in a bit more detail? Also, why it has the resistors in the right hand example?

Would this be a "safe" way to wire 4 strings of LEDs in parallel off an ELN-60-24 or will a thermal runaway still most likely occur?

Basically what happens is that if an LED fails open, there is a path around the LED to equalize voltage and current. When running three strings or more, the failure of a single LED can be handled pretty easily. More than one failure can become problematic as the array is trying to compensate too much. I can't see any good reason to use a resistor at the begining of each string unless the forward voltages of each LED were wildly varied and needed something to limit the current.

Thermal runaway is a failure event due to poor current regulation, like when you use a current limiting resistor.

Hey evil any updates on the 556 pwm circuit for dimming these. My array is almost done and could use some dimming.

In my "Where do I begin" LED thread, Evil mentioned it should be done this week.

http://www.nano-reef.com/forums/index.php?...t&p=2349003

Ive lost track on this whole dimming thing since the last time i posted... Please refresh me quickly.

1. The P models that we got through the GB are able to be dimmed by a LM317 circuit?

2. The 556 circuit is a PWM circuit that can also be used to dim the LEDs?

3. The LM317 circuit is adequate enough to dim the P meanwells that we have?




On a side note... Im tired of waiting for Digital Aquatics to release the update to use the ALC with the RKL controller. i purchased the ALC months ago thinking it wouldnt be long for the update to be released. Its really annoying that it was promised and the promised date has come and gone.



I just called powergate and they said they cant sell to you unless you have a company?

It will come, don't worry

The "P" version needs the 555/556 pwm circuit to dim. The "D" model needs the LM317/pot to dim. To use the "P" model on the ALC, you will need a slightly different circuit to convert 0-10v from the ALC to a pwm signal. I will have a circuit for you for that also, but a little later on. The 555/556 pwm circuit will be later this week once I get a chance to finish it (sorry Deep )

I know the fealing. I am harrassing them both on these forums and theirs. It will be out soon, hopefully by my birthday in 19 days.

You mean you can't just plug the ALC straight to the meanwells?

You can plug them straight into the "D"s, but the "P"s will need a little conversion circuit. It's actually easier to build than the pwm circuit.

then that your next project! haha

I have been running 2 drivers like this for a few days now and I am happy with the results. I also tested it on 2 other drivers and all drivers dimmed smoothly. I also tried hooking up a 5v power supply and it worked about the same but could only dim down to 220ma before it dropped off.

So u hooked up the 5V DC and it gives u the full intensity when the pot is at no dimming position? If that's the case, a 5V PWM should be able to dim it fully, maybe the Arduino board will work on the P.

Does anyone know if suffix P or D will work with the Reefkeeper ALM and how does it hook up to the ALM, the 0-10v inputs? If its that then how does the suffix P work because from reading it is PWM signal? All this electrical stuff is giving me headaches, thanks.

Jason

The "D" can be connected directly to the ALC. Just hook the DIM+ and DIM- wires to the connector on the ALC. The "P" model can also work, but with an additional converter circuit in between. It's simple to build and I will be posting details of the circuit soon.

Thanks evil, that was some fast reponse. What is the difference between PWM and analog 0-10v, is one better or is it about equal and what is your preference? How many LEDs do you recommend for a 55 gallon standard tank im planning to seperate the LEDS on 2 arrays and when I upgrade to a bigger tank I will just make add another array. Oh yeah sorry for such a long post but what do you guys think between MH vs LED because I was bored at work and was calculating the expense of bulds and cost of new MH setup vs LED for 10yrs and it seems like the difference isnt that much.

I have uploaded a vid of my Cree MC-E (752 lumen) LED being driven by the Meanwell ELN-60-24D @ 2.3A.

The dimming is done via the bench power supply in the bottom right corner.

What isn't obvious is that at 0V there is still some light emitted. I turned off the auto-exposure on the cam. The auto-exposure made the dimming effect hard to see.

http://www.youtube.com/watch?v=XOszrRRy__Y

It's two different methods for doing the exact same thing. 0-10v is a typical method for dimming residential and commercial lighting. Pwm is common for automated controls like what we do with Arduino microcontrollers.

You could light a 55g with about 60 LEDs (20x3) and 60 degree lenses. This will get you equivalent to about a 2x250W MH setup. Comparing power consumption, you are looking at ~222W vs. 500W. With an average of 10 hours of light per day at $0.15 per KWH, you are looking at $121.55 vs. $273.75 to run them for a year. Next, you will save $120-$200 per year on bulb replacements. You also potentially save on not running a chiller in both energy costs and the cost of the chiller itself. So even if you you don't factor in the chiller savings, you have potential costs to run the MH setup of $2273.75 for 10 years. The LED setup would cost $1215.50 to run over 10 years. Fixture costs can vary wildly depending on what you pick. You can go from the super cheap MH setups of about $300 to the high end $1000 setups. The LED fixture will cost about $750 to build, so you sit right in the middle of the MH fixture costs. The savings are there. It's not as dramatic as some companies drum it up to be, but when you factor in many of the features, it's still a great option.

Thanks evil, I never thought about that. I think it was the initial cost of LED thats made me think the savings would not be that big until later on in the years. Actually I was thinking about using 60 LEDs and about a 50:50 split between white and blue. I am abit confused about the meanwell drivers though, does that mean I need like 5 meanwells with 12 LEDs/driver? Like I know with the specs of the whites are 1000mah and the blues are like 700mah? How would I run the blues on the meanwells when the lowest I can adjust it is down to about 1000mah? Im not too familar with the difference between running LEDs in series and parallel.

if you get the crees, the blues and whites both run at 1000mA. the meanwells that we use can power upto 13 LEDs in series. so yes 5 drivers would do it. if you wanted to do a dawn/dusk effect or keep the colors separate so you can controll you color, you would need 6 drivers.

Thanks, can anyone explain to me what is the dfference betweetn running LEDs in series and parallel and what it does to volt and current ratings of a driver but in particular to the meanwells.

when you wire in series, the current is constant (amp) and the voltage varies or the voltage gets divided up by the number of objects.....that is why we use the constant current drivers

when you wire in parrallel. the current varies or in other words get divided by the number LEDs/strings and the voltage is constant. when using this type of wiring, you need a current mirror circuit to protect your LEDs. the current mirror protects your LEDs by dispersing the energy through out an array when one LED fails. without it you can burn up all of the LEDs if one decides to fail. very risky and lots and lots of extra wiring.

stick with series wiring for the LEDs, if one burns out, the driver will adjust for the remaing LEDs and no other damage to the good LEDs will occur, and a lot simpler wirind. to wire in series, you would start with the positive lead from your driver and connected to the positive on the LED. then from the negative of that LED to the positive of the other LED, then negative from that one to the positive of the next one. the repeats itself until the last one then the negative is connected to the negative of the driver to complete the loop

Thanks for the detail explaination, it was very helpful because im still in the planning stages of my project and looking to get everything down before I go buying things. Does anyone have any idea of any buckpucks that have 0-10v dimming?

according to the specs, the buckpucks will take 0-10v dimming from a controller.

So, what's the trick to opening one of these up? I got mine from the group buy and was going to trim the internal pot down but was having troubles busting it up without breaking anything

If the clips are stuck try shifting the top to one side or the other.

Don't know if that helped but I had one that did not want to come apart so I had to pry at it and when the top shifted it came apart.

0-5 or 0-10? I was under the impression that it was 0-5... if it's 0-10 that sure changes my plans.

I think the meanwell drivers are 0-10v dimming. I am not sure but i think regular buckpucks are 0-5v or are there other buckpucks besides the meanwells that have 0-10v dimming?

Buckpucks will not successfully take 0-10v for dimming. Luxdrive does not recommend this at all (I had posted this in another thread specifically about this).

Hey evil will you do another led, driver, optics GB anytime soon?

(Man, this gets asked a lot) Maybe, but not until the end of the year.