DIY : Soldering Leds to heatsink ,without MCPCB.Doable ?

stardustsailor

stardustsailor

Well-Known Member
- Epoxy doesn't stick to Anodised Aluminium .


-Almost all alloys of soft alum ( 5xxxx 6xxxx 1xxxx ) they form a protective- to further corrosion -,superficial layer of Alum .oxide ..
Alumina .Which is transparent .
And while soft alum is way heat conductive,alumina has ~10x less thermal conductivity values ...
Alumina is the main reason that aluminium cannot be painted or adhere with glue ,easy ....


-To overcome this problem ,brushing the alum with hard Scotchbrite sponge can remove Alumina
or using a fine ( ~100-200 ) sandpaper is another way ...
But alum reacts with oxygen of air ,and alumina is forming rapidly again ....

-Epoxy will stick ,but eventually it will peel off (as a thin film of ~ 0.1mm ),after a certain period of time ...

-So "passivation" of alum surface is needed ...
Aluminium etch primers ....
Toxic and dangerous.
(Alodine ,ect...Usually part A/first step =defatting sovent +Phosphoric Acid Part B /second step = Chromic +6 acid )
And epoxy has to be applied really soon after the surface priming procedures ....
http://www.jamestowndistributors.com/userportal/show_product.do?pid=12214

But it will adhere ,once and for all ....
.........................................................................................................................................................
Now .....

Turn the epoxy ,into thermal conductive ,dielectric bonding agent...
Tricky .....
Cost Drivers For MCPCBs
Dielectric Substrate
The # 1 cost component of the MCPCB is the dielectric substrate between the copper traces and the metal heatsink / core.
Definition
An insulating medium which occupies the region between two conductors. In this case, the copper circuits and the metal core heat sink.
https://heatsinks.files.wordpress.com/2011/02/mcpcbs-for-led-applications.pdf
Click to expand...
-Thermal conductivity of epoxy should rise from 0.2-0.5 W/m.K to > 0.8 W/m.K

Best market-available epoxies have 0.8 to 3 W/m.K :

Bergquist : 1.1 -2.2 W/m.K
Laird : 2-3 W/m.K
DuPont : 0.8 W/m.K
Fischer Electronik : 0.8 W/m.K
Iteq : 2.2 W/m.K
Taconic : 2 W/m.K
ResinLab : 1-2 W/m.K

Make your own ....
-thermal epoxy has to have the lowest viscosity possible .
(ensuring good wettability ,consistence,coverage and homogeneity -no air bubbles/pockets )


-Normal,low viscosity, liquid epoxy has to be mixed with dielectric ,thermal conductive filler .
Filler load should be low to moderate ( 5-20% of volume ),otherwise the epoxy will become thick (putty..)
Smallest particle Dia of ceramic is desirable ...

Two main ceramics here ..

BeO ( Beryllium Oxide ) with highest thermal conductivity ,after diamond ,
but way toxic ...
https://en.wikipedia.org/wiki/Beryllium_oxide
https://en.wikipedia.org/wiki/List_of_thermal_conductivities

BN ( Boron Nitride )
https://en.wikipedia.org/wiki/Boron_nitride

http://www.asburystore.com/ProductDetails.asp?ProductCode=BN5800-1LB

....................................................................................................................................

-Now ...After those .....
Epoxy will be dielectric and thermally conductive ,while it will adhere to alum surface of heatsink ....



What remains to be solved is the exact procedure of "sandwiching " heatsink -epoxy(dielectric layer ) and copper foil ....

Epoxy should be applied in one-phase or two phase procedure ?

If one phase ,the epoxy has to be applied in relatively thick layer and more "dense" (more BN load ) ,evenly and carefully ,
to the freshly primed heatsink and the copper foil should be layered on top ,and all together , be put inside the hot-press for curing ..
(or clamped and onto the oven ) ...

Thickness of epoxy should be no more of 0.1 mm (100 um ) ..
Heatsink and copper foil should not contact at all ...


If "two phase ",one first layer of epoxy ( ~50 um ) is applied and let to fully cure (harden ).
Then copper foil is layered on top ,with another ~50 um layer of epoxy ...
Again pressed together and let cure in elevated temps ( see epoxy manual for that .. )


....
Then on to copper etching ..

Carefull there also ...
-Iron (III ) Chloride ,should not come in contact with rest of aluminium heatsink ,except copper(& epoxy underneath ) surface ..
 
stardustsailor

stardustsailor

Well-Known Member
Ceramic Heatsinks ..


http://www.power-mag.com/pdf/feature_pdf/1222890448_PEE_Issue_2_2008_Thermal_Management-Ceramic_Heatsink_Provides_Innovative_Thermal_Management.pdf







Made of BeO ,BN ,AlN ,Al2O3,SiC ,ect ....
Almost the same way like pottery ....
(water -binder -ceramic powder -other )

Ceramic sintering

Sintering is part of the firing process used in the manufacture of pottery and other ceramic objects. Substances such as glass, alumina, zirconia, silica, magnesia, lime, beryllium oxide and ferric oxide. Some ceramic raw materials have a lower affinity for water and a lower plasticity index than clay, requiring organic additives in the stages before sintering. The general procedure of creating ceramic objects via sintering of powders includes:

  • Mixing water, binder, deflocculant, and unfired ceramic powder to form a slurry;
  • Spray-drying the slurry;
  • Putting the spray dried powder into a mold and pressing it to form a green body (an unsintered ceramic item);
  • Heating the green body at low temperature to burn off the binder;
  • Sintering at a high temperature to fuse the ceramic particles together.
All the characteristic temperatures associated to phases transformation, glass transitions and melting points, occurring during a sinterisation cycle of a particular ceramics formulation (i.e., tails and frits) can be easily obtained by observing the expansion-temperature curves during optical dilatometer thermal analysis. In fact, sinterisation is associated to a remarkable shrinkage of the material because glass phases flow, once their transition temperature is reached, and start consolidating the powdery structure and considerably reducing the porosity of the material.
There are two types of sintering: with pressure (also known as hot pressing), and without pressure. Pressureless sintering is possible with graded metal-ceramic composites, with a nanoparticle sintering aid and bulk molding technology. A variant used for 3D shapes is called hot isostatic pressing.
To allow efficient stacking of product in the furnace during sintering and prevent parts sticking together, many manufacturers separate ware using Ceramic Powder Separator Sheets. These sheets are available in various materials such as alumina, zirconia and magnesia. They are additionally categorized by fine, medium and coarse particle sizes. By matching the material and particle size to the ware being sintered, surface damage and contamination can be reduced while maximizing furnace loading.
Click to expand...
https://en.wikipedia.org/wiki/Sintering#Sintering_mechanisms




Now ...
Heatsinks made out of the approx or exact same material like the ceramic package of leds ..

http://www.hotfrog.co.uk/Companies/YEG-Opto/Ceramic-Heat-Sinks-36612


http://www.ceramtec.com/ceramcool/

http://www.americanberyllia.com/?gclid=CLHb_ZT557QCFUW_zAod6G0Avw
[url]http://www.americanberyllia.com/BeO_conductivity.html
[/URL]
 
guod

guod

Well-Known Member
for etching copper use ammonium persulfate instead of Iron (III ) Chloride. it´s faster and you can see the process
 
stardustsailor

stardustsailor

Well-Known Member
ganja 2 said:
begin the test
Click to expand...
I'm into it ...

First I'll test some epoxies (liquid -low viscosity )

I.e.
Epoxy #1 Physical Test.
Aluminium not de-fatted,brushed,scraped or treated in any physical or chemical way .
Epoxy layered with soft synthetic brush.
Layer is way thin ( <0.1 mm )
~48 hours after applying epoxy #1 ..
epoxy physical tests.jpg

1)Scrapping :Poor Resistance .Almost peels off .
2) Impact with steel sharp point : Poor .
3)Bending back'n'forth (same as =>elongating & pressing the film ) after >20 cycles back-forth : Good to Excellent .
4)Heating (400°C max for 2' ) : Good to Excellent .Epoxy expanded ,but is elastic enough to compensate with Alum. expansion .
Didn't peel off or show any bubble craters...


I think we'll manage pretty easy ,to make a really decent Circuit Carrier Alum. Heatsink,there ....
Now ...
We need fillers .....
And
Alum.Etch primer ...
 
guod

guod

Well-Known Member
real tough Guys use a mixture of hydrochloric acid and hydrogen peroxide

!! Warning - don´t try this at home !!
 
stardustsailor

stardustsailor

Well-Known Member
Way ....No...I'll stick to persulfates,Guod ...
[video=youtube;BIQKPkABgE4]http://www.youtube.com/watch?v=BIQKPkABgE4[/video]
 
stardustsailor

stardustsailor

Well-Known Member
I want to design some prints for Oslons ....
For new Astir models (high quality parts,there ...Astir OSL 25 and/or OSL 30 .... )

Those SSLs are way tiny ......
Heatsink will seem ...empty ......
Oslon 2.jpg


BTW.....
oslon 1.jpg
(Ganja we need Nitrogen ....For those panels ...)

And ...

Additional Product Features and Benefits:


  • Air flow is adjustable using knob on the side of the power unit
  • Compact design is ideal for soldering, desoldering, and rework of SMT components
  • Uses fast acting shop air or compressed nitrogen
  • Air flow is actuated by a finger switch on the hot air pencil
  • Full range of hot air nozzles is available on page 21
  • ESD safe to protect sensitive components

http://www.lashen.com/vendors/coopertools/weller_high_end.asp
Click to expand...
Hot nitrogen reflow soldering station ...
Weller ..~ 570$ ....
 
loergias

loergias

Member
Hi, I've been checking your write-ups on your development of the Astir© line of LED panels and couldn't help but notice the very evident influences of KNNA and his write-ups on yours. Up until now, everybody's been offering very constructive criticism, as it seems you are taking this to a level beyond mere hobby and into the commercial level. So, is the reason you're trying to make your own thermal pad from scratch purely based on economic viability, as opposed to DIY practicality? Before, you tried to emulate KNNA's work by using Kapton and cheap, hand-mounted LEDs. Couldn't you just use Kapton or another flexible copper clad cb product with your Oslon SSLs? Or does the need for the reflow soldering station and the fact Oslons are SMD make the SSLs less compatible with Kapton? Aren't there other high power, white LEDs that have similar efficacy as the Oslons that can be hand-mounted?

It seems you're trying to improve your panels by giving it extra thermal conductivity mainly at the LED-Heatsink interface with just the right amount of insulation. Am I seeing thing correctly here? I want to construct some LED panels that doesn't require expensive equipment (i.e. ~600 dollar solder stations). Kapton-like products work as far as providing the insulation and thermal resistance. If they work adequately enough in the role they play in dissipating the heat from the LED junction, why not just promote their use for DIYers as KNNA did? Or, are you simply trying to improve upon it further for the commercial success and profitability of your Astir© line of products? If you did develop your own line of more feasible thermal tape, would you market it for less than your competitor's prices on www.customhydronutrients.com???

I don't quite understand the intent of trying to make a start-up company based on the model of DIY proposed originally by KNNA. What benefit does it do for us DIYer's but help drive up the cost of individual LED lamp components, due to the growing demand and large competitors gaining a monopoly over the industry in their continual, unsustainable effort to meet that demand? The currency that a majority of these LED companies seem to dedicate their time and resource to, I believe, is the planned obsolescence and the insufficient technical understanding/know-how possessed by your average horticulturalists to repair and salvage their own broken lamps. So, these large companies bank on that understanding to inflate the prices of otherwise cheaply produced goods and make the lives of DIYers miserable. Their (the large corporation's) credos is to demoralize the ethos of DIY and its rugged attitude towards wasteful, mass consumerism. Sometimes, i wonder if all products that have great potential in the market follow this boom-and-bust trend, giving little or no incentive for DIY enthusiasts to thrive. Something worth pondering... But then again, mass production of LED grow panels overseas in mechanized factories do have their benefits in their Eco-friendliness. It's no wonder China comes out among top in GDP but also for being one of the most polluted and health-hazardous areas to live.

Sorry, if I've gone a little rogue into talking about the economic/business aspect of making LED panels. I still have a great appreciation for all the valuable knowledge and information being transmitted through this forum. It's just a harsh reality when owing up to the Fact: the machines overseas are about a hundred times more efficient in putting together panels that actually work, while asking only a meager pence more and saving consumers all the cost and labor of building and maintaining LED panels for themselves. Please, do take my deconstructive criticism with a grain of salt; I'm sure your LED panels possess something that sets them light years ahead of the competition (both large and small), carrying some kind of future-proof seal of quality. :clap:


-genpah
 
You must log in or register to reply here.
Top