Heatsinks for DIY LED lamps

AquariusPanta

Well-Known Member
I had a similar idea and a pair of 5000K CXA3590s hanging around. Running them at 250mA and no temp droop to speak of, they are about 65% efficient or 210lm/W. The 5000K Vero 29 at 250mA should be about 58% efficient or 188lm/W. Awesome vegging lights! Cost is $5-$5.50 /PAR W, much less than I recently paid for my Oslon SSL deep reds LOL.

Mounted on Alpine 11s so they can be cranked up if necessary.
View attachment 3306992
Crude math, just stinking.

1x vero 29, 70cri, 5000k+molex EZ-mate: $35
1x 2"x5.886" Heatsink: $3.50
1x Fasttech driver Link: $2.50
=$41+($41*0.15)= $48

Estimating about 61% efficient even after temp droop, 10.5w power draw.

A bit expensive but it would beat the pants off a CFL and would be fairly easy to assemble.
Your all taking the light vs. heat efficiency wayyyyyyyyy to far lol, to a point of noooooooo return. The middle ground is very much a'lush.
 

SupraSPL

Well-Known Member
Do you guys paint the heatsink or the cob with paste?

If you are making a bunch, why not use a stencil? Or if youve got the tools a silkscreen?
I have never tried it. I normally dab the paste onto the back of the COB, spread it and then press hard on the four corners for 30-60 seconds to get as thin a layer as possible. Then I clean up the edges with ISO alcohol. Recently I tried using the stock paste that came on a few CPU coolers. And I did not bother to sand the surface. I am in the process of testing the performance difference between, polished and unpolished CPU coolers. So far I can say, the difference ay 900mA is measurable, but possibly negligible. Now testing at 1.4A.
 
Last edited:

SupraSPL

Well-Known Member
Your all taking the light vs. heat efficiency wayyyyyyyyy to far lol, to a point of noooooooo return. The middle ground is very much a'lush.
True, at that point heat is a complete non issue. But since vegging lights are on for 18-24 hours a day, it is a nice place to try and save some kWh. When I get power outages I am able to power the vegging areas with backup batteries which does come in handy from time to time. Also, I have enough nerd in me to be excited about putting LEDs to work that are over 200lm/W.
 
Last edited:

Abiqua

Well-Known Member
Supra, I put a dab on the back and don't spread it initially, not until I place it down on the sink and then working it back and forth like you, let's it spread. Then I hold, which is great advice, by the way.

Back in the day on Cpu mounting..... I was taught, that if any comes out, you have too much.......
 

AquariusPanta

Well-Known Member
True, at that point heat is a complete non issue. But since vegging lights are on for 18-24 hours a day, it is a nice place to try and save some kWh. When I get power outages I am able to power the vegging areas with backup batteries which does come in handy from time to time.
The electricity costs do add up quickly.

For my next DIY COB fixture, I'll be ordering a super fancy dimmable driver so I can control costs a little better during vegging season. It should be worth the investment.

Always room for improvement
Ain't that the truth.

But like anything else, too much of one thing typically results in the lack of something else important, such as money in this instance.
 

SupraSPL

Well-Known Member
This test examines the value of sanding/polishing and expensive thermal compound versus stock CPU cooler.

Ambient temp 22C
Heatsinks: Arctic Alpine 11 (2250cm²)
One has a flattened polished surface and the other is stock with stock paste (MX4)
CXA3070 AB @ 1.45A = 53W dissipation
43cm²/W
fan at 5V = .55W
fan at 9V = 1.87W

Stock heatsink with MX4:
fan 5V - 2.9% temp droop
stable heatsink temp 28.5C

fan 9V - 1.87% temp droop
stable heatsink temp 26C


Polished heatsink with Prolimatech PK3
fan 5V - 2.9% temp droop
stable heatsink temp 28C

fan 9V - 1.88% temp droop
stable heatsink temp 27C

So in this test the results were exactly the same. If the polished heatsink had any advantage, it was lost within the margin of error. Ouch that hurts, I have been promoting polishing and I have spent an awful lot of time doing it, although I maintain that 20mm stars are unruly. I would like to repeat this test with the Vero29 running at 2.4A to see if there are advantages at high current. Working on finding the right driver for that setup.

Also, worth mentioning, this is the 4th group of tests that has demonstrated the 5V outperforming the 9V
Baseline 42%
5V - 40.36%
9V - 39.8%
 
Last edited:

epicfail

Well-Known Member
Yessr it looks like you could improve efficiency significantly by swapping in a 5V, 6V or 7.5V adapter, easy upgrade :)
Remember I'm using 4-wire pwm fans, I can get them to run stable at 20% speed off the 12v power source. I built a PWM circuit with integrated 5v regulator to control the speed with a 100k pot. The fans, PWM circuit and arduino timer which controls the timing of the main white lights and the 730 flower initiator, all run off the same 12v 60w power supply. I need to get another kill-a-watt type device to get a reading on how much power they all draw when running at 30%. Supposedly controlling 4-wire fans with pwm is more efficient than using 3-wire with voltage, I'm no expert though.

I have a quantum meter (PAR), I could perform a test with the pwm fans running low and off and see what kind of difference in output there is. Without the fans the black anodizing would be doing more than just protecting from oxidation. I have read that heatsinks, when black anodized can reduce thermal resistance by up to 25%. This really has no effect when you use a fan but for passive cooling it does actually help a bit. I'm just not sure how much.
 

SupraSPL

Well-Known Member
This test is to compare the Arctic Alpine 11 ($10) versus the Rosewill RCX-Z1 ($13). The Rosewill received bad reviews, probably due to its lack of capacity (surface area 1525cm²). This means it would perform poorly under large loads of heat. But for our purposes, it has a nice flat copper pad, a thick base plate and thick fins. It has a high RPM so I didnt bother testing it at 9V. In summary, the Rosewill very slightly outperforms the Alpine 11, and that is using the stock paste with no surface prep for both.

I added the Arctic Freezer 7 Pro for a direct comparison.

Tested using CXA3070 AB bin running at 1.45A. Ambient temp was 23C

---------------------------------
Stock Alpine 11 (MX4 paste)
fan 5V - 2.9% temp droop
stable heatsink temp 28.5C
fan .555W

fan 9V - 1.87% temp droop
stable heatsink temp 26C
fan 1.87W
----------------------------------
Stock Rosewill RCX-Z1
fan 5V - 2.11% temp droop
stable heatsink temp 28C
fan .72W
-----------------------------------
Arctic Freezer 7 Pro
fan 9V - 1.33% temp droop
stable heatsink temp (25C fins, 29C copper pad)
fan 1.63W

fan 5V - 2.43% temp droop
stable heatsink temp (28.5C fins, 31C copper pad)
fan .504W
-----------------------------------

Baseline 42%
Alpine 11 @5V - 40.36%
Alpine 11 @9V - 39.8%
Rosewill @5V - 40.56%
Freezer 7 @5V - 40.6%
Freezer 7 @9V - 40.2%
 
Last edited:

Abiqua

Well-Known Member
Invaluable info!

One thing that has been crossing my mind, small detract, is how much rpm and millibar pressure come into play thru the various fan options? Its is probably a small impact if any at all, but...

for instance, anyone contemplated using a small 12v blower? rather than an axial fan? they usually have lower rpm's but they usually have millibar pressure in magnitudes greater than double most axial fans..

when I started out with mini-cabs, one of the solutions for small fans and Diy carbon filters was to use a 12v blower with slower rpm but far higher back pressure, which was able to push thru the filter much more efficiently.

Probably nothing, but thought I would add it, to the convo....:peace:
 

SupraSPL

Well-Known Member
A good example of that, I tried using an axial fan to exhaust a grow box, but when I switched it to a bath fan, I got much more air flow because it could handle the pressure so much better. As far as the CPU coolers are concerned, I expect we are dealing with very low pressures, especially at 5V. For those who are venting a lamp that is enclosed, it may come into play.
 

bicit

Well-Known Member
This test is to compare the Arctic Alpine 11 ($10) versus the Rosewill RCX-Z1 ($13). The Rosewill received bad reviews, probably due to its lack of capacity (surface area 1525cm²). This means it would perform poorly under large loads of heat. But for our purposes, it has a nice flat copper pad, a thick base plate and thick fins. It has a high RPM so I didnt bother testing it at 9V. In summary, the Rosewill very slightly outperforms the Alpine 11, and that is using the stock paste with no surface prep for both.

Tested using CXA3070 AB bin running at 1.45A. Ambient temp was 23C

Stock Alpine 11 (MX4 paste)
fan 5V - 2.9% temp droop
stable heatsink temp 28.5C
fan .555W

fan 9V - 1.87% temp droop
stable heatsink temp 26C
fan 1.87W

Stock Rosewill RCX-Z1
fan 5V - 2.11% temp droop
stable heatsink temp 28C
fan .72W

Baseline 42%
Alpine 11 @5V - 40.36%
Alpine 11 @9V - 39.8%
Rosewill @5V - 40.56%
built in dimmer anyone?
http://www.newegg.com/Product/Product.aspx?Item=N82E16835200014

Seems like the pre-applied tim is adequate for lower power lamps. Which is great for the uninitiated and the lazy :P

In light of this, makes one wonder how the thermal 'pads' and thermal adhesive stack up.
 

SupraSPL

Well-Known Member
O forgot to mention that, the Rosewill RCX-Z1 does come with a 12V fan dimmer. It works great from the few minutes I played with it, but I wonder about the efficiency. Dimmer efficiency test soon.

The Rosewill was $12.99 at Newegg and Amazon for black Friday so I grabbed a few.
 
Last edited:

Dloomis514

Well-Known Member
I think COBs are flat enough, but your idea about stars needing some flattening seems like a good idea to me.

This test examines the value of sanding/polishing and expensive thermal compound versus stock CPU cooler.

Ambient temp 22C
Heatsinks: Arctic Alpine 11 (2250cm²)
One has a flattened polished surface and the other is stock with stock paste (MX4)
CXA3070 AB @ 1.45A = 53W dissipation
43cm²/W
fan at 5V = .55W
fan at 9V = 1.87W

Stock heatsink with MX4:
fan 5V - 2.9% temp droop
stable heatsink temp 28.5C

fan 9V - 1.87% temp droop
stable heatsink temp 26C


Polished heatsink with Prolimatech PK3
fan 5V - 2.9% temp droop
stable heatsink temp 28C

fan 9V - 1.88% temp droop
stable heatsink temp 27C

So in this test the results were exactly the same. If the polished heatsink had any advantage, it was lost within the margin of error. Ouch that hurts, I have been promoting polishing and I have spent an awful lot of time doing it, although I maintain that 20mm stars are unruly. I would like to repeat this test with the Vero29 running at 2.4A to see if there are advantages at high current. Working on finding the right driver for that setup.

Also, worth mentioning, this is the 4th group of tests that has demonstrated the 5V outperforming the 9V
Baseline 42%
5V - 40.36%
9V - 39.8%
 

AquariusPanta

Well-Known Member
This test is to compare the Arctic Alpine 11 ($10) versus the Rosewill RCX-Z1 ($13). The Rosewill received bad reviews, probably due to its lack of capacity (surface area 1525cm²). This means it would perform poorly under large loads of heat. But for our purposes, it has a nice flat copper pad, a thick base plate and thick fins. It has a high RPM so I didnt bother testing it at 9V. In summary, the Rosewill very slightly outperforms the Alpine 11, and that is using the stock paste with no surface prep for both.

I added the Arctic Freezer 7 Pro for a direct comparison.

Tested using CXA3070 AB bin running at 1.45A. Ambient temp was 23C

---------------------------------
Stock Alpine 11 (MX4 paste)
fan 5V - 2.9% temp droop
stable heatsink temp 28.5C
fan .555W

fan 9V - 1.87% temp droop
stable heatsink temp 26C
fan 1.87W
----------------------------------
Stock Rosewill RCX-Z1
fan 5V - 2.11% temp droop
stable heatsink temp 28C
fan .72W
-----------------------------------
Arctic Freezer 7 Pro
fan 9V - 1.33% temp droop
stable heatsink temp (25C fins, 29C copper pad)
fan 1.63W

fan 5V - 2.43% temp droop
stable heatsink temp (28.5C fins, 31C copper pad)
fan .504W
-----------------------------------

Baseline 42%
Alpine 11 @5V - 40.36%
Alpine 11 @9V - 39.8%
Rosewill @5V - 40.56%
Freezer 7 @5V - 40.6%
Freezer 7 @9V - 40.2%
O forgot to mention that, the Rosewill RCX-Z1 does come with a 12V fan dimmer. It works great from the few minutes I played with it, but I wonder about the efficiency. Dimmer efficiency test soon.

The Rosewill was $12.99 at Newegg and Amazon for black Friday so I grabbed a few.
It looks like the Arctic Alpine 11 Plus is the underlying winner so far, as for only $10 (free shipping XD) you get near efficiency in comparison to those models that are a tad more expensive. I do like the feathers the RoseWill RCX-Z1 carries; the red and black colors look nice.

If one didn't want to use the dimmer (PWM?) for the fan because one had paid attention to your enlightening posts and bought a OEM 5V PSU, could they use the dimmer along with their lights, to control power output? I understand some drivers include the PWM nob but most of the cheaper ones don't.
 
Top