Heatsinks for DIY LED lamps

i

bicit said:

It doesn't need too

Click to expand...

many designs use a wide thinner heatsink for optimal passive flow. the whole width needs airflow.

so how is it not needed ?

I can use one 120 mm fan which happens to have a higher cfm/w, and get air flow through all of the fins ....... how is an 80 mm better?

PurpleBuz said:

many designs use a wide thinner heatsink for optimal passive flow. the whole width (assuming fins are oriented lengthways) needs airflow.

so how is it not needed ?

Click to expand...

In the post I quoted you were talking about the physical dimensions of the fan and it's orientation. Not so much the amount of air flow.

As long as the fins directly above the cob are cooled it's usually sufficient. You don't have to cool the entire width of the heatsink. It's not going to drastically alter performance if the outer fins don't get cooled directly. This is all very abstract not talking in specifics.......

Now if we're talking about something like realstyles setup, where he staggered his cobs on a particularly wide heat sink, then yeah, you would want to cover the whole width of the heatsink with fans. Or if you have particularly long heatsinks. Otherwise the size of the fan doesn't make that big of a difference....

course there are many variable not being taken into account here...... surface area of the heatsink in question, cob orientation, and power displacement being the elephants in the room.

well open your eyes bud, cause you are generalizing based on false ASSumptions

Alright, simmer down peeps. This is how we get to conclusions, nobody (in this thread anyhow) is purposely giving ASSumptions for us to believe. It's hard to wholly communicate in a thread. Sometimes new info comes to light, etc etc. we are all mostly spacey and it's easy to overlook stuff. I am lazy so I read less, lol.

Cmon! Peace love and friggin rainbows!!

See I just read all this...and now I'm like ok...purp has a legit issue, but we all are here to learn. Constructive criticism will help us, and our plants grow.

It's ALL good!!!!!!

PurpleBuz said:

well open your eyes bud, cause you are generalizing based on false ASSumptions

Click to expand...

To what exactly? As far as im aware we were discussing vague hypothetical situations. Were you referring to a specific case?

I sized the sinks big enough for passive so I imagine a 240 cfm fan flow through the fixture will be plenty.

Bonjour
I have a 1000X165X35 H.S cut in 4 pieces with 4 Xcxb 3590 3500k cd 36v and I am going to buy fan for this, but there's more, cheaper and better 120mm than 140mm (and I will not talk about 160mm...) and I do not want bigger one it must fit the h.s.
How many cfm or rpm should I aim for?
And when I am going to use 3590 @700mA (for now it is going to be 1400mA caus' I have 4) is half of the h.s would be enough? I mean 8 3590 instead of 4 on the same h.s? But with 8 120mm fans!
Have a great day★

Wanting to add some red stars on an 18" heatsink, no more than 25 vf total @ 700ma. What size heatsink to passive cool would I need from heatsinkusa?

Experimenting with some thermal paste. The working theory, since sanding the heastsinks seems to make no difference, probably the thermal paste doesnt either. I already tested the stock Rosewill paste, Arctic MX4 versus PK3 and there was no difference that I could measure.

New test using CXA3590 driven at 1.05A (80W). Arctic Silver Ceramique Versus Prolimatech PK3. The Ceramique came up with 4.5% temp droop and the PK3 was 4.3%.

Next test will be Arctic MX2, 65g tube for $35 . If the thermal paste makes no difference, might as well use a cheap, non toxic one that is easy and fast to apply.

SupraSPL said:

Experimenting with some thermal paste. The working theory, since sanding the heastsinks seems to make no difference, probably the thermal paste doesnt either.

First test using CXA3590 driven at 1.05A (80W). Ceramique Versus Prolimatech PK3. The Ceramique came up with 4.5% temp droop and the PK3 4.3%.

Next test will be Arctic MX2, 65g tube for $35 . If the thermal paste makes no difference, might as well use a cheap, non toxic one that is easy and fast to apply.

View attachment 3464574

Click to expand...

That's good to hear, since i used Ceramique on my Vero 29s. Only because that's what i had in my drawer - it was leftover from a computer build a from few years ago. Keep up the great work Supra!

since sanding the heastsinks seems to make no difference

Click to expand...

I must have missed when and where this was discussed. In everyones experience, are most of the wider profiles from heatsinkusa good to go. No prep other than cleaning, applying thermal paste, and mounting COBs?

NastyN8t said:

so far like 4 months at 66% power. And yeah I realize if u don't run cree or Vero no one takes u serious here. But I'm pioneering my own water cooling set ups and didn't want to fry hundreds of dollars in cobs to prove my design. So far it kicks ass. I was familiar cree quality way before anyone had any serious grows w them my lil bro and his friends have been building cree powered flashlights for almost a decade. Chinese cobs are $4 a piece and made me some frosty harvest on my first run indoors. I'll definitely build a cree or Vero array eventually. My Chinese cobs+power supply is about $50 for every 400w. Are they as bad ass as the cree? Doubt it ......have they let me down?not yet but I have spare components on hand just in case.....theoretically we are in the same club but I don't have the expensive shiny toys yet ......

Click to expand...

Could you share some links

Growmau5 said:

I must have missed when and where this was discussed. In everyones experience, are most of the wider profiles from heatsinkusa good to go. No prep other than cleaning, applying thermal paste, and mounting COBs?

Click to expand...

Further back in the thread I was very surprised when I noticed that. I tested temp droop at high dissipation rates to try and exaggerate the difference between stock heatsinks and sanded heatsinks and consistently there was no difference. I also noticed that the stock Rosewill paste, stock Arctic paste (MX4) and PK3 seemed to perform the same even at high dissipation (75%-100% max current) on unsanded heatsinks. Based on that, the difference would be even less running at lower dissipation (12.5-50% of max current where we often run COBs) and even less again in the case of the Vero line.

Apparently, because of the relatively thin interface layer, relatively high cooling performance of our heatsinks, and natural resistance to temp droop of blue/white LEDs, it makes little or no difference to invest much into the thermal interface. Efficiency gains will come from investing into larger/better heatsinks/cooling, bigger/higher bin COBs and/or by running softer to address both current and temp droop.

When testing temp droop of a large COB on a large heatsink, you can see the droop caused by internal thermal resistance and the reistance of the thermal interface because it happens almost instantly. Especially noticeable at high dissipation, the Vf drops instantly and then stays put until the heatsink temp begins to rise. As the heatsink begins to warm up you can see the droop that is caused by the heatsink temp increase. Because of that it should be relatively easy to analyze the two separately maybe I will start trying to measure both.

Bonjour
@SupraSPL could you please give me the specifications for 120mm and 140mm fans, that you advice! (Sorry for my bad English) i know you talk about it and said:use it at 5, 9 or 12V, but all the fans i saw have différent spec.so i would like to know how many cfm/rpm/m2H/sqft2H or any other one mesure you want, i should aim minimum.
Bigger they are and slower they go apparently!?
Have a great day ★

Hi
Is a 140mm fan who turn at 1400RPM and have 87, 7CFM is ok? I found it on aliexpress at 30 $ for 5 fans and they are purple (this is going to be the only purple thing in my grow box...if you know what I mean! )
Have a great day ★

This is the 140mm fan I have the most experience with:

It is a low RPM fan so even at 12V relatively quiet and relatively low power consumption. If you are blowing into a heatsink with long fin channels it might be worthwhile to run it at a higher voltage but if you have short fin channels and just want slight air movement you can run it lower voltage and it will be silent and even longer lasting.

SPECIFICATIONS
Dimensions: 140x140x25mm
Bearing Rating: Sleeve Bearing
Rated Voltage: 12 VDC
Operating Voltage Range: 7 to 13.8 VDC
Rated Current: 0.2 Max Amps
Rated Power: 2.4 Max Watts
Rated Speed: 1000 +/-100 RPM
Airflow: 87 CFM
Noise Level: 18.1 dBa

This Aliexpress fan? Seems like a good value. A higher RPM fan and higher power consumption. You might want to experiment with 5V 7.5V 9V or 12V depending on your setup.

@SupraSPL how long is long fins, and how short are short fins ? Is 35mm long ?

I've been looking at different heatsink possibilities these days, and have been wondering for passive cooling, what width is reasonable in ratio to length.

I found an interesting HS 10cm wide, which seems to fit the bill for passive with 30cm length. (140cm²/W heat)
So ratio width to length is a little stretched ... 1:3
Heat surely sreads and evacuates better, **at equal surface cm²**, with a HS 15cm wide and 20CM long ... am I spliting hairs, or is this ratio worth taking into consideration ?

I've also been wondering where to draw the line on fin spacing for passive.
eg unusual HS : http://www.bal-group.com/heatsink_detail/43 would this work okay passively ? I I think it falls a little short when it comes to base thikness too.

It would be nice as I could use it now with 56% efficiency (semi) actively cooled, and upgrade later with @67%+ efficiency cobs, and then use it passively ...

Any thoughts ?

Bonjour
Yes good question (caus i have a 250X165X35mm h.s to cool a 3590 @1400mA with the 140mm fans we talk about)
Thank you again supra
Have a great day ★