Let's talk DIY lights

Ok @dbrn32 and @MAXHeadRoom @1BigFella
I’ve separated these into “categories” based on size. Larger ones with more surface area we should put something like Cree’s on them. Unless anyone can think of a larger just as efficient one. Then the round and a few square ones aren’t as large, but are still just as tall for cooling. Those are 1” 1/2 across. Then we have 5 small heat sinks that are thin, and I think would probably overheat even with fans. Those are 1” 1/4 in more of a rectangular shape being 1” 1/8 on the skinnier side.
We discussed the bridgelux vero 8 for some of those, and I think those would work. Suggestions for the others would be great. I thought I would get everything together at least. I have yet to plug in the fans, however these did come from working units so they more than likely work.

5 Likes

You could actually do experiments to see how well they do with different levels of power: Get a heating element similar to a big COB, connect it to the mounting surface, and then drive it from a variable power supply. As long as it was DC, you should be able to measure volts and amps to get watts. Then watch the temperature of the heat sink with and without a fan. A good heat source would be a simple soldering iron but it has to be able to run at 80 watts or so. You could just solder the soldering iron tip to a little copper block and then mount the copper block on each heat sink with one screw and some thermal goo.

After lots of experiments, you should get a maximum number of watts each heat sink can dissipate with and without a fan. Most COBs put out about 50-60% of their power as heat, so you would be able to predict all the performances with different COBs without having to glue COBs on heat sinks and then not be able to use them.

You could also do the same experiments with AC if you had a variac (variable transformer) but not a light dimmer. Hard to measure current and AC voltage accurately with a dimmer.

You could also just measure each one and count the fins to calculate surface area. (Don’t forget to count both sides of each fin.) Then use the rule of thumb. I think it’s 17 square inches per watt passive and 5 square inches per watt with a fan. Something like that anyway.

4 Likes

We’re you following along with @Bogleg and I yesterday? We found citi 1212 for $9. About same efficiency running them at 40 watts as cxb at 50 Watts. Not sure you can run them passive on the smaller ones there. But may be worth throwing into mix at around 30 Watts passive or little higher with fan.

If your looking for an alternative to cree with watt for watt efficiency, luminous cxm-22. The buzz is that it’s about same as Cree at 50 watts and about the same as bridgelux at 75 Watts. At least one of the citizen cobs now falls in that range too, maybe the 1818. Pretty sure they are both 52 volt nominal. The 52 volt vero 29 isn’t far behind either.

I like what you’re looking at doing, but I feel obligated to at least caution you about over building. You don’t want to spend the cash to end up having to dim or run your lights higher than needed. Have extra is obviously better than not having enough. But if you get to the point of covering every square foot of your tent with a cob, 30 or so Watts per is plenty.

2 Likes

I think those numbers are pretty close as far as sizing of the heatsinks. Off the top of my head, I was thinking 4.75 for active and like 18 for passive. So definitely in the neighborhood. I think the thicker the base usually gives you some cushion too.

Good idea about the testing too. Makes sense if you can find a reasonable way to test them.

1 Like

Well, you could just glue a small COB on one and see. Might have to run with a fan, but you do have plenty of fans. I think you probably want to glue them on with that alumina epoxy, so you would not be able to remove them. I’ve got some 22 watt red COBs coming from China and recycled CPU heat sinks would probably be fine. Those Intel processors are around 60 watts, but they always run with a fan.

On another topic, I was thinking about the inverse square law today so I took some lux measurements. At the top of my plant it was about 60000 lux. At the bottom (about twice the distance) it was 30000! Not the 15000 you would expect. Then I measured the side light being reflected off my flat white side panels: About 10000 coming off each panel! The leaves at the bottom are actually getting 30000 + 40000 lux in side lighting. So the reflector side panels work great.

The only explanation I can think of for the top-to-bottom measurements is that the plant is between 4 COBs and at the top of the plant all the light comes at more of an angle from each COB. At the bottom of the plant the light comes from the four COBs at less of an angle so it’s stronger. That effect probably works with LED panels too.

1 Like

I would say they are blending and possibly getting some reflection as well.

I’m thinking you could test it. Try to get centered between cobs at the floor and then raise your meter. Should increase until you get high enough that you’re not in both beams at the same time. Then move slightly left and right. Get readings of the individual beams and then compare to the readings slightly lower when you’re in both beams at the same time.

Far from scientific, but you should be able to support what you’re thinking.

1 Like

I’ll try it tomorrow. My wife is sleeping right now and the lights are off. But that makes sense. The COBs have a very definite beam effect though it is continuous. Right below a COB is a lot stronger than off to the side even 30 degrees, but 30 degrees off you may be directly under another COB. And down near the floor, you can see all the COBs without much angle.

That’s actually good news: It means more distance makes a more uniform light field.

2 Likes

For sure!

On a side note, did you catch any of the traffic on the prices of the citizen gen 5 cobs at cobkits? 1212’s are dirt cheap. And I haven’t looked extremely close, but light spectrum of the gen 5’s appears to be smidge a better than gen 6. It’s almost as if they shifted phosphor to the greens a little to show an increase in efficacy. Good for warehouse lighting, but doesn’t do much for us. Speaking of 3000k 90 cri specifically.

May be worth checking out if you’re looking at doing some stuff for flowering.

1 Like

The spacing between the COBs really matters, in terms of the uniformity of the light field. Going from the floor to near a COB is a continuous intensity increase but not as bad as you would get with just one COB. Because down near the floor, the meter can see all four COBs I am using now. When you come up from the floor between COBs you do see continuous intensity increase but if you move the meter so it’s under one COB, the intensity goes down! Between the COBs, it sees both beams, The very highest intensity is equidistant from all four COBs right in the middle. But it is a small difference, like 60000 lux versus 50000.

1 Like

The spacing is pretty relative to the height I think. There’s a stoner report somewhere that they messed with cob spacing and heights. I think it was 3590’s at around 40 Watts, and they claimed at 12” centers they got the most amount of blending vs par level at 14-15”. I don’t remember where I seen it, or who it was, so I definitely could be off on the numbers.

I want to make sure I understand this, so I am going to us the light I am building as an example. Initially we had looked at 4 COBs along a rail long enough to cover 4’ (so let’s say 42"). With 4 cobs centered on that you’d need to space them every 10.5 inches. But to get the COBs to run more efficiently, we’re going to run them at a lower wattage and use 5 of them instead. 5 of them over that 42" should spaced every 8.4 inches, let’s say. I’m just using basic math on the spacing.

So with the 4 COB set up, because they are spaced wider, you’d need the light to be higher before you’d get the most amount of blending (cones of light starting farther apart). With the 5 COB set up, you’d end up being closer to the light to get the same amount of blending.

Is that correct? But the 5 COB set up would be more intense at the high of blending than the 4 COB at the height of it’s blending.

I might have eaten too much of that oatmeal cookie bar this morning…

2 Likes

Not if it’s in a straight line but yes if it’s built in a square formation. My frame essentially uses 3 cobs (blended) per plant. In a straight line you will on get 2 cobs per plant blended.

1 Like

Ya you’re not looking at math right but we will have to figure out the spacing.

Let’s say doing 4 cobs at 10” spacing, that only takes about 30” obviously more because you have your heatsink diameter to accommodate that’s larger than the cob. But look at this way… the first cob is essentially zero. Then it takes 10” to get to the second cob, another 10” to get to third cob, and another 10” to get to fourth. So a direct measurement from cob one to cob four on 10” centers would be 30”. Make sense? So you’re overall bar length in that scenario would just have to accommodate the distance from the center of the end cobs to the end of the heatsink.

1 Like

:upside_down_face: got to look at equations like this!!

1 Like

Pretty much. I know the average cat says 10x4 is 40. It’s hard to tell them it isn’t haha.

1 Like

@Bogleg when the heatsinks come in I would take whatever you need need off of each piece of 48” angle to accommodate width of your heatsinks into frame. They are 133mm sinks right? So measure out how you intend to install, the distance will vary whether you mount the cross bars inside your angle or outside, but should be about 5” or so off each piece with about a 43” overall bar length. That should give you a couple inches of clearance on each end of tent. Then install a cob on each outer end and one dead center on bar. After that simply split the difference between center and outer heatsinks. That will get you even without stressing about the math and coincidentally at about that 10” cob spacing we are looking for.

1 Like

That makes sense! Thanks (again).

1 Like

dbrn32 is right about 5 COBs with only 4 gaps between them. It’s a cardinal versus ordinal thingie. But really you just have to lay the heat sinks on the table along the rail. It will be obvious where to mount them to get even spacing. The COBs illumination width is so wide that at any reasonable distance many plants will see strong light from a COB, and any plant will see many COBs. I decided mine are putting so much light out to the sides I have ordered some Angelina reflectors from Rapid LED. They should be coming any day now and I will post the difference in lux since they reflect more of the side light down at the plants.

$9.50 each with the mounting ring.

3 Likes

Good clarification on laying out the bar. @Bogleg you can certainly lay them out like that prior to physically mounting. It didn’t dawn on me to explain it that way. You can really do it any way you’d like. I just meant from a measuring standpoint, spread them out evenly along the whole length.

If and when he gets to transforming his entire grow switched over to cobs, he can look at reflectors then. For now, it doesn’t make sense to give up any photons to have a more direct beam. His cob holders should already function as a mounting point for the little angelina, so it’s a simple upgrade. But I think even when it’s complete he’ll be better off without them.

I have a pretty good idea of how I want to do it based on my space and my future expansion plan. But now I understand the basic carpentry term of measuring from center. Haha! I’m surprisingly lacking in “handiness” experience for a dude my age. I can build a computer without any references but ask me to do carpentry and I’m like Gene Hackman’s character in Unforgiven.

3 Likes