DIY LEDs - The write-up - (split again)

[2Quills]

Hi all,

Hope I am not hijaking here. was advised this is a place we can post ?

I bought 4 of the meanwell HLG-240-48b drivers and I finally have 2 of them wired up to my LED array. I have 1 running 6 strings of 11 (66) XP-G cool whites and the other driver running 5 strings of 12 XP-E royal blues (60). This is half my led array. I have the reefled dimming kit (2 pots in the case). I have these wired up and when I put a meter on the pot outputs, I can see the volts go from 1 - 10, I also see the voltage change on the wire that I am running to the blue (+) wire of the driver. When I turn the pots up and down, the LEDs get brighter and dimmer but my concern is that each string is running about 600 to 1000 mA and I can not get them to turn down any lower than 600. I know I need to balance my strings but my concern is that these do not dial down to the 200 or 300 mA range. I do not think I will ever want to run these up in the 800 - 1000 mA range. Can anyone throw me any suggestions on why they will not dim below 600mA?

I have my driver leads running to wiring blocks. wire in , fuse, 1ohm 3 watt resistor and then out to the first LED of the string. I measured vf across the resistors and they are all over the board, 600, 800, 800+, etc. I know I need to start swapping some LEDs to get the strings balanced but it just seems odd that even when these are dialed all the way down, the vf across the resitors is never lower than 600 mA.

I have not measured from first LED to last LED and I have not measured the vf of each LED yet. I did not want to let them run that high and that out of balaance for long, plus it is almost impossible to work on them at 600 mA. I could post some pics of the wiring and such, just wondering if there may be something I missed inthe build. internal pot on the driver, etc.?

Here is the basis for my parallel wiring layout.

These drivers (the 240's) have caused a few headaches for some folks. There was a change to the hardware made by the manufacturer last summer. If you don't have the ones made after July 13 I believe it was then they're not going to be compatible with 1-10v or modulated signal.

Can you post your serial numbers on the drivers? That will tell us when yours were manufactured.

I just have a friend go through this very same thing.

 

[zachts]

Anyone on else on here think this circuit might be useful for large arrays such as the one reeferdale is decribing now having to go thru and ballance? Maybe not good for a first timer, but after someone has managed to soder 100+ LEDs I think it is within the realm of possibility. I have it matching two otherwise intentionally very unballanced strings and it works beautifully!

http://www.reefcentral.com/forums/showthread.php?t=2097086&highlight=led+current+mirror

so far my experimentting seems to make it much easier and quicker to build a few of these rather than have to mess with trying to ballance your strings manually. Seems to work great thus far in my little experiment.

Any thoughts folks?

 

[kcress]

Yeah Zachts there are a couple of reasons those mirrors are less than desirable. They're essentially just variable resistors. The excess is going to be dumped on the transistor. That's wasted energy.

How does the whole mirror scheme work with six or seven strings? Who's mirroring who?

A really large issue is implementation. Where's all this electronics paraphernalia going to physically reside? Without this stuff you pretty much have a simple, durable, electrical assembly that's wires and screws and terminal blocks. That mirror stuff is essentially "electronics" stuff and short of making circuit boards of some sort, there's no practical way of implementing them.

 

[reeferdale]

http://www.reefcentral.com/forums/showthread.php?p=19811732&posted=1#post19811732

These drivers (the 240's) have caused a few headaches for some folks. There was a change to the hardware made by the manufacturer last summer. If you don't have the ones made after July 13 I believe it was then they're not going to be compatible with 1-10v or modulated signal.

Can you post your serial numbers on the drivers? That will tell us when yours were manufactured.

I just have a friend go through this very same thing.

thanks, These are able to dim, they just don't dim very much. I can turn them up and down but I would expect the vf to drop well below 300 when the pot is all the way down. I am eventually goiong to put these on my Apex but for now just to get everything running and transfer tanks I want manually dim them.

here are the serials Kinda odd that each 2 are close to each other.
RB18073709 & RB18073711 RB02224635 & RB02224662

If you have info I appreciate it.

I accidentally have this post in 2 different theads

The original thread I started is
http://www.reefcentral.com/forums/showthread.php?p=19811732&posted=1#post19811732

 

[kcress]

Can anyone with
I've never used an HGLxx-xxB type drivers. Can anyone using them tell us what theirs turns down to?

 

[reeferdale]

Anyone on else on here think this circuit might be useful for large arrays such as the one reeferdale is decribing now having to go thru and ballance? Maybe not good for a first timer, but after someone has managed to soder 100+ LEDs I think it is within the realm of possibility. I have it matching two otherwise intentionally very unballanced strings and it works beautifully!

http://www.reefcentral.com/forums/showthread.php?t=2097086&highlight=led+current+mirror

so far my experimentting seems to make it much easier and quicker to build a few of these rather than have to mess with trying to ballance your strings manually. Seems to work great thus far in my little experiment.

Any thoughts folks?

I looked at this, sounds right but was way over my head! I think I have to balance the old fashioned way.:crazy1:

 

[2Quills]

http://www.reefcentral.com/forums/showthread.php?p=19811732&posted=1#post19811732

thanks, These are able to dim, they just don't dim very much. I can turn them up and down but I would expect the vf to drop well below 300 when the pot is all the way down. I am eventually goiong to put these on my Apex but for now just to get everything running and transfer tanks I want manually dim them.

here are the serials Kinda odd that each 2 are close to each other.
RB18073709 & RB18073711 RB02224635 & RB02224662

If you have info I appreciate it.

I accidentally have this post in 2 different theads

The original thread I started is
http://www.reefcentral.com/forums/showthread.php?p=19811732&posted=1#post19811732

RB18073709 & RB18073711 These were manufactured in August 2011 they should work for what you want to do.

RB02224635 & RB02224662 These were manufactured in Febuary 1010 and will not work for what you want to do. They are standard resistance dimming.

Where did you get them if I may ask?

 

[zachts]

Yeah Zachts there are a couple of reasons those mirrors are less than desirable. They're essentially just variable resistors. The excess is going to be dumped on the transistor. That's wasted energy.

How does the whole mirror scheme work with six or seven strings? Who's mirroring who?

A really large issue is implementation. Where's all this electronics paraphernalia going to physically reside? Without this stuff you pretty much have a simple, durable, electrical assembly that's wires and screws and terminal blocks. That mirror stuff is essentially "electronics" stuff and short of making circuit boards of some sort, there's no practical way of implementing them.

those don't really seem like that big of issues to me. energy wise were talking a pretty miniscule amount of wasted energy, they barely get warm to the touch. think of all the "wasted" energy for those using metal halide, :lol: I guess if your main goal is über energy efficient lighting then this would not be the route for you, but even some of the most inefficient LED builds possible still use half the power of a MH setup.

I should point out I'm not refereing to the more complex circuit at the top of the thread I linked to that has built in protection elements. I'm just talking about using two transistors along with the fuses and 1ohm resistors we already would be using.

As far as implementation that doesn't really seem that hard, we have to locate our drivers somewhere don't we? and then there are those pesky controllers, many of which we have built from scratch. That's alot of "electronics" that also need to go somewhere Just look at the pic of my test array I posted on the other thread. It's all just wires, and screws, and terminal blocks

For multiple strings, just pick one and use it as the reference and connect it to all the base legs of the transistors. I don't see any reason it would not work just the same as with two strings.(though I have not YET tried this) All the transistors would just need to be close together preferably back to back on the same heatsink. (possibly on one of our heatsink fins? or just a scrap of aluminum left over from the build) they could be neatly arranged along with your fuses and resistors wherever you put those. If you were going to make circuit boards even better, many people have done that to make drivers so why not for these?

I just wan't to have a fair disscussion of the pros and cons. This wont be for everyone but for some it may be just the ticket. Me, I'm inherently lazy and only like to do things once. If I can do a bit of extra work up front and never have to worry about strings getting un ballanced with age and heat, I'm all for it.

 

[reeferdale]

Can anyone with
I've never used an HGLxx-xxB type drivers. Can anyone using them tell us what theirs turns down to?

kcress, I actually emailed Bill at reefledlights and his reply was
". . . They are rated down to 10% at 1 vdc."

I am going to wire up these same two drivers tomorrow night and take new reading. I will get them a low, mid and almost full POT (1v, 5v, 8v). I will take vf readings for each. Wiring these POTS to the driver is pretty straight forward there is a + & -, the POT feeds Voltage to the blue + wire. I am hoping that now that they are in the frame and the wiring and drivers will all be wired up snug, maybe I just over looked something with the temporary wiring, it was a LOT of wires running in and out.

Plus, I have 2 other drivers I can try in place of the ones I tried initially? I think I may be a day or 2 premature in posting this and asking for help. Let me get the actual vf numbers at different dimmer voltage and then I will be able to post some pics and discuss any issues with a repeaatable set of standards.

thanks

 

[TheFishMan65]

reeferdale,

You pointed out what you did with the DIM + wire, but where is the DIM - wire connected? If not connected this could be your problem.

 

[reeferdale]

reeferdale,

You pointed out what you did with the DIM + wire, but where is the DIM - wire connected? If not connected this could be your problem.

The DIM - wire is connected the to correct lead on the POT. I guess I should have made that clear, I was just trying to point out that I can 'see' the voltage change on the drivers DIM+ wire so I know that the signal being input is being controlled by the POT properly. I wish it would have been this easy. I am going to throw some pics up tomorrow

 

[reeferdale]

xxxxxxxxxxx

 

[2Quills]

I would double check the serial numbers on the drivers youve got hooked. The two newer ones should work but the problems you are describing sounds like youve got the two older ones hooked up. If thats the case then id try swapping them out and test everything again.

 

[reeferdale]

I would double check the serial numbers on the drivers youve got hooked. The two newer ones should work but the problems you are describing sounds like youve got the two older ones hooked up. If thats the case then id try swapping them out and test everything again.

I will try that but I am 95% sure I happened to have had 1 of each serial# range hooked up. I am going to do the wiring for 1/2 the array tonight and I will use the to drivers with the higher serioal numbers and psot results with some pics
thanks

 

[lowfi]

Hi guys what self tapping bit do you use for mounting the led's?? Also need the screw size...should I forgo all of this drilling and tapping and just go with the solderless?

Thanks for the help!!

 

[gguertin]

I'm looking at the idea of removing my mh's that has small spider reflectors and replacing it with cree leds. I have just enough room for a 1.1" wide heat sink like one of Rapid leds. I have a 4" fan on each end of the fixture. If I use there good 4 fin heat sink with fans for cooling. How close together can I space them.
Thanks

Sent from mobile

Mine run pretty cool so I would think you would be ok with a very close spacing if you are going to run forced air over the leds.

 

[brassmonkyballs]

Hi guys what self tapping bit do you use for mounting the led's?? Also need the screw size...should I forgo all of this drilling and tapping and just go with the solderless?

Thanks for the help!!

#40 drill bit and #4 x 3/8 sheet metal screws will screw in that hole. Always screw for future maintenance and customization purposes.

 

[lowfi]

Will do thanks. I've got an 11" opening at the top of my tank so I'm thinking about installing the LED's on 2" aluminum u channel. I'm thinking 4 runs at 44" long or so. Still deciding whether or not to go bridgelux or cree...its a really tough decision for me as I'm not an electronics guru at ALL. It's hard enough for me to keep up with all the different types of cree models on the market atm. I've been reading about how some of the crees can be run at higher levels...so you might not need as many as originally thought. I'm thinking about running 72 on my 90, with the front supports angled in towards the live rock. This stuff is really exciting. Thanks for all the wonderful posts everyone!

 

[TheFishMan65]

I used 4/40 with 3/32 bit. Don't forget the nylon washers.

 

[brassmonkyballs]

I used 4/40 with 3/32 bit. Don't forget the nylon washers.

+1 on the nylon washers! Fish I too drilled and tapped my first fixture....and called BS! Oh man too much work when doing a large fixture. Thats when I picked up some #4 screw and starting drilling holes and found the #40 bit but makes a nice hole and the #4 screw self taps into it! Much easier.