Homebrew PCBs

Didn't have as much time today as i hoped.
I discovered the first PCB i made yesterday was not mirrored when printed  
out - one must mirror PCBs with this method from the normal "see through"  
view! With TT that happened during transfer... I only noticed when i  
finished drilling :-(
Anyway, this PCB should go out in the mail tomorrow so i'm currently  
making a corrected replacement.

I have many more photos to make on friday (no time tomorrow). I only  
managed a few (blurry ones) today while the replacement board was curing:
<http://dipcb.schtuff.com/?action=gallery&q=020806>

James, somewhere i wrote that cutout on the carrier is "edit" long. It is  
90mm.
You wondered about placement of tape - just on the edge to hold the PCB in  
place. probably one would do but i use two. Will be obsolete soon when i  
finish the holder. I made screenshots and i will check on friday what  
images you would further like.

Two more things to think about:
My clothes iron has just about the right temperature for curing on the  
highest setting, i think, must try.
Volkan uses a much thinner layer of ink than i do, i think. Why? Is my  
thick ink layer still needed now that i found out how to cure this  
properly? May it even be detrimental? Must try...
Another note: if the head height is wrong (too close) a vertical line in  
"fast mode" (bidirectional) turns out as two lines.
And another one: with the photo sensor glued to the metal frame where i  
have it the printer starts printing the very top of a page exactly at the  
point where the detected edge is on the carrier.


What i did do however is work on the alignment mechanism.
I installed the plastic guides to test the rail system, and the  
registration in that direction (across the page) is perfect, i can not  
detect and shift between two vertical lines printed one after the other.

OTOH there was a small shift in the lengthwise direction. Apparently the  
printer can not be trusted to use the photo sensor to detect the leading  
edge of the paper precisely. The shift was 1mm worst case, usually less.
My fix for it involves a mechanical stop to push the carrier against.  
Obviously this stop can not be in front of the carrier or the printer can  
not pull it, so the stop is on the back and you first insert the carrier  
further into the printer and then pull it back against the stop. Because  
my carrier is longer than the ways where it rides on i could not simply  
make the stop on the very end but instead had to build it into the rail  
itself. It's simply a piece of PCB glued into the rail and a nut on the  
ways against which it is pushed.

So now i have a printer that prints on exactly the same spot on the  
carrier every time. Should be great for double sided and further stuff  
like soldermask or legend.

Now the next thing is the PCB mounting mechanism. I found quickly just  
putting the thing on the carrier will allow it to slide around even at the  
relatively low speed photo quality setting (so that's why i tape it down  
for now). In draft mode the PCB would probably be a lethal projectile  
catapulted across the room ;-) (Or maybe it would just stay perfectly  
still with the carrier shooting away uder it....)
I have worked out a way to do this, and already built some components ->  
watch this space.

Photos of the carrier rail system will follow soon...
I'll also have to do a lot of PCBs in the next days so plenty of  
opportunity to test!
That's always when i made the most progress with TT, with only a PCB every  
now and then you forget what worked and what didn't.

ST