LED DIY : Making of a "DJ SET "-style grow led light .
- Thread starter stardustsailor
- Start date
stardustsailor
Well-Known Member
Now ....
Literally ,across the street at the electronic parts shop ,found some .......
PLAIN METAL CORE PCB (Made in Germany ! )..
Not much info yet ...(In a week or so ....)
But ...some here =>
....
I've DipTrace ...
I've ordered some plain mcpcbs ..
And a laser printer ...
And etching chemical stuff....
Some solder paste ....
Nitrogen is easy to find ....(to create an .."atmosphere" during ..Kitchen pcb hot plating ....)
And some leds .....
I'll try it ....
I 'll give it a go ...
The "French " want just 130 Euros for 3 x 50x60 mm MCpcbs ...& 55 Euro for soldering ,plus ~2Euro per led ...Yeah right ! .
Oslons SSL cost ~ 1 Euro per piece ....(At a german web- shop )....
Yeap ...
I'' try to make two of those lights ...
DIY everything from scratch ....
The whole thing ...
Literally ,across the street at the electronic parts shop ,found some .......
PLAIN METAL CORE PCB (Made in Germany ! )..
Not much info yet ...(In a week or so ....)
But ...some here =>
....I've DipTrace ...
I've ordered some plain mcpcbs ..
And a laser printer ...
And etching chemical stuff....
Some solder paste ....
Nitrogen is easy to find ....(to create an .."atmosphere" during ..Kitchen pcb hot plating ....)
And some leds .....
I'll try it ....
I 'll give it a go ...
The "French " want just 130 Euros for 3 x 50x60 mm MCpcbs ...& 55 Euro for soldering ,plus ~2Euro per led ...Yeah right ! .
Oslons SSL cost ~ 1 Euro per piece ....(At a german web- shop )....
Yeap ...
I'' try to make two of those lights ...
DIY everything from scratch ....
The whole thing ...
guod
Well-Known Member
list of PCB Makers
http://www.mikrocontroller.net/articles/Platinenhersteller
http://www.pcb-pool.com/ppde/order_productconfiguration_js.html#00
the PCB for the Bucks can be made at Home. Material here is FR4
MCPCB or IMS needs a lot of skills...
the alu-core has to be isolated or it is faster away then copper on etching.
Toner-transfer also not easy on a metal-core. you have to heat up the Mcpcb to 180°C for the transfer.
i would start on a prototyp with Stars, test the concept, and than...
i did the same, and now i´m in the market for about 10 MCpcb > PCP-Pool says 280€ for all
20 of them will cost 442€.
going Stars is cheaper but looks not so nice.
http://www.mikrocontroller.net/articles/Platinenhersteller
http://www.pcb-pool.com/ppde/order_productconfiguration_js.html#00
the PCB for the Bucks can be made at Home. Material here is FR4
MCPCB or IMS needs a lot of skills...
the alu-core has to be isolated or it is faster away then copper on etching.
Toner-transfer also not easy on a metal-core. you have to heat up the Mcpcb to 180°C for the transfer.
i would start on a prototyp with Stars, test the concept, and than...
i did the same, and now i´m in the market for about 10 MCpcb > PCP-Pool says 280€ for all
20 of them will cost 442€.
going Stars is cheaper but looks not so nice.
guod
Well-Known Member
)!
stardustsailor
Well-Known Member
With two 48 V 80 mm fans (PELKO MOTORS*),electronics and connector can not stay as they were ...
Maybe moved to the side small angle or to the lid ....Connector facing the sky ...Maybe ...
Needs a bit of thinking ...
..
*Most common automotive , pro-sound and industrial fans,used ..
Their HQ in Canada ,factory in -where else ? - RPC .....
And with a greek CEO!!
( Do I see ancient greek pillars there in the firm's logo ? Forming a " π " letter ?..WTF ? )
Good motors , but they do not deliver the quality of Ebm-papst ..
Nevertheless they also have a really wide selection of fans ,with "weird specs " ,to suit every need....
Motors are certified : http://www.pelkomotors.net/quality/quality.jsp
And belong to the " D" series of DC fans ...
80x80x25 mm D series," power group " H ...
http://www.pelkomotors.net/profile/dcfan/Dseries/D8025.pdf
*Each delivers 61,62 m^3/h ( 1,027 m^3/min *60 min/h ) .
* 3000 rpm
*0,036 A * 48 V = 1,728 W ( ~1,74 says the manufacturer )
*Frame: Thermoplastic, PBT (30%), UL94-V0
*Impeller: Thermoplastic PBT (GF 15%), UL94-V0
*Bearing System: Ball Bearing or Sleeve Bearing
*Protection: Locked rotor protected for minimum 72 hours..
* 32 dB +32 dB = 35 db both working ....
....I'll give 'em a try ...
If they fail to the task ....
Then ,back to the ebm papst range of choices ..
Maybe moved to the side small angle or to the lid ....Connector facing the sky ...Maybe ...
Needs a bit of thinking ...
..*Most common automotive , pro-sound and industrial fans,used ..
Their HQ in Canada ,factory in -where else ? - RPC .....
And with a greek CEO!!
( Do I see ancient greek pillars there in the firm's logo ? Forming a " π " letter ?..WTF ? )
Good motors , but they do not deliver the quality of Ebm-papst ..
Nevertheless they also have a really wide selection of fans ,with "weird specs " ,to suit every need....
Motors are certified : http://www.pelkomotors.net/quality/quality.jsp
And belong to the " D" series of DC fans ...
80x80x25 mm D series," power group " H ...
http://www.pelkomotors.net/profile/dcfan/Dseries/D8025.pdf
*Each delivers 61,62 m^3/h ( 1,027 m^3/min *60 min/h ) .
* 3000 rpm
*0,036 A * 48 V = 1,728 W ( ~1,74 says the manufacturer )
*Frame: Thermoplastic, PBT (30%), UL94-V0
*Impeller: Thermoplastic PBT (GF 15%), UL94-V0
*Bearing System: Ball Bearing or Sleeve Bearing
*Protection: Locked rotor protected for minimum 72 hours..
* 32 dB +32 dB = 35 db both working ....
....I'll give 'em a try ...
If they fail to the task ....
Then ,back to the ebm papst range of choices ..
stardustsailor
Well-Known Member
....
stardustsailor
Well-Known Member
Maybe with a different driver pcb ...
....
And if I've got it right ....
-Two fans on top of each other ,would have increased air pressure ,not air-flow-which is the most wanted variable / factor ....
-Two fans longside to heatsink's fins -one behind the other - would increase ,airflow ...
But over the same fin and fin-channel ,"air current path "...
-Having them wide-side to heatsink (as in model ) airflow is increased again ,but this time ,almost all over ,the width of the heatsink ,thus cooling a larger area ....
2 x 1.75 (say to be a "round " number ..) = 3.5 Watts ...
All in fan power ...
35 dB is an average level of noise ...
"Silent " ain't ....
Neither " a fighter-plane ,taking off " ,though ,sound volume ...
....And if I've got it right ....
-Two fans on top of each other ,would have increased air pressure ,not air-flow-which is the most wanted variable / factor ....
-Two fans longside to heatsink's fins -one behind the other - would increase ,airflow ...
But over the same fin and fin-channel ,"air current path "...
-Having them wide-side to heatsink (as in model ) airflow is increased again ,but this time ,almost all over ,the width of the heatsink ,thus cooling a larger area ....
2 x 1.75 (say to be a "round " number ..) = 3.5 Watts ...
All in fan power ...
35 dB is an average level of noise ...
"Silent " ain't ....
Neither " a fighter-plane ,taking off " ,though ,sound volume ...
stardustsailor
Well-Known Member
Yes ,you're right ...
Exactly ...
Thank you ,once more ..
Don't know what I would've done without your contribution ...
You're tops in engineering ....And not only there...
...
Cool ..90° to the side ..Yes..Of course ! ...
Exactly ...
Thank you ,once more ..
Don't know what I would've done without your contribution ...
You're tops in engineering ....And not only there...
...
Cool ..90° to the side ..Yes..Of course ! ...
stardustsailor
Well-Known Member
I let the old pads stay ..
This way pcb can be placed in many variations in top cover in relation with pot , connector and led cables ...
Something like that ... :
..or with a trace corrected ...
..
(holes need correction also ..What I was thinking ? 1 mm !!!! )
This way pcb can be placed in many variations in top cover in relation with pot , connector and led cables ...
Something like that ... :
..or with a trace corrected .....(holes need correction also ..What I was thinking ? 1 mm !!!! )
stardustsailor
Well-Known Member
And those are about , the two pcbs I need for my version of GD-SDS box ...
(in a corner of A4 sized page )
(led thermal pad connection spokes to copper pour ,are missing )
The smaller -easy- one is made with FR4 ...One piece per box ....
The other one -the really tricky one- has to be made x3 times ...
And is MCPCB ..
Thermal transferring the printed circuit ,doesn't worry me that much ...
Soldering the tiny SSLs with "hot-plating " doesn't worry me either ....
What worries me ,is the soft ally and the chemical copper etching agents ...
Not good combo ,there ...
I keep thinking ....
Alum chemical Isolation ...
Covering with :
Wax,PVA ,silicone,grease, plain acrylic glue...
Even Laser-printer toner powder,as a thick coat ,melted afterwards in kitchen oven forming a solid film ! ...ect...
Need something with the least porosity ...
All of the above resist etching agents ...
And I'm thinking of -if it is enough cheap to do - having them all goldplated,not
for cosmetic reasons ,but for oxidation protection of copper ...
(specially on the leds pcbs ,which have large area of thermal copper pour )...
And of course ,MC-PCB of leds will be cut-off a bit larger in size-before etching- ,so if any ally is lost at sides ,the
final pcb area will remain intact ,with some light "post- etching trimming " work .....
Also ....Etching here is not done by " diving " the pcb in etching agent ....
Here a small size plastic soft art brush ,
will help applying small quantities of the etching agent carefully at a time ,
on appropriate place,until etching is complete ...
Then with cloth/paper wipe ,etching agent is removed ....
After "hand-crafted" micro-etching the mcpcb ,
a lot of warm water will wash away any etching agents residues ...
Then the ally isolation will be removed ...
Holes opened ...
Size trimmed down,with i.e. a sanding belt machine ...Or with Dremmel cutter (the best tool ever ! ) ...
And move on to gold-plating ,maybe ,before soldering the leds ....
(Hot plating in kitchen or with cloth iron ,high precision temp gauge ,timer
and Reflow soldering time pattern of leds-from pdf- ,really well understood and studied...
And try to follow it as close as possible ...)
(in a corner of A4 sized page )
(led thermal pad connection spokes to copper pour ,are missing )
The smaller -easy- one is made with FR4 ...One piece per box ....
The other one -the really tricky one- has to be made x3 times ...
And is MCPCB ..
Thermal transferring the printed circuit ,doesn't worry me that much ...
Soldering the tiny SSLs with "hot-plating " doesn't worry me either ....
What worries me ,is the soft ally and the chemical copper etching agents ...
Not good combo ,there ...
I keep thinking ....
Alum chemical Isolation ...
Covering with :
Wax,PVA ,silicone,grease, plain acrylic glue...
Even Laser-printer toner powder,as a thick coat ,melted afterwards in kitchen oven forming a solid film ! ...ect...
Need something with the least porosity ...
All of the above resist etching agents ...
And I'm thinking of -if it is enough cheap to do - having them all goldplated,not
for cosmetic reasons ,but for oxidation protection of copper ...
(specially on the leds pcbs ,which have large area of thermal copper pour )...
And of course ,MC-PCB of leds will be cut-off a bit larger in size-before etching- ,so if any ally is lost at sides ,the
final pcb area will remain intact ,with some light "post- etching trimming " work .....
Also ....Etching here is not done by " diving " the pcb in etching agent ....
Here a small size plastic soft art brush ,
will help applying small quantities of the etching agent carefully at a time ,
on appropriate place,until etching is complete ...
Then with cloth/paper wipe ,etching agent is removed ....
After "hand-crafted" micro-etching the mcpcb ,
a lot of warm water will wash away any etching agents residues ...
Then the ally isolation will be removed ...
Holes opened ...
Size trimmed down,with i.e. a sanding belt machine ...Or with Dremmel cutter (the best tool ever ! ) ...
And move on to gold-plating ,maybe ,before soldering the leds ....
(Hot plating in kitchen or with cloth iron ,high precision temp gauge ,timer
and Reflow soldering time pattern of leds-from pdf- ,really well understood and studied...
And try to follow it as close as possible ...)
guod
Well-Known Member
wax, grease no go in 45°C hot acid
Paint could go...
or this sort of tape....Name?
Resistant to many acids, bases and solvent
http://www.yatego.com/onlinepack/p,4e447140480b6,4872188a515b37_2,1-pp-paketklebeband-32ac-50my-klebeband-paketband-braun
Paint could go...
or this sort of tape....Name?
Resistant to many acids, bases and solvent
http://www.yatego.com/onlinepack/p,4e447140480b6,4872188a515b37_2,1-pp-paketklebeband-32ac-50my-klebeband-paketband-braun
stardustsailor
Well-Known Member
This simple and ordinary packet-tape ?
!!!
WOW !
Will it work ?
But why wouldn't ?
!!!
WOW !
Will it work ?
But why wouldn't ?
stardustsailor
Well-Known Member
Plenty use this marker widely for pcb DIY making ...
The one with the lit' ball inside (paint as you've stated ,not ink ..)...
MC_PCB etching is more of an "art-work" ...
Brush,wipes ,good lighting ,maybe a large magnifying lens ...
A lot of patience ,time and steady hands ...
But I really do not think that can not be done,with rather great success ...
(Always in small DIY scale making .... )
It needs a bit of skill ....
Not much ,though ...
It'is just another "micro" work ....
This marker might prove an "ace in the sleeve ",to have , for such a work . ....
And at these thermal copper poured mcpcb for leds ,the amount of copper ,that has to go ,is really small ,
contrary to the normal FR4 pcb work ...
The one with the lit' ball inside (paint as you've stated ,not ink ..)...
MC_PCB etching is more of an "art-work" ...
Brush,wipes ,good lighting ,maybe a large magnifying lens ...
A lot of patience ,time and steady hands ...
But I really do not think that can not be done,with rather great success ...
(Always in small DIY scale making .... )
It needs a bit of skill ....
Not much ,though ...
It'is just another "micro" work ....
This marker might prove an "ace in the sleeve ",to have , for such a work . ....
And at these thermal copper poured mcpcb for leds ,the amount of copper ,that has to go ,is really small ,
contrary to the normal FR4 pcb work ...
stardustsailor
Well-Known Member
Can't also the tape ,"turn" at the edges ...?
If pcb is cut -off ~2mm bigger at every side (for safety reasons ) ....
You know what ?
Something tells me ,that this is going to be way easier than we think ....
If pcb is cut -off ~2mm bigger at every side (for safety reasons ) ....
You know what ?
Something tells me ,that this is going to be way easier than we think ....
stardustsailor
Well-Known Member
Also have to study Kitchens/ cloth's iron "attitude " ...
Have to test to find out exact temps at every switch position ,ratio (°C/sec ) of temp rising/dropping ,ect ..
Get to know ,the soldering ..."tools" .....
Do some testing soldering on small pieces of mcpcb with the asian cheapos ,also ....
And then go for the big step ...
Have to test to find out exact temps at every switch position ,ratio (°C/sec ) of temp rising/dropping ,ect ..
Get to know ,the soldering ..."tools" .....
Do some testing soldering on small pieces of mcpcb with the asian cheapos ,also ....
And then go for the big step ...
stardustsailor
Well-Known Member
Guod ,this is the solder paste I got ..
Would it be good enough ?
http://www.termopasty.pl/pg_440_357_3_36_3_easy_print_sn96_5ag3cu0_5.html
Tin-Sn:96,5%
Silver-Ag:3%
Copper-Cu:0,5%
With anti-bubble activator .
Melting point of solder in the paste ...
Exactly at the TL(liquid state -bathing state) for the SSLs ....
..
The -main- thing / procedure ( ambient 25 C to max ~245 ) has to take 8 mins max ...
Generally not a very difficult or complex soldering pattern ...
Lots of time ...
Cooling down is important ...(if done rapidly, solder or led or pcb copper pad,might crack ....) ..
A hair-dresser hot air gun ...
Would it be good enough ?
http://www.termopasty.pl/pg_440_357_3_36_3_easy_print_sn96_5ag3cu0_5.html
Tin-Sn:96,5%
Silver-Ag:3%
Copper-Cu:0,5%
With anti-bubble activator .
| Solder type | for soft soldering |
| Shape | paste |
| Alloy composition | Sn96,5Ag3Cu0,5 |
| Flux type | No Clean |
| Melting point | 217°C |
| Grain size | 25...45 µm |
| Properties | lead free; residues need no cleaning; residues can be cleaned with alcohol-based agen |
Exactly at the TL(liquid state -bathing state) for the SSLs ....
..The -main- thing / procedure ( ambient 25 C to max ~245 ) has to take 8 mins max ...
Generally not a very difficult or complex soldering pattern ...
Lots of time ...
Cooling down is important ...(if done rapidly, solder or led or pcb copper pad,might crack ....) ..
A hair-dresser hot air gun ...