I'm about to buy 20 of these heatsinks for the 20 cob fixture, or shouldn't I?

I did some research and it seem pretty easy to solder. Just need a hot heatgun or a torch and melt a ball of solder on a non oxidated surface where you want chip. Them just add the chip with fluss on it. A non acid type fluss so one doesnt melt the electronics :)

Should make a great connection for heat transfer.
sounds interesting, try it out and let us know how it went :D
 

Nugachino

Well-Known Member
Yes but are you going to cook the chip by putting that much heat through it? That's the other thing with soldering.

Its usually high heat over short periods. Otherwise you risk burning the components.
 

furalle

Active Member
sounds interesting, try it out and let us know how it went :D
I will give it a go
Yes but are you going to cook the chip by putting that much heat through it? That's the other thing with soldering.

Its usually high heat over short periods. Otherwise you risk burning the components.
Yeah thats the concern but I will check it out. I have some one dollar chips from china to try on. i hope most of the heat will be transfered to heatsink and it should cool down very fast.

Will try with a leadsolder that melts at 182c. Just have to get a small propane torch
 

welight

Well-Known Member
what you need to consider is the redundancy factor, active heatsinks are by nature something that you drill a fan onto, little more, thats why their cheap, if the fan is highly reliable, 100k hours then no issue, if the fan fails the COB will be momentarily behind it as there is no where to hide thermally and if one fails what does it say about the rest. Passive cooling is a full nights sleep, just MHO
Cheers
Mark
 

Danielson999

Well-Known Member
Personally I'd rather spend a few more bucks for the piece of mind that a passive design gives you. Never having to worry about a fan failing and having everything running more silent is nice too. I see your dilemma though, trying to buy 20 cobs and all the associated hardware can get expensive real quick.
 
Personally I'd rather spend a few more bucks for the piece of mind that a passive design gives you. Never having to worry about a fan failing and having everything running more silent is nice too. I see your dilemma though, trying to buy 20 cobs and all the associated hardware can get expensive real quick.
I understand your concern, but then again:

40 cob's, which will need 40 heatsinks will cost me either:

40 x $20 passive + shipping VS. 40x $1.50 active

The price difference is so great between these 2, that a burned out chip caused by a failing fan is easily covered by the savings on heatsinks.
All my fans are powered individually, so 1 failing fan won't do harm to the rest of the fixture.

Besides, this cob project is just an experiment that I'm enthusiastic about. I'm replacing 2x Gavita 1000's which already get me very decent harvests, so I really can't justify spending that much more on this just for the heck of it.
I wen't for citi 1212's for a simple reason: price-quality. I would rather have waited for the new Cree's or something more efficiënt, but my budget wouldn't let me. I'm not a commercial grower, I'm a smoker
 

MasterpieceNutes

Well-Known Member
Did you build one bar first I hope? I cringe a little watching someone build 10 pieces of something with no proto.

You are 'ok' with the extra heat from those 1212s rather than utilizing a bigger chip for less heat?

I originally started my light builds using vero 18's and cree 25xx. If I continued to build with the smaller chips, the extra heat in filling my room with these would require active AC cooling of the room. (kind of self-defeating the $$ we are pouring into these lights).

I used 11 cxb 25xx (1200ma) in a 4x4 tent and heat was gradually building without near constant air exchange. Switching to 6 vero 29 (1400ma), no heat problems. I ended up mixing and matchin chipsets on rows for larger rooms. At the end of the day: :Still no ac required in the summer, so a win imo.

All active cooling on my builds too. I use cpu fans also. $5 each on ebay/amazon. Super quiet. -I expect I will hear a fan get loud if it decides it will fail. Simple DC Fans dont usually just 'break' or 'not turn on'. -I've never seen one break unless I dropped something on it ;)

So anyways, what I'm saying is you might want to proto one of each and check it out. Other options: you could put two 3618's on an HLG320 @ 1400ma and add a chip if it runs too warm (3 chips @ 1050, 4 @ 700ma).. Expand-ability.. With the 1212's you are looking at chip replacement or dependent on ballast: working more 1212s on your string.

Dont solder if you dont have to. Chip holders are so much easier. They look cleaner. They build faster.

Hope any of this helps. -Learn from my mistakes! lol. GLHF.
 
Did you build one bar first I hope? I cringe a little watching someone build 10 pieces of something with no proto.

You are 'ok' with the extra heat from those 1212s rather than utilizing a bigger chip for less heat?

I originally started my light builds using vero 18's and cree 25xx. If I continued to build with the smaller chips, the extra heat in filling my room with these would require active AC cooling of the room. (kind of self-defeating the $$ we are pouring into these lights).

I used 11 cxb 25xx (1200ma) in a 4x4 tent and heat was gradually building without near constant air exchange. Switching to 6 vero 29 (1400ma), no heat problems. I ended up mixing and matchin chipsets on rows for larger rooms. At the end of the day: :Still no ac required in the summer, so a win imo.

All active cooling on my builds too. I use cpu fans also. $5 each on ebay/amazon. Super quiet. -I expect I will hear a fan get loud if it decides it will fail. Simple DC Fans dont usually just 'break' or 'not turn on'. -I've never seen one break unless I dropped something on it ;)

So anyways, what I'm saying is you might want to proto one of each and check it out. Other options: you could put two 3618's on an HLG320 @ 1400ma and add a chip if it runs too warm (3 chips @ 1050, 4 @ 700ma).. Expand-ability.. With the 1212's you are looking at chip replacement or dependent on ballast: working more 1212s on your string.

Dont solder if you dont have to. Chip holders are so much easier. They look cleaner. They build faster.

Hope any of this helps. -Learn from my mistakes! lol. GLHF.
Thanks for your reply!
Heat is not my concern as I'm replacing 2x 1000W Gavitas. Besides, I'm running these 1212's low at only 50w each.
Soldering the chip interests me because it seems the most efficient way of heat transfer into the heatsink.
 
Top