Yahoo Groups archive

Homebrew PCBs

Index last updated: 2026-04-28 23:05 UTC

Message

Re: [Homebrew_PCBs] Gluing 2 single sided boards to make a double sided board.

2015-07-05 by Petar Delic

Hi,


It is not worth the effort.


The best way of doing it is just drilling 3 referent holes (that should 
be placed in advanced in layout) and using them for mask alignment for 
both side of double sided boards.


I tried to make multilayer boards and I almost done it (I have diy CNC 
and heating press with 20 tons for gluing cores and prepregs and 
chemical process for copper plating holes....).


It is interesting that my CNC has 0.01 mm precision but holes positions 
weren't precise (let say that the worst precision was 0.1 mm). I later 
discovered that I can use some aluminum sheet to enhance hole positions ...


Prepreg is a uncured laminate that acts as a glue (but in my case I 
manage to get 20 years old prepreg.... and actually it was working where 
FR4 prepreg faced the FR4 core, despite 20 years old the it glued well, 
but where prepreg glued to copper that was weak.... and I had problems - 
micro air bubbles where prepreg meets the copper and when I tried to 
solder the board in the oven it slightly delaminates ....).


I had 3 major problems: weak connection between prepreg and copper (I 
tried to make the copper surface rough by micro etching but that didn't 
work), weak connection between copper that plates the hole and the hole 
wall (I didn't find out the stuff that dissolve FR4 and make surface 
rough for better plating, well... I didn't find commonly available stuff 
for dissolving FR4).


Other than that my multilayer boards prototypes worked even at 6 Gb/s 
speed (I made some FPGA transceivers small boards....).


Thanks,
Petar










On 07/05/2015 07:13 AM, James bitsyboffin@... [Homebrew_PCBs] wrote:
> Double sided boards as we all know can be a bit of a pain, the biggest 
> problem is getting that precise alignment between sides correct so 
> that when you drill in the center of your pad on one side, it comes 
> out in the center of the pad on the other (and not, as is not 
> uncommon, breaking out, and probably messing up the trace, or a trace 
> nearby...).  Especially if your only able to expose/imprint artwork 
> one side at a time.
>
> Even more frustrating when one side comes out perfect in part of the 
> process (eg, developing) and the other side is too flawed to be 
> useful, meaning you get to start from scratch again.
>
> So, the thought randomly occurred to me today, if you could prepare, 
> expose, etch, drill the two sides of a double sided board completely 
> separately, and join them up at the end of the process, it would make 
> getting precise alignment of those sides almost a non-issue, assuming 
> you can drill even roughly vertically and hit the center of the holes 
> it's just a matter of lining up the holes you drilled on each side.  
> And also means that the success or failure of each side is independent 
> of the other.
>
> Of course, it does double the drilling, but that's the easy bit really.
>
> This is also of course more or less how multi layer PCBs are made by 
> fabricators, a stackup of already prepared pcb layers and resin 
> impregnated sheets.
>
> Single sided laminate at 0.8mm thick (and you can get down to 0.5mm) 
> is available out of China for cheap.
>
> Eg:
> http://item.taobao.com/item.htm?spm=a1z10.5-c.w4002-1192550948.12.TIEXsq&id=35870048695
> http://item.taobao.com/item.htm?spm=a230r.1.14.4.VdTsRT&id=21105435798&ns=1&abbucket=3#detail
>
> Has anybody done this... am I reinventing a triangular wheel here, is 
> it doomed to failure? Suggestions on suitable adhesive... maybe just 
> spray on contact adhesive might work, it's a pretty big flat contact area.
>
>
>

Attachments

Move to quarantaine

This moves the raw source file on disk only. The archive index is not changed automatically, so you still need to run a manual refresh afterward.