Pictures of your DIY lights - Post your pics!!!

[mrandersonnen]

I would definitely stay away from kapton fixture.

i have thermal compound on the back of the chips. what about kapton concerns you?

 

[SomeGuy]

Here is an "other" light I slapped together. I just nicknamed it el cheapo after a few rips and beers. It cost 90$ or close to that. It is 250ww induction. Pulling 273 at the plug.



Yes..... It's plywood framed and has a roasting pan reflector. . Lol

 

[cdgmoney250]

8-12" here hortilux shps 1kw, horizontal though, in a sealed sun systems reflector.

Gnarly. But if it works it works. Do you ever get any bleaching or heat stress?

 

[MrEus1]

plants have "Active transport" for specific ions. Its not a matter whether they want or not. mostly only the smaller ions like NA or K that are absorbed passively.,
there is also the fact that while a plant does absorb some ions by osmosis, if they don't "metabolize them" ie incorporate them into proteins, cell walls etc. it just sits in solution, just like the blue dye in celery uptake experiments

You are right my friend. What is happening, with sugars and many other "plant fertilizers" is that we are feeding the microbes, or some would say the soil and they have to break it down into a small enough molecule of something that the plant may actually use. That's fine for organics and is a sloooowww process. Still, in my opinion a great percentage of nutrient additives are a waste. Now some fine tweaking and an ability to read a plant, those are true skills that can prove beneficial. Please know that I'm not trying to tell any one how to tend their garden. There are many ways to have success. No worries.

 

[SPLFreak808]

Gnarly. But if it works it works. Do you ever get any bleaching or heat stress?

My veg is approx 6 weeks, during that time, if they cannot handle my hortilux blue at 5-10" away, i just toss them because the hps WILL bleach them if the MH/Blue does. i dont have any sealed place higher than 5ft to flower. If i get desperate, ill dim the bulb but that always sucks!

 

[Castaman]

Are you sure that Kapton played any part?

What did the back of the diode look like? Was the surface area impregnated by anything?

Something doesn't add up all the way, thanks

No, there wasn't any bubbles or impregnancies at the back, but i think Kapton loose it's tension due to temperature fluctuations, as well as compound loose it's "vacuum" because it dried. So, if there was a holder, i think temps didn't get that high.

 

[Castaman]

i have thermal compound on the back of the chips. what about kapton concerns you?

You can face with situation like mine.

 

[Meinolf]

@SomeGuy, super cool study! It should feature 50-100 PARW of highly efficient LEDs and a wee bit more sophistication on cooling, but I like the compact form very much.

 

[SomeGuy]

@SomeGuy, super cool study! It should feature 50-100 PARW of highly efficient LEDs and a wee bit more sophistication on cooling, but I like the compact form very much.

I had the parts for a while and needed more light quickly. Trying to avoid hid all together. Still have another cob bar to build also...

FYI. It's going over plants now. They seem to like it with the couple kessils from @Mohican

 

[MrEus1]

CXB3590 36v
Meanwell 185h c1400bb
Arctic alpine 11 gt

So if you would be so kind, I am about to put the same combo together and I'm trying to find out if when I run the driver at full capacity, are each of the 4 cobs receiving 50 watts? Or do I need to turn it down and if that's the case how do I determine how much power each cob is receiving? If you can answer this, I'd be greatly appreciative. If you can't answer it, thank you all the same.

 

[Abiqua]

No, there wasn't any bubbles or impregnancies at the back, but i think Kapton loose it's tension due to temperature fluctuations, as well as compound loose it's "vacuum" because it dried. So, if there was a holder, i think temps didn't get that high.

Do you know have many CC's of material were used?

Any pics?

Something is not adding up, not trying to Call you out, just trying to figure this out....

I have used a handful of Vero 10's for well over 14 months of 18 hour use everyday. Personally I haven't experienced this problem, so I am trying to figure out how to avoid it in the future....

Based on testing of Kapton @SupraSPL, something is a amiss....Any thoughts?

 

[SupraSPL]

I recently took apart a fixture from 2010 that used kapton to insulate copper traces, it still stuck like it was brand new and left no residue.

If applied to a clean aluminum surface it will not lose its grip with time or temperature fluctuations. It will hold the COB in place so that even if the TIM dries out it will still be making full contact.

Castaman do you think the tape was applied tight or was it never holding the COB in place? If so, the TIM would have been doing all the work eventually allowing the COB come loose. Once there is an air gap the COB is toast, even if running soft. We saw another case (dloomis?) where a kapton mounted COB came loose due to anti corrosion film coating on a copper heatsink. This film has to be cleaned off for the kapton to adhere. Even without a film, it is a good idea to clean the surface before applying tape.

Kapton mounting is a viable option for those who do not want to drill holes and may be less intimidating for some DIYers. From an engineering standpoint screwing them in is the "proper" way to install a COB, yet I have never screwed in a single one or else my heatsinks would be swiss cheese by now. So kapton vs screws, it comes down to your preference.

 

[Castaman]

Do you know have many CC's of material were used?

Any pics?

Something is not adding up, not trying to Call you out, just trying to figure this out....

I have used a handful of Vero 10's for well over 14 months of 18 hour use everyday. Personally I haven't experienced this problem, so I am trying to figure out how to avoid it in the future....

Based on testing of Kapton @SupraSPL, something is a amiss....Any thoughts?

Yeah pics will be soon. Can't get to the place right now. I'm IT engineer for 15 years, applying TIM isn't much differ from CPU installation to get the "vacuum seal". Almost identical light is working without any issues. I think my TIM doesn't meet Cree requirements for TIM's, as well as kapton tape isn't recommended to fix Cree Cob's. I also saw incidents when cxa cob degraded under lens because of phosphor evaporation. Cree also has recommendations for ventilated lenses in datasheets for cx** series. Yes kapton was applied tight, but after 3 cycles i think it might loose it's tightness because of numerous temperature drops.
Moreover i found some info about cxa phosphor degradation and will publish it here after i translate it to English. In the meantime here is some spoiler photo:

 

[Castaman]

I really admired of your work Supra, you are the person who brought me in to cob's universe, but let me disagree with you.
Your and other "lucky" (sorry for using this word) diy builds with capton tape can't be generalized.
It's just imho, i'm not trying to insult you.
I think my case is proving that. Here is some new photos:
General view - second cob undamaged

TIM isn't dry, it's not the air gap - i touched it with my finger at the bottom of the cob and heatsink, it's condition is like new, mb a bit solid:

"Sanded" surface

Second cob

So, you decide to use kapton or not. I'm not recommend it, although it can work fine with some other setup.
Even some thing like this

will do better:

 

[werm11]

Just curious...is everyone still sanding their heatsinks before applying the tim/cob? I thought I read in a thread that it was determined that was an unnecessary thing to do. Anyone know about this?

 

[Al Yamoni]

Just curious...is everyone still sanding their heatsinks before applying the tim/cob? I thought I read in a thread that it was determined that was an unnecessary thing to do. Anyone know about this?

I thought that is what I had read as well.. I believe it was that the difference was extremely small for the amount of work put in. Don't quote me on that one, It's certainly been a while.

 

[SupraSPL]

I really admired of your work Supra, you are the person who brought me in to cob's universe, but let me disagree with you.
Your and other "lucky" (sorry for using this word) diy builds with capton tape can't be generalized.
It's just imho, i'm not trying to insult you.
I think my case is proving that.

Good info, we can agree to disagree on this. Kapton+paste is a viable way to mount COBs. Like anything it does have to be done correctly and with some care. Many DIYers prefer screwing them in which is understandable and "proper" installation.

This is a CXA3070 3000K Z2 that I keep around as a shop light because the ceramic has a crack (caused by me prying on it long ago). It was installed on this heatsink with PK2 12~24 months ago. Running at 38W, it hangs with the COB facing down. Surprisingly the COB has remained in place without kapton and maintained a perfect thermal interface. I am not recommending this of course, I do consider it "lucky" but it proves the point, as long as there is no air behind the COB it will stay very cool and very low temp droop. We just have to do whatever it takes to keep the thermal interface intact.

 

[SupraSPL]

I thought that is what I had read as well.. I believe it was that the difference was extremely small for the amount of work put in. Don't quote me on that one, It's certainly been a while.

@werm11

That is correct, I had to retract my recommendation to flatten heatsinks. When I started temp droop testing I measured no difference between a sanded heatsink vs stock, even though the stock heatsink was far from flat. No difference between PK3 paste vs MX4 or MX2 paste. The reason I am so convinced of this is because I did a direct comparison with COBs running at 100W, and yet I measured no difference. Most of us run at 50W or even 25W so it is even less of a thermal bottleneck for those designs.

The reason for this, much of the temp droop is occurring inside the package and increases with current, beyond our control for the most part.

I am not saying the thermal interface is unimportant because it is critical, the COB will die without it, but as long as there is sufficient heatsink and the interface is sound (filled with paste and no air gaps), you will get very good temp droop figures.

 

[Al Yamoni]

@werm11

That is correct, I had to retract my recommendation to flatten heatsinks. When I started temp droop testing I measured no difference between a sanded heatsink vs stock, even though the stock heatsink was far from flat. No difference between PK3 paste vs MX4 or MX2 paste. The reason I am so convinced of this is because I did a direct comparison with COBs running at 100W, and yet I measured no difference. Most of us run at 50W or even 25W so it is even less of a thermal bottleneck for those designs.

The reason for this, much of the temp droop is occurring inside the package and increases with current, beyond our control for the most part.

I am not saying the thermal interface is unimportant because it is critical, the COB will die without it, but as long as there is sufficient heatsink and the interface is sound (filled with paste and no air gaps), you will get very good temp droop figures.

Excellent info! Thanks as always.

 

[werm11]

@werm11

That is correct, I had to retract my recommendation to flatten heatsinks. When I started temp droop testing I measured no difference between a sanded heatsink vs stock, even though the stock heatsink was far from flat. No difference between PK3 paste vs MX4 or MX2 paste. The reason I am so convinced of this is because I did a direct comparison with COBs running at 100W, and yet I measured no difference. Most of us run at 50W or even 25W so it is even less of a thermal bottleneck for those designs.

The reason for this, much of the temp droop is occurring inside the package and increases with current, beyond our control for the most part.

I am not saying the thermal interface is unimportant because it is critical, the COB will die without it, but as long as there is sufficient heatsink and the interface is sound (filled with paste and no air gaps), you will get very good temp droop figures.

That's very good to know. Thank you for clearing that up. Have you ever tested the pre applied tim that comes on the arctic alpine 11?