Homebrew PCBs

Hi,

Can someone please help me interpret the results I'm getting from a step
wedge test?

I'm working with "Microtrak" pre-coated boards and JetStar Premium
inkjet film. The problem with Microtrak is that when developed... it
doesn't really change colour (it stays a light yellow-green which is
nearly invisible on top of the copper).
To work around this, I scanned the boards into GIMP, then ran these
filters across the image:
  - Colours -> Components -> Channel Mixer; R=+100, G=0, B=-100,
      Monochrome on.
  - Colours -> Auto -> Stretch Contrast
  - Image -> Mode -> Greyscale

The board was inserted in the UV unit, with the step wedge on top of the
photoresist, and a clear piece of JetStar Premium film on top of the
step wedge.


Here's my first test, done at 340 seconds (standard test time for my UV
box):
  http://img860.imageshack.us/img860/7346/201203240001340secgry.jpg


The first visible step appears to be #4, with #3 only faintly visible.
To me, that suggests I need to adjust the exposure down around two stops:
  340 / sqrt(2) / sqrt(2) = 170 seconds


And here's my second try, at 170 seconds:
  http://img535.imageshack.us/img535/2597/201203240002170secgry.jpg

I also tried immersing the board in etchant for a few seconds -- on the
340-second side, step 4 became more prominent, step 3 and below
disappeared almost completely.

On the 170-second side, step 2 became more prominent (but was only
partly visible). Step 1 was clear (aside from the "Stouffer T2115" text).

Is this roughly what I should be aiming for?

Also, based on the step wedge, it looks like I need a HD-LD of around 5
stops (log-density 0.7) in order to get a reliable transfer. JetStar
provides a measured HD-LD of D=1.75, giving a ~3:1 margin. LaserStar
provides a margin of only 1.414:1... talk about working on a knife edge.

I think I might have to write this up for the PCBWiki :)

Thanks,
-- 
Phil.
ygroups@...
http://www.philpem.me.uk/

On 24/03/12 21:24, Philip Pemberton wrote:
> On the 170-second side, step 2 became more prominent (but was only
> partly visible). Step 1 was clear (aside from the "Stouffer T2115" text).
> 
> Is this roughly what I should be aiming for?

And the answer is.... no!

I exposed a board at 170 seconds. It had a thin film of photoresist
coating the entire board -- almost invisible, except when it was dunked
in the etchant and the copper didn't change colour. No matter what I
did, this film wouldn't go away.

So I figured I was about half a stop low on the exposure:

  170 * 1.414 = 240
  (240-170) / 2 = 35
  170 + 35 = 205 seconds.

Exposing at 205sec produced a good image (the developer turned a VERY
deep blue this time around -- something to watch out for, at least). The
"clear" areas really are clear (down to copper), and the non-clear areas
are covered with photoresist.

Steps 2 and 3 have now merged (3 was just about visible, but washed off
in the etchant; 4 is ~50% covered). So it seems the rules for Microtrak
are quite simple:

  1) If step 2 is visible, you're underexposing.

  2) If you can see the outline of the step wedge, you're DEFINITELY
underexposing.

  3) If in doubt, add half a stop. You most likely won't overexpose the
photoresist, but you will improve your safety margin.

Lord only knows how this applies to negative photoresist. I'll have to
order some from the previously mentioned Greed-bay sellers and try it.

One of these days I'll figure out the secret to this... and maybe write
a little shareware app to calculate exposure figures based on density
values.

-- 
Phil.
ygroups@...
http://www.philpem.me.uk/

Assuming your film wants a "hold at 8 on a 21" (riston films do), you
want the results of your test to show that step 8 had film, but step 9
didn't.  More exposure = higher steps.  So if you're holding step 4 at
340 seconds, and you want to hold step 8, you have to expose for 4x the
time, or 1360 seconds (~22 minutes).

You can get a much better idea of what's happening if you include a
print of some stripes under the wedge; if you then overexpose you'll see
a range of "dark" steps, followed by some striped steps, followed by
some "light" steps.  The number of striped sets tells you your effective
constrast ratio, and you want to adjust your exposure so that range of
steps is centered at *step zero* (step zero is "fully transparent" so
this is what your board will eventually see).  So if you overexpose, and
see stripes on steps 4-8, you want to reduce your exposure by 6 steps.

You do not want the *held* step to be near step zero.  That means you're
just barely exposing it enough.  You don't want "just barely" you want
"just right".

> I think I might have to write this up for the PCBWiki :)

I've written many posts about how to use a step wedge, so check the
archives first and make sure you capture it all :-)

PCBWiki ???
Where would I find this  PCBWiki ??



--- In Homebrew_PCBs@yahoogroups.com, DJ Delorie <dj@...> wrote:

> 
> > I think I might have to write this up for the PCBWiki :)
> 
> I've written many posts about how to use a step wedge, so check the
> archives first and make sure you capture it all :-)
>

First link in Google: http://www.printedcircuitsboards.com/wiki/


On 28/03/2012 21:20, John wrote:
>
>
>
> PCBWiki ???
> Where would I find this PCBWiki ??
>
> --- In Homebrew_PCBs@yahoogroups.com 
> <mailto:Homebrew_PCBs%40yahoogroups.com>, DJ Delorie <dj@...> wrote:
> >
> > > I think I might have to write this up for the PCBWiki :)
> >
> > I've written many posts about how to use a step wedge, so check the
> > archives first and make sure you capture it all :-)
> >
>
> 



[Non-text portions of this message have been removed]

On 28/03/12 21:26, Bert Van Kets wrote:
> First link in Google: http://www.printedcircuitsboards.com/wiki/

Wrong.

http://pcbwiki.philpem.me.uk/

I started this ages ago, but nobody really contributed to it -- it ended
up being more or less a dumping ground for my notes.

You need to register (name and email address) to edit, but that's only
to stop the spambots.

-- 
Phil.
ygroups@...
http://www.philpem.me.uk/

On 28/03/12 07:00, DJ Delorie wrote:
> Assuming your film wants a "hold at 8 on a 21" (riston films do), you
> want the results of your test to show that step 8 had film, but step 9
> didn't.  More exposure = higher steps.  So if you're holding step 4 at
> 340 seconds, and you want to hold step 8, you have to expose for 4x the
> time, or 1360 seconds (~22 minutes).

That's fair enough for a negative resist... the thing is, Fotoboard and
Microtrak are positive-resist boards (they're coated with a liquid
photoresist in various ways; Fotoboard is dip-coated, Microtrak is
roller-coated). Basically, it starts out "hardened" and UV exposure
softens the resist enough to wash away in the developer (typically
sodium metasilicate).

Riston is, IIRC, a negative dry-film resist which hardens when exposed
to UV, and is developed in a sodium carbonate based developer.

Exposing for as long as that will undercut the artwork (if you have a
VERY good phototool master) or just plain fog the whole thing (so all
the resist washes away). It seems with positive resists you want enough
exposure to clear the photoresist from the background, plus about half a
stop.

Thanks,
-- 
Phil.
ygroups@...
http://www.philpem.me.uk/