Building a Home Made LED

[DrFever]

hey guys wad up have you ever seen one of these 3 boxes somehow miraculously fell off a truck in front of my house )
spinning LED ????




PS: just kidding but this would be kinda cool way to go wondering Vuk somehow you make yours spinning ????
using a used ceiling mounted fan show the boys you can do it

 

[vukman]

hey guys wad up have you ever seen one of these 3 boxes somehow miraculously fell off a truck in front of my house )
spinning LED ????




PS: just kidding but this would be kinda cool way to go wondering Vuk somehow you make yours spinning ????
using a used ceiling mounted fan show the boys you can do it

ahahahh, that's freaking hilarious bro!!! LOL...fell off the back of a truck!!..I used to use that saying a lot in my youth... Shit was always falling off of trucks back then...heheh <wink wink>

What are you doing with that thing bro?? It looks like you've got it set up in a tent or something? You're not actually trying LED's are you!??! Guess all this talk about them has you thinking and wondering, huh?

That thing does look damn neat though...........

 

[rives]

PF, I completely agree that there are many, many ways to get where we want. While it has been shown that plants can grow relatively well with just red/blue light, I think that it would be pretty damn silly to think that they evolved for millennia under broad-spectrum light and only utilize those two specific components. Besides, it is so much easier for me to tell what is going on with the girls under white light. Our vision obviously didn't evolve under a red/blue mix!

 

[repuk]

Hi!

been veeeery busy these days, plants from seed have almost 6 weeks, needed to setup everything to flip to 12/12.

Lots of changes at once, something not advisable as if anything goes wrong there will be lots of variables to check, but... had no other chance.

More details in my grow log.

I tested all the LEDs, tinned the MCPCBs, mark their placement on the heatsinks, placed them w/o gluing yet to check the MCPCB "orientation" to ease wiring, then glued them and finish the wiring, tested for shorts with a multimeter, checked visually three times that the wiring was right, then hooked each string to its driver and lit them.

I'm lazy specially when it come to screws, so I used Artic's Alumina thermal Epoxy to glue the MCPSBs to the heatsinks.

I put an inline fuse in the postivie rail with each of the strings: 1,5A for the XP-G's and 3A for the XM-Ls.

Mad brightness, had to be careful not to look ever directly to the lit chips.

The heatsinks get hot, but you can touch them and hold the finger there.

I have put them 15" (40cm) above the canopy. Still in 18/6, I intend to flip to 12/12 by mid week, still have to repot one clone and prop its screen.

Everything in place:

 

[PetFlora]

PF, I completely agree that there are many, many ways to get where we want. While it has been shown that plants can grow relatively well with just red/blue light, I think that it would be pretty damn silly to think that they evolved for millennia under broad-spectrum light and only utilize those two specific components. Besides, it is so much easier for me to tell what is going on with the girls under white light. Our vision obviously didn't evolve under a red/blue mix!

I did not write the info, but when it was posted in my thread, I saved it to learn from, and to help my IC homies

 

[repuk]

BTW, I checked AC amperage, 0.25A the LPC-60-1050 (AC voltage read 232V), or 58W. The LPC-60-1750 0.4A. That's 90Watts, had me puzzled as expected the 1750 to be much more efficient!?

 

[guvoo]

I put an inline fuse in the postivie rail with each of the strings: 1,5A for the XP-G's and 3A for the XM-Ls.

what a clever idea on a constant current driver!

 

[tenthirty]

BTW, I checked AC amperage, 0.25A the LPC-60-1050 (AC voltage read 232V), or 58W. The LPC-60-1750 0.4A. That's 90Watts, had me puzzled as expected the 1750 to be much more efficient!?

That's strange, put a kill a watt on the 1750. There is no way that the 1750 will draw 90 watts in my experience.
Does the driver get hot? (should be slightly warm)
If you need a trouble shooting procedure, I'll write one up for you.

By the way, nice lumiaires!! I can hardly wait!

 

[tenthirty]

I put an inline fuse in the postivie rail with each of the strings: 1,5A for the XP-G's and 3A for the XM-Ls.

what a clever idea on a constant current driver!

The leds/diodes are much faster than any fuse.
A zener diode would be a better choice.
I personally, run bare. No guts, no glory.

 

[repuk]

That's strange, put a kill a watt on the 1750. There is no way that the 1750 will draw 90 watts in my experience.
Does the driver get hot? (should be slightly warm)
If you need a trouble shooting procedure, I'll write one up for you.

I thought apart from measuring global DC amp output, to measure voltage across each Led vs total to calculate DC amp draw but had to do a lot of things and left that... The driver spec 0.7A draw for 240V so thought that either the driving outputs more than 1.75A (no big deal) or is rather inefficient. What would you check?

Is just warm to the touch.
I used the fuses I had at hand, intend to use 1.2A for te XP-G and 2.5A for XM-L.

My concern was that being the PCBs pads exposed I could shortcircuit something while working in the grow area.

By the way, nice lumiaires!! I can hardly wait!

thanks!!!! Me neither!!

 

[vukman]

Congratz repuk!!!!!!!!!! nice...very nice....now it's time to sit back and take notes and watch the show..

Good Luck

 

[tenthirty]

I thought apart from measuring global DC amp output, to measure voltage across each Led vs total to calculate DC amp draw but had to do a lot of things and left that... The driver spec 0.7A draw for 240V so thought that either the driving outputs more than 1.75A (no big deal) or is rather inefficient. What would you check?

Is just warm to the touch.
I used the fuses I had at hand, intend to use 1.2A for te XP-G and 2.5A for XM-L.

My concern was that being the PCBs pads exposed I could shortcircuit something while working in the grow area.



thanks!!!! Me neither!!

I see where you are geting the .7a figure.
http://www.meanwell.com/search/lpc-60/default.htm

I don't know what they meant by .7a at 230v typ.
but my strings of leds seem to be drawing spot on.
The efficiency rating seem about right as well.

The first and cheapest thing to do is to get a kill a watt or the equivalent and get a measurement. If you're happy with that, I would call it good.
If you want to go further, you could get out the Simpson meter and put it in series with the string of leds and get a current reading.
Then measure the voltage for the whole string and do the math.

Judging from the fact that everything looks good, I would say just get a kill a watt, just to do a spot check.

Running 9 XM-Ls and 9 XP-Gs I get something less than 90w at the plug, temp dependent.

I would also get a IR temperature gauge, though the rule of thumb is,
if you can't hold onto the heat sink it's too hot!

 

[repuk]

I see where you are geting the .7a figure.
http://www.meanwell.com/search/lpc-60/default.htm

I don't know what they meant by .7a at 230v typ.
but my strings of leds seem to be drawing spot on.
The efficiency rating seem about right as well.

Yep, it specs 1.2A for 100VAC and 0.7 for 240VAC

The first and cheapest thing to do is to get a kill a watt or the equivalent and get a measurement. If you're happy with that, I would call it good.
If you want to go further, you could get out the Simpson meter and put it in series with the string of leds and get a current reading.
Then measure the voltage for the whole string and do the math.

Judging from the fact that everything looks good, I would say just get a kill a watt, just to do a spot check.

I measured AC amperage by putting my (very accurate) multimeter in series with the driver AC input, I don't have a killawatt but guess it will measure the same?

Yep, I will check the whole string amperage and voltage.

Running 9 XM-Ls and 9 XP-Gs I get something less than 90w at the plug, temp dependent.

I would also get a IR temperature gauge, though the rule of thumb is,
if you can't hold onto the heat sink it's too hot!

I will measure temps and do a thorough check as soon as I can, though it will difficult this week...

 

[repuk]

Congratz repuk!!!!!!!!!! nice...very nice....now it's time to sit back and take notes and watch the show..

Good Luck

thanks vukman!! I've never done a SCROG neither used blumats so here we have three variables where I can fail, but I think hopefully the performance of the HML will be easy to see.

 

[hempfield]

@repuk : Excellent job my friend !

I have to suggestions for you :
- the wires between the LEDs could be a little shorter (just for aesthetic reasons - but it seems the plants does not care about that)
- you could mount the drivers on top of the heatsinks, using some metal/plastic spacer. This will give you a much robust light panel and also decrease the power loss on the wires ( we should not forget that leds are using DC , which is less suitable for long way transport. for every Ohm of the electrical path, at 1000mA you loose 1W - actually it's converted on head)

I also like the calibration of the water plastic drum :bigrin: A couple of years ago I bought a similar canister which supposed to have 20L. When I started to fill it up with a 1L bottle, I was disappointed to see that the real capacity was only ~17 Liters .


L.E. Could you guys post some temperature measurements of your heatsinks ? I usually measure temperature on :
- right next to the led star PCB ;
- right under the LED, on the other side of the heatsink ;
- 2-3" apart from each LED.

On my V3 LED panel, the maxim temperature (in center of the heatsink) is closed to 37C (98.6F) and no more than 35C (95F) on the edge.

Right now I work on to supplementary LED panels , having 5 x 10W leds on each of them. The heatsink is just a flat 3mm thick aluminium sheet (with some fins attached on the back). The temperature right next to the LEDs and under them is 51~53C (123~128F). I can't measure the chip temperature , but I suppose it'a a few degrees higher. I go with no driver at all, because I power this led panel using and IBM Server Power Supply Unit, and it's voltage is very stable 11.98V. I have put two 1N4007 diodes in series with each LED, and this give me a maximum voltage of 9.87V on the led, and a current of 780mA . This is the cheapest driver ever(no protection, but what the hack, I'v been using high power leds this way for years with no failure at all).

I will post some pictures soon.

 

[repuk]

thanks hempfield!

@repuk : Excellent job my friend !

I have to suggestions for you :
- the wires between the LEDs could be a little shorter (just for aesthetic reasons - but it seems the plants does not care about that)
- you could mount the drivers on top of the heatsinks, using some metal/plastic spacer. This will give you a much robust light panel and also decrease the power loss on the wires ( we should not forget that leds are using DC , which is less suitable for long way transport. for every Ohm of the electrical path, at 1000mA you loose 1W - actually it's converted on head)

I left them a little longer on purpose, they could had been a little shorter, but was concerned if I cut them too "tight" the warming/cooling cycles that make them to expand and contract could make solder joins to crack.

Good point on the driver dc cables!

I also like the calibration of the water plastic drum :bigrin: A couple of years ago I bought a similar canister which supposed to have 20L. When I started to fill it up with a 1L bottle, I was disappointed to see that the real capacity was only ~17 Liters .

hahah You can see I used a 2L bottle!

I got this one for free from a car wash, washed and rinsed it thoroughly and let it sit filled for a couple weeks changing water in between... it's supposed to be 25L but it can hold 26L for sure, though I didn't want to fill it completely for the first time to have some room with the PH down adjustment: I haven't made a mix that big for some years and taking out water would have been a royal PITA...

L.E. Could you guys post some temperature measurements of your heatsinks ? I usually measure temperature on :
- right next to the led star PCB ;
- right under the LED, on the other side of the heatsink ;
- 2-3" apart from each LED.

On my V3 LED panel, the maxim temperature (in center of the heatsink) is closed to 37C (98.6F) and no more than 35C (95F) on the edge.

Right now I work on to supplementary LED panels , having 5 x 10W leds on each of them. The heatsink is just a flat 3mm thick aluminium sheet (with some fins attached on the back). The temperature right next to the LEDs and under them is 51~53C (123~128F). I can't measure the chip temperature , but I suppose it'a a few degrees higher. I go with no driver at all, because I power this led panel using and IBM Server Power Supply Unit, and it's voltage is very stable 11.98V. I have put two 1N4007 diodes in series with each LED, and this give me a maximum voltage of 9.87V on the led, and a current of 780mA . This is the cheapest driver ever(no protection, but what the hack, I'v been using high power leds this way for years with no failure at all).

I will post some pictures soon.

That is with each LED in parallel with the supply?

Will look for my IR gun and take some measurements, though it will be difficult this week.

For the touch I'd say right now my heatsinks should be near 55C.

 

[hempfield]

When I touch the radiators at temps over ~50C it feels almost hot , but in a few seconds I can adapt my fingers and it's bearable. If I could lower the temps under 45C it will be awesome - and I intend to do this like I did in the V3 panel : I will attache 3 or 4 "U" aluminium profiles on the back of the flat heatsink. It's very important to have the fins vertically so the heat could escape via natural convection and in the same 'migrate' in the entire body of the heatsink, away from the LEDs.

I measure the temps with the thermocouple probe which came with my multimeter and it's pretty accurate (not like an laser IR gun, but it gives me a good measurement of the temps).

 

[repuk]

I just had an unexpected chance to be with the girls, but couldn't find the IR thermo... you're gonna laugh... I'm afraid is inside the toolbox that I used as a reservoir raiser! damn!

I had a precision thermo probe for my multimeter, and used that instead... (just noticed you just used that hempfield! coincidences, coincidences! ) put the probe touching the heatsinks 15-20 minutes after switch on and it read 44C (111F) next to the star heatsinks, the hottest position... I read 41-42C on the heatsink side, sorry but forgot to measure on the heatsink fins...

Measured DC amperage: 1.770A; voltage: 30.5VDC, so DC wattage was something expected: 53.9Watts. Didn't measure AC Amps again as I guess they will be the same.

I will try to get a killawatt as it would be handy for other things here, but dunno if it will be easy or not to source here.

Glad LEDs doesn't burn, as the working position is rather awkward, Houdini tricks are like lying on the couch in comparison!!

BTW I think I may have found a cheap source of heatsinks... while buying some methacrylate for another project on a sign shop, I saw they sell aluminum U profiles in differing sizes by lenght... a 2 meter piece was like 80€. It was about 4-5mm thick, and the bottom of the U like 6" or 15cm, sides about 8-10cm. So maybe hunting for scrapes could result in good luck and a rather cheap heatsink... it may need profiles riveted to add fins or not if you use fans...

 

[hempfield]

Under 45C , that's great repuk !

The AC power is surely a bit higher, but this is normal: those driver achieve maximum efficiency when they have the biggest load as possible (maximum led number), and even then is not higher than 90~95% (I guess).

I think it's hard to get a Kill-a-watt here in Europe but I found a very useful device of this kind which measure : voltage, current, watts, vat, frequency, power factor, and in addition count the kwh and if you enter the price per kwh , it will show the total cost. I reset my counter when I start my current grow so it will be so easy to find the total cost of the electricity and to calculate the real g/kwh ratio.

I bought mine here in Europe, from the hardware store and it was around 20EUR. On it' back it is writen DUWI (with two horizontal dots over U - I don't have this caracter on my keyboard , it says that is imported in Hungary (I'm not in Hungary, but who know how knows what path did this device followed until it's final destination on the hardware store self) by a third german or polsky company , and of course that is was made in ... where else ... China. I am very happy with it .

 

[tenthirty]

Measured DC amperage: 1.770A; voltage: 30.5VDC, so DC wattage was something expected: 53.9Watts. Didn't measure AC Amps again as I guess they will be the same.

53.9w draw and and 90w input. If that is true, something is getting damn hot.