when mine is hot to touch (like 3 seconds and then ouch!) its about 45-50 degrees celcius.. that is still safe just on the warmer side..

Mine should be close to that, 3 seconds and then ouch. but don't have a clue about temp. Maybe I am worried cause my other fixture was over cooled. I had a 24V noisy fan that pushed a ton of air over that 5"x5" heatsink.

Another question, Do I have to put the 10V power supply connected to the DIM + - on meanwell in the same timer as the lights or I can leave it plugged directly always even with the LEDs Off?

I have experienced it and itrs been said before that it doesnt take much air to keep it cool.. if your fan facing the right way (blowing onto heatsink) i have my 12V fan on about 6-7V i think.. cant hear it at all, and keeps the heatsink from a passive temp of 53 down to about 32...

About a volt should be fine, but the final measurement should be taken once the circuit is hooked to the driver.



Just think of the meter as another LED and wire it as such. You won't cook the meter hooking it backwards.


As long as the solder isn't touching expsed aluminum, you are ok.


Without measuring the die temperatures, you won't really know what is safe. Is it running that hot with the fan, or without?

I'ts running that hot with the fan on. I just though maybe the thermometer probe could give an idea of how hot is. I put it in the middle of the heatsink fins, right under the the fan. Its constant at 110F

110F is fine. That shouldn't be much more than uncomfortable to the touch. I wouldn't consider that to be hot. Normally, I like to keep heatsink temps under 140F.

Oh, good to know. I was about to panic. Then I was overcooling my other heatsink cause it never got even hot. Not even half this one.

I've been looking into acrylic splash guard lately. What thickness is good to with ? Is it even a must-have ?

As far as thickness you should be fine with 1/8"

Hey, yall.

I put together the LM317 circuit using a 150mA, 12V power supply.

For the output voltage (no load, only mulitmeter hooked up to dim- and dim+) I am getting 8.2 to 16.5 from the pot, which I know is not within spec.

I've looked at my circuit board and cannot see what I've done differently than the one Artnsx posted--

Any ideas where to begin here figuring this out?



Also, the LM317 was getting pretty warm after about 5 minutes. The resistors did not, only the LM317. Uncomfortable to the touch-- is this normal?

This is not normal. Your voltages are way off, and you should have no heat in the LM317. Sounds like something isn't connected correctly. Let me see if I can get the diagram done at lunch.

OK, thanks.


I noticed that the resistors I have are a smaller gauge wire than what Artnsx posted. Mine are 1/8 watt resistors, are these sufficient?


As a side note, the 12v power supply, with no load, is giving me 19.9v. I've verified this with two different multimeters. Could this have anything to do with this?


I have double checked the back of the circuit board to make sure I didn't have any solder joints that were touching (where the LM317 is). It's such close quarters that it's hard to see. I used a razor blade to clean out in between them just in case, and this had no effect.

Many thanks, Evil!

Here is the image I reproduced, provided by Artnsx.




As a side note, for the input voltage on the LM317, should that be the - or the + lead from the 12v ps? Artnsx says in the pic that it's +, but I thought (if I recall correctly from physics in college) that input on DC was -

Always DC+ in.

The resistors are fine, provided they are the right resistance value. Every resistor has a deviation from that that value, but not enough to skew the outputs that bad. The voltage of the power supply shouldn't make any difference. Anywhere from 12-32v should be fine.

When my lamp turns off you can see the LEDs still on very very little like dimmed. Not enough to work as moonlights though but they still very low. Why is this happening? Is it normal?

I removed everything and resoldered it on the perf board, in case I had accidentally bridged between the solder points (causing a short).

It came out somewhat better--- I'm getting (with no load) 2.5 to 14.5v

The LM317 still warms up. Should this stay cool the whole time or should it generate something at all?

Is that with the AC off and the dimmer still on? The "P"s don't seem to have that issue, but I haven't played with the "D"s yet.


Better, still high. I'm going to have to look into this. You are using 1K and 470 ohm resistors for R1 right? And a 10K pot?

Yes, I verified all three.

Do you have any tips on soldering the backs of these perf boards? I am using the finest tip I have available to me, but it's just too easy to puddle up too much and risk overlapping. They don't visually overlap, but they are oh-so-close together.

Nope. Both the meanwell and the 10v psu are in the same timer. I don't think is a grounding problem or something like that cause when I test run everything out of the heatsink (LEDs over a wood board) it happened the same.

Smaller diameter solder can help. Also, not feeding as much in helps too


Hmm. Does it only stay dimmly lit for a short period of time? Could just be internal capacitance.

I guess I'll try resoldering it again, but honestly I don't think there's a short between the solder joints on the LM317. I used a blade to clean out between them best I can.



Couple of questions on the Meanwell.

--SRV2--- turn this all the way CLOCKWISE for minimum setting (around 975 mA right?)




We tried to test this out this afternoon, with SRV2 turned all the way clockwise. I used the multimeter to make SURE that my dimmer circuit/pot was putting out around 7.5 volts (random choice). I assume this should have been around 750 mA or so.


When we connected the LED to the energized Meanwell, it blew. Bright spark inside the LED and a "click".

What would you say we did wrong here? I was under the assumption that the Meanwell automatically detected and adjusted, so perhaps 1 LED was just too little of a load and it hit the thing with a full 48?

We did verify with the multimeter that, at no load, was putting out 48.2v. We didn't try to check the current.

Not for a short period. That was what i was hoping. Last night was my first night with the new setup and this morning they were still the same

You now have a dead LED. You should never connect an LED to a live driver. Also, you should be turning the SVR2 pot counter clockwise to drop it to it's minimum.

The reason you killed it is because it basically dumped all the stored energy that was built up in the driver into the LED. Two things you should never do: connect an LED to a powered driver, and remove and LED from a powered driver. Two good ways to kill things.

EXTREMELY IMPORTANT INFORMATION:

The Meanwell ELN-60-48 has a limitation that I had not noticed before, that seems to be causing some people some headaches. It will require a minimum of 6 LEDs connected to regulate properly. I know I have said that one LED should be fine for testing, but it seems that this is a recipe for disaster. I know this message is a little late for some, but for those looking to use these, please pay attention to this.

Sorry for any inconvenience that this has caused people.




RedfishSC, your problem is that you connected to a live driver, but this is something to pay attention to from this point on.

Ah. Question totally answered. THANKS!

Considering that I had ramped the internal pot all the way UP by mistake, no freakin wonder it blew the LED.

I only roasted one LED. $8 in education. So now I'm down to 7.... can't fry many more than that, lol.


Evil, does the picture that Artnsx posted look right, for the dimmer board?

I am going to pick up a new LM317 (one that I haven't trimmed short yet) and spread the two furthest prongs out to space the circuit out a bit to give me more room for soldering. Any reason that's not a good idea?

It wasn't the pot so much that killed the LED. It a difference in potential, and how it changes as a load is connected, especially considering the inductor in the driver. Basically, the magnetic field in the inductor collapsed and dropped a bomb on the LED. Most likely the LED saw a brief current spike of a few amps. It would have done that regardless of how the pots or the dimmer were set.

Art's picture looks fine. I'm going to build one up today just so I can start testing to see where potential issues are. This is a stupic simple circuit, and I can't believe it's creating so many issues.

So not unlike a big capacitor dumping a huge charge, relatively speaking, right? Makes sense.




Very cool. The problem here is the "stupid simple" part. The assumption here is that the person building it is, at the minimum, stupid. That may be my problem

Hey guys,

I'm about to build my lighting set up and have a couple simple questions. What is the way to go : soldering and wiring the leds together and then place them onto the heatsink or the other way around (place them and then wiring) ?

Also, what kind of wires should I use : stranded or solid wire ?

Thanks !

Its easier to place the LEDs in place then solder. Its a little more difficult due to the heatsink removing heat from the PCB so you will need at least a 40W soldering iron. You can use solid or stranded.... i prefer stranded cause its easier to work with, more flexible. For wiring the LEDs you can use 22gauge wire but should use at least 18 from the drivers to the LEDs

Damn, really? I've ordered 22 AWG for all the wiring. MPJA power supplies to buckpucks. Buckpucks to LEDs. LEDs to LEDs.

If the drivers are close to the LEDs, 22awg will be fine. 18awg is important for long runs (greater than 3ft) so the voltage drop is minimized. 18awg wire is easy to get at hardware stores though.

I'm going to have about 4 feet total from driver to LED's. I have 20awg, will that be sufficient, or should I make a trip for some 18?

20awg is fine.

LEDs arrived from Nanotuners just now. First thing to arrive. HIGHLY recommend them!!!!
Can't wait to get the heatsink to begin the project. (oh, and Nick answered an Email question on a Saturday. Most excellent service!)

Michael

Evil, did you have a chance to wire up a perf board for 10v? If so, what results did you get?

I picked up a new LM317, and I'm going to spread the prongs out a little further to give me more clearance between solder joints.

Didn't get a chance to hit Radioshack today. Gonna get down there tomorrow.

Hi evilc66,

I have being using it for more than 6 months but have not play around the SVR thingy.

Am I right to said (Voltage) SVR1 and (Current) SVR2, both set to clockwise for minimum and anti-clockwise to Maximum?

I have a lots of wires of different sizes but none of them mention any kind of size measurement. How could I know what AWG these are ?

Also, I've tried to hook up a 12V fan to the LM317 circuit, and using a multimeter, it says there is 10.17V. I don't know if it's a good thing considering that I want to hook it up to the DIM+ of my MeanWell driver (that takes normally up to 10V). What do you guys think about it ?

I wish I could get 10.17v out of my rig. The lowest I can get is 14.5

You can set SVR2 COUNTER CLOCKWISE to the minimum, but SVR1 will do nothing.



Sounds damn near perfect.

I've purchased the thermal pads that nanotuners have. Is there a specific procedure to follow to make it very sticky or what ?

Hi,
I think I am going to try to build my own LED light fixture, instead of buying a ready product, (thank god my wife does not read these forums )

what type of a setup would I need for a 24"x24"x24" cube,
would a 24 Cool white Cree XR-E, and 24 Royal Blue Cree XR-E setup be enough, or can I do with less.

do I need lenses ?
If so what degree would I need.

Thanks,
Yasar

Just make sure your heatsink and the back of the PCB's are clean. I buffed my heatsink to a literal mirror finish using a buffing wheel w/tripoli compound (jeweler's rouge) and another wheel with white diamond compound... but only b/c I had the equipment handy.


I cleaned the PCB's and heatsink with acetone since that's what I had, but anything like that would work--- denatured alcohol, MEK, even laquer thinner (wear gloves if you use this).

Other than that, the ones I have are VERY strong.

The pyromaniac in me was wondering the other day what would happen if the AC input polarity was switched when a Meanwell was wired up.

I have mine wired appropriately (the wide blade of plug = blue wire on Meanwell = neutral), but wondered what would happen if someone swapped them by accident, or if the house's plug outlet was wired wrong (ie, someone's DIY maddness on an electrical outlet swapping poles). I've seen this happen in apartments with maintenance guys who have had a few too many vodka doses.

AC, by definition, doesn't have polarity. So there is no problem with swapping the leads. The reason for maintaining the neutral on the wide blade is a bunch of National Electrical Code mumbo jumbo. And so that the wall switch will deenergize the light when you turn it off. If you hook it up backwards, you will still have voltage all the way to the load, so if the load gets grounded, you won't be able to turn it off.

I followed you all the way up to that last point. I don't know what you mean by "if the load gets grounded, you won't be able to turn it off".

How would this get grounded--- you mean like a short (ie, I become the ground wire and get zapped)?

A good example of how it gets grounded is when you touch it and scream, "It shocked me!". It could short (or ground, means the same thing) out to the case, and make the circuit complete. There should always be a case ground or else you're begging to get shocked. The electricity needs a path, but it doesn't care where.

Ok, I see.

As far as grounding the whole thing, I think Evil recommended against grounding the whole unit at some point. I can't remember why. I may be wrong..... I'm used to that by now

Redid the whole board. Still getting 2.4 to 14.8v (no load). Verified with two different multimeters.





I purchased a new LM317 so I can spread the prongs apart for more soldering room so I can be sure of avoiding a bridge in the soldering. That worked like a charm. Absolutely no solder bridging between the connections of the 3 prongs.



Also, bought a new 10K ohm pot. The other pot, for some reason if you bumped it, would give you a voltage jump (from 14 to nearly 20). This new pot does not do it.


Still using a 470 ohm and a 1K ohm resistor. Verified their resistance with the multimeter, got some variance. Getting 462, 980, and 10.01K. According to the calculators, that should be around 12.5v or so. That's high, but not as high as I'm getting.


QUESTIONS:

Is there a difference between an LM317 and LM317T? The "T" was all they had at Radio Shack, no slot for anything but that.


Last, is the attached, cheezy Paint pic what I'm shooting for? This is what I have wired up.

Hey Evil, I got to page 10 and got tired of reading.

I'm planning on lighting a Biocube 8 with the list below. I don't intend to keep too much that requires a lot of light - maybe a ways down the road. What I'm curious about is what light increase I'd see at full brightness over the stock 36W combination? Care to take a guess?

4x Cree XR-E 10,000k
4x Cree XR-E Royal Blue

No optics planned.