LED DIY : Making of a "DJ SET "-style grow led light .

[stardustsailor]

My clothes iron hits 207 C on max setting.

Used the EDSYN cr44 on Guod's reccommendation. It was a super silky job (TY guod, if i didnt say already... )

If you are doing a "home job" i would definitely consider the oslon temperature recommendations pessimistically! hitting the limit they set cant be healthy, right?

I've also wondered if the flat plate of the iron, being heated by a coil, has an irregular temerature output across its' surface...

Of course it would 've been a silky job ...
Those SnPbAg solder alloys need shorter preheating/soaking times ..
And they get liquid in lower temps than other alloys ...

Clearly ,Osram ,suggests SnAgCu alloy solder ..(easily understood from solder profile ..) ..

I do not want to question Osram's guidances ...
Or try to find out why they choose such ...
Healthy ?
Maybe on the long run is ....

Maybe solder alloys have different thermal expantion coefficients ,and SnPbAg are more subject to cracking ....
Who knows ?

Thing is that SnAgCu solder makes DIY reflow soldering , kinda more difficult ....

 

[stardustsailor]

This guy is doing really awesome job ......

....

This is approx the size of a Oslon SSL led (3.1 x 3.1 mm with pad of 2.8 x 2.8 mm )..
Look what he has done !
He could have made a pad for a led ,with same size like Oslons ,having more than 10 contacts ..(say ,i.e. a multichip led ...)
Awesome job !

 

[stardustsailor]

Very interesting ....

http://www.sparkfun.com/tutorials/59

I liked specially this one :

" One very old trick for getting an even heat distribution is to put a layer of clean dry sand at the bottom. I think about 1 cm (0.5") would be fine. The sand will spread the heat and you will get a “softer” and more evenly spread heat.
Since the sand will also retain or store heat you will not be able to accomplish any high speed changes in temperature, but I do not think that will be a problem. The sand is dry so there should not be anything to clean from the soldered boards unless some paste has found its way to the bottom side."

Clean dry sand ....

Quartz sand ...or quarzt balls for sterilisation

i.e .
http://www.imaxshop.com/en/sterilization-ultrasonic/quartz-balls-sterilization.html
http://www.imaxshop.com/en/micro-beads-for-apus-quartz-steriliser.html

Powderized perlite will do also,just fine ....

Lay a layer over stove ..
put carefully the mcpcb on to the layer ...
And start the whole procedure ....

(a layer of quartz (SiO ) either is ordinary sand ,pure quartz sand,crushed perlite,or quartz balls ,
will spread heat evenly ,and protect from sudden ramp up or downs of temps ..Very nice old trick ,indeed .... )

 

[stardustsailor]

Ok ....
Guod ,my brother ,say -just say- that you've me convinced about that >60% reds ...
Say ,I 'm going to invest on trying this layout :
......

4 x Oslon SSL 80 LHCP7P-3T (645-660 nm )
&
8 x Oslon SSL 80 LCW CQ7P.CC-JUKQ ( WW 2700K CRI 95 ) .....

And because the heretic inside me doesn't give up easy .....

Say,I wanted one driver to control the 12 reds (with its pot ,of course ) ...
and the rest two bucks ,to control the 24 WW remaining (12 leds each driver ) ..With their-common-pot ,also ...
Something like this :
....

Now ...I need your help ...
What would be the best choices (value-wise..) for :
-R1 (pot for reds ) Vref of one driver ...
-R3 SMD resistor for R1
&
-R2 (pot for WWs ) Vref for two drivers ...
-R4 SMD res for R2 ...

That's about all ..
(Wiring & mcpcb tracing/routing are next problems...)

 

[guod]

use the same values as for the version with 3 drivers.

2.2k pot + 420 R
1.0K pot + 180 R
both are good

 

[stardustsailor]

Today two plain metal core pcbs,have arrived ...
Made in Germany .
Both sides are protected by plastic foil .(transparent on copperside ,blue at alum side .)
Size each : 245 x 175 x 1.2 mm
100um (μm ) dielectric at 1.75 W/m-K.
Oxygen free copper 99.99% 70 um .
..

 

[stardustsailor]

Now ...
From shipping (most probably ) one of the mcpcbs' was slightly bended on one corner ..(from impact ....)
Anyway ....
I've decided to cut it out and try to make one pcb, like the ones the box is going to have ...
-Never-ever made before any kind of this work..Not even one single FR4 ...So first time was on a metal core pcb ....

I've preheated it at the oven at 180°C for 10' ...
Put on the kitchen's hot plate ( at '7'/'9'=max)..
Took the iron and started transferring the laser print ( printed on a magazine's page) .....
Many mistakes ...
Too much heat ...
Too much pressure ...

Some traces melted and "fused" together ,other didn't stick good to copper ....
What the heck ...I've learned a lot .....Next time will be way better ....

So I've started ,using a brush ,to etch the copper ..
Looooooooooooong time .....
What the heck ,who gives a shit about the aluminium .....
Half dipped into etching Sodium Persulfate .....


..After a while .....

...And then ...
My first mcpcb ever ...
(I wanna see yours first mcpcbs ..... )
Anyway ...
It's pathetic ..
But it was just a test,anyway ....



Didn't have any acetone ,but had some fine #1000 sandpaper !!!
........


...Hmmmm..

I'll get it better next time..
I've learned from my mistakes ....

Ohhh.Almost forget .....

Which part of pcb's alum,was dipped in Sodium Persulfate ?

....
Can you tell ?
I can't ......

No problem with alum and sodium persulfate ......


As for the rest .....
Only ,just a glimpse,for now .
3 ,in fact .....

 

[stardustsailor]

Because the print had to be on glossy magazine paper ,I fed it selotaped on a normal matte A4 page ...
But first print was not from the "magazine-page doped" side ..
(Murphy's Law ...)
It'was from the side of plain matte A4 ....
So ....
Just kiddin' around ....
Still learnin' ,though ....

-Matte paper used (holds tiny traces in place ? Doesn't let them "spread " ? )
-Not well preheated iron , not at full temp ....
-not much pressure applied ...
-not much time transferring ....

-Carrier used ?

A non-defatted or sanded (at anyway prepared ) or preheated ....
....Stainless steel ashtray's bottom ....

.....

Hmmm ..Matte paper for mcpcbs and tiny traces ?
.....Hmm....

 

[PSUAGRO.]

is this what your doing with your ashtrays nowadays?

 

[guod]

some tips
thin paper with a silk surface, not glossy. the toner has to stay on it, not go into it.
the transfer need heat and pressure. 2 min. ironing is not to much.
tune your laser for max. toner output.

spraying the paper with a mix of 20-25% aceton and spirit this will merge the toner-particel
you can also buy it...
http://www.reichelt.de/Aetzmittel-Entwickler/TVD-LF-A/3//index.html?ACTION=3&GROUPID=3375&ARTICLE=89770&SHOW=1&START=0&OFFSET=16&&SID=13URrL4X8AAAIAAHq1Jlce2fd2d8fda0720ff3ac812c96c8555bd&LANGUAGE=EN

for cleaning the copper use aceton, it has to be fat free.

 

[stardustsailor]

Silk surface ...?
OI'm trying to find that "silk surface " ..."Glossy but ruberrish "..."Clay filling " ...
Ok ....yeah ..Like I was a paper-surface critic all my life ....
Shit ..What's that silky ?

Anyway ...Even normal matte A4 paper does the job ....
But it needs heat,time and pressure and it is not 100% guaranteed that it is transferred well ..
But I'm trying to find that "non-absorb" paper...

Thanx for the toner output tip and for the link ...

 

[guod]

daily newspapers have a matt surface
expensive magazines have glossy or high glossy surfaces. ..shiny?
silk is betweens this. look at cheap catalogs or flyers.

 

[stardustsailor]

That would do ....
Anyway ...
DIY making a metal core pcb ,isn't so hard after all ...
Tricky yes ,hard no ...

 

[stardustsailor]

Oh..And a small fine cheap saw is needed for fine cutting them ...
Electric saw or dremmel won't do as fine job ...
Or a belt grinder ..That's the best option ,to "trim-shape" final PCB.....

 

[guod]

i do my cuttings with this

http://www.proxxon.com/eng/html/27006.php+
Tungsten tipped saw blade NO 28 017

don´t know if this can cut MCPcb, i work mostly with FR4

 

[stardustsailor]

Now ..
Some tools for the led mcpcb DIYer ......

-A set of small files of different profiles ...(square,round ,flat,ect ) ,to shape curves,offsets ,screw holes /' gulfs ',ect
-A some kind of medicinal/veterinal small pliers ....(A friend vet gave it to me ...)VERY-VERY -useful for really tiny parts ...
Full sense of holding power and the ability to get a grisp/grip of tiny things or ...hot things ....
Really usefull on hot-plate soldering,also....
(<=holding an 1 Watt blue laser..friode ... )

-A fine saw for metal work ....cheap one will do ... )


-A pair of tweezers ...(long and wide opening ) .....



Top Cover for the box .....

Today's visit to local scrapyard ,gave fruit ....
I found a piece(in really bad shape ) of a weird material ..
First time I'm seeing something like this ...
I was told it is used for (mainly ) industrial buildings/facilities construction..
Anyway ..
It is a sandwich of aluminium (metal grey anodised / thick ~0.7 mm ) -High Density Polyamide ~4.6 mm thickness and finally again aluminium of ~0.7 mm thickness (white anodised) ...

Really lightweight and really tough ! ..
I managed to save a flat piece from the really messed-up material ..
200 x 200 mm ..
With an un-avoidable hole ...(it'is covered by small side lide angles -25 x 25 mm..So no problem for air escaping .... )
Anyway ...
Fine material ....
Holes for fans aren't so easy to make ...
Not with an electric saw for sure ...
Won't "make " the "close turns" ,needed ....
With what ?

 

[guod]

...think this is alu-dibond

http://www.modulor.de/shop/oxid.php/sid/1e44221c6800b690139d0c0cb5b6616b/cl/alist/cnid/BFD/tpl/-/lang/1

 

[stardustsailor]

So..it is Pet inside ..That's why it is so strong ...And light ...(really light ... )
Guod ...It is the perfect material for top cover / lid ....
If you can get some to test it out ...
Really good ....

Oh ..I wish we had shops like modulor ....
We are more than 100 years left behind from rest of the world ...
I had to go to scrap-yard ,to see such a thing ,for first time in my life ...
And who knows from where it came ....
...

 

[stardustsailor]

i do my cuttings with this

http://www.proxxon.com/eng/html/27006.php+
Tungsten tipped saw blade NO 28 017

don´t know if this can cut MCPcb, i work mostly with FR4

Ok ....For those who can afford such a beauty on their workbench ...(like me ... )...
A fine saw will do ....And some muscle power ...

(Although there's an old greek saying going like this :
-Tho mouni ke tho prioni ,hopios dhe ta kseree ,idhronee ...-
Meaning : "Saw and Pussy .One who doesn't 'know' them well enough ,he sweats -a lot .... " )

 

[guod]

i know this from our fair-booth. they glue poster on it or they print the pics direkt on it.

hard to work with for nice edges . my top is without the alu and can be cut with a knife.