Homebrew PCBs

Hi guys,

Here's a step wedge from my latest attempt to make Fotoboard2 work properly:
http://img413.imageshack.us/img413/8161/xyzvm.jpg

Exposure time was 340 seconds for Step 1; step 2 is 1.414x less (divide 
time by 1.414), step 2 is another 1.414x less, and so on. Layering was 
as follows:

Base: PCB
L1: OHP transparency, inkjet printed, printed side flat on PCB
L2: Stouffer 21-step exposure wedge

What I'm getting is severe pitting of the photoresist layer, which is 
causing the photoresist to be breached during etching. My ground planes 
look like they've been applied with a sponge...

I've aimed for a Step Held of 8, which worked fine for other photoresist 
laminates (Microtrak and CIF), but doesn't seem to work on Fotoboard2.

I've also noted that the coating quality on FB2 seems to be somewhat 
poorer than MT -- for a start, the coating seems to "blob" and "bubble" 
towards the edge of the PCB, and in a few cases I've had boards where 
patches and strips in the middle have been had photoresist missing. Most 
odd. Although maybe this is down to the coating method used (FB2 is 
roller-coated; MT is dip-coated).

So onto my questions:
   1) Am I doing something wrong here? Should my "step held" be lower 
for this laminate? The image (link above) suggests a step reduction of 
around 3 or 4 (i.e. reduce time from 340secs to 120 or 85 seconds) might 
be more suitable, but this seems a bit low based on my experiences with 
other photoresists.
   2) What would be a typical exposure time for FB2? Mega seem to shy 
away from making any recommendations, simply stating in the datasheet 
that "exposure times will depend on the UV unit and artwork in use".
   3) For those who are using Fotoboard2, what process parameters are 
you using? Exposure time, developer type, that sort of thing would be 
useful.

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

Step 8 is a good exposure point for negative films, I don't know about 
positives.  Based on the photo, I'd shoot for an exposure time of around 
85 seconds - your step 1 (340 seconds) is overexposed, but by step 8 (21 
sec) you're underexposed.

Note that "step held" is the highest step that's still getting barely 
enough light to do something, so in your picture you're holding 
approximately step 12 - the highest step that still shows any trace of 
your inkjet pattern.  Correcting for that also results in an exposure of 
around 85 seconds.

When using a step wedge, you don't always go by which step has the best 
*pattern*, but usually by which step has *anything*.

DJ Delorie wrote:
> Step 8 is a good exposure point for negative films, I don't know about 
> positives.  Based on the photo, I'd shoot for an exposure time of around 
> 85 seconds - your step 1 (340 seconds) is overexposed, but by step 8 (21 
> sec) you're underexposed.

I just did a test run at 75 seconds based on some times I found posted 
to a forum. That worked pretty well -- the board is in the etch tank at 
the moment. It does seem to need a shade more exposure, though -- there 
was a little bit of partly-developed coating left in places. Basically, 
it's turned blue but not detached from the PCB.

> Note that "step held" is the highest step that's still getting barely 
> enough light to do something, so in your picture you're holding 
> approximately step 12 - the highest step that still shows any trace of 
> your inkjet pattern.  Correcting for that also results in an exposure of 
> around 85 seconds.
 >
> When using a step wedge, you don't always go by which step has the best 
> *pattern*, but usually by which step has *anything*.

So let me see if I've got this right... On positive-acting resists, the 
step held is the first step that isn't completely black; on a 
negative-acting resist, it's the first one that isn't completely clear?

Well that explains a lot. It also implies that my exposure times for the 
0.8mm CIF laminate are way off in the weeds somewhere. Although if that 
stuff can handle 4x the required exposure and still produce an 
almost-perfect image, it has a VERY impressive exposure latitude.

I've got a couple of half-inch wide FB2 offcuts in the bin... I might 
have another go at the step-wedge adjustments with the LaserStar film. 
Who knows, maybe my laser printer is putting enough toner down, although 
I bet I'll still need to use my Isograph pen (a Rapidograph with a 
refillable ink tank instead of a cartridge) to fix the pinholes in the 
ground planes...

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

On 22/03/2010 18:04, Philip Pemberton wrote:
> Hi guys,
>
> Here's a step wedge from my latest attempt to make Fotoboard2 work properly:
> http://img413.imageshack.us/img413/8161/xyzvm.jpg
>
> Exposure time was 340 seconds for Step 1; step 2 is 1.414x less (divide
> time by 1.414), step 2 is another 1.414x less, and so on. Layering was
> as follows:
>
> Base: PCB
> L1: OHP transparency, inkjet printed, printed side flat on PCB
> L2: Stouffer 21-step exposure wedge
>
> What I'm getting is severe pitting of the photoresist layer, which is
> causing the photoresist to be breached during etching. My ground planes
> look like they've been applied with a sponge...
>
> I've aimed for a Step Held of 8, which worked fine for other photoresist
> laminates (Microtrak and CIF), but doesn't seem to work on Fotoboard2.
>
> I've also noted that the coating quality on FB2 seems to be somewhat
> poorer than MT -- for a start, the coating seems to "blob" and "bubble"
> towards the edge of the PCB, and in a few cases I've had boards where
> patches and strips in the middle have been had photoresist missing. Most
> odd. Although maybe this is down to the coating method used (FB2 is
> roller-coated; MT is dip-coated).
>
> So onto my questions:
>     1) Am I doing something wrong here? Should my "step held" be lower
> for this laminate? The image (link above) suggests a step reduction of
> around 3 or 4 (i.e. reduce time from 340secs to 120 or 85 seconds) might
> be more suitable, but this seems a bit low based on my experiences with
> other photoresists.
>     2) What would be a typical exposure time for FB2? Mega seem to shy
> away from making any recommendations, simply stating in the datasheet
> that "exposure times will depend on the UV unit and artwork in use".
>     3) For those who are using Fotoboard2, what process parameters are
> you using? Exposure time, developer type, that sort of thing would be
> useful.
>
> Thanks,

I don't have any problems using Fotoboard 2 with my home-made UV 
exposure unit (two tubes about 15 cm from the glass. I found the optimum 
exposure level (11 minutes) by using the old strip technique that used 
to be common in photography. Transparencies are printed on Mega's 
Jetstar Premium film using an HP 5940 printer (1200 dpi). Over exposure 
doesn't seem to matter too much, I've missed the alarm before now and 
probably left it for 20 minutes, on one or two occasions.

I develop in NaOH (12 gm/litre). It takes about 20 seconds.

I've found that the edges of the material sometimes don't work properly, 
probably because it's been exposed to fluourecent lighting, which has 
leaked under the protective coating.

It's important not to use outdated material, the resist seems to go off 
after about six months.

Leon
-- 
Leon Heller
G1HSM

On 22/03/2010 19:17, Philip Pemberton wrote:
> DJ Delorie wrote:
>> Step 8 is a good exposure point for negative films, I don't know about
>> positives.  Based on the photo, I'd shoot for an exposure time of around
>> 85 seconds - your step 1 (340 seconds) is overexposed, but by step 8 (21
>> sec) you're underexposed.
>
> I just did a test run at 75 seconds based on some times I found posted
> to a forum. That worked pretty well -- the board is in the etch tank at
> the moment. It does seem to need a shade more exposure, though -- there
> was a little bit of partly-developed coating left in places. Basically,
> it's turned blue but not detached from the PCB.
>
>> Note that "step held" is the highest step that's still getting barely
>> enough light to do something, so in your picture you're holding
>> approximately step 12 - the highest step that still shows any trace of
>> your inkjet pattern.  Correcting for that also results in an exposure of
>> around 85 seconds.
>   >
>> When using a step wedge, you don't always go by which step has the best
>> *pattern*, but usually by which step has *anything*.
>
> So let me see if I've got this right... On positive-acting resists, the
> step held is the first step that isn't completely black; on a
> negative-acting resist, it's the first one that isn't completely clear?
>
> Well that explains a lot. It also implies that my exposure times for the
> 0.8mm CIF laminate are way off in the weeds somewhere. Although if that
> stuff can handle 4x the required exposure and still produce an
> almost-perfect image, it has a VERY impressive exposure latitude.
>
> I've got a couple of half-inch wide FB2 offcuts in the bin... I might
> have another go at the step-wedge adjustments with the LaserStar film.
> Who knows, maybe my laser printer is putting enough toner down, although
> I bet I'll still need to use my Isograph pen (a Rapidograph with a
> refillable ink tank instead of a cartridge) to fix the pinholes in the
> ground planes...
>
> Thanks again,

I orginally used a laser printer with Laserstar film, but get much 
better results with an inkjet printer. No pin holes, for starters.

Leon
-- 
Leon Heller
G1HSM

Leon Heller wrote:
> I orginally used a laser printer with Laserstar film, but get much 
> better results with an inkjet printer. No pin holes, for starters.

What are you printing on with the inkjet? Tracing paper, JetStar 
Premium, JetStar, normal OHP film or something else?

I find the Stabilo inkjet OHP film works pretty well but is a pain to 
find, and takes a bit of time to dry. I've been trying to get my mitts 
on a pack of 3M CG6000 Universal Inkjet/Laser OHP film, but %#$!ing 
Royal Mail have gone and "lost" it. Grrrr.

I've used JetStar Premium as well, but the coating seems quite thin and 
easy to scratch -- sometimes the act of removing a sheet from the pack 
is enough to scratch off a piece of the coating. The Stabilo OHP film 
doesn't seem to scratch at all, and with LaserStar the toner tends to 
scratch off long before the coating.

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

Philip Pemberton wrote:
> I just did a test run at 75 seconds based on some times I found posted 
> to a forum. That worked pretty well -- the board is in the etch tank at 
> the moment. It does seem to need a shade more exposure, though -- there 
> was a little bit of partly-developed coating left in places. Basically, 
> it's turned blue but not detached from the PCB.

Image here:
http://img403.imageshack.us/img403/3346/xyzbrdtl.jpg

The photoresist seems to get stained by the FeCl - it starts out dark 
green/blue, then goes a shade lighter when exposed to UV. The UV-exposed 
bits go blue/black in the developer before dissolving. Finally, when it 
hits the FeCl3, it goes dark green again... The board seems fine. Not 
quite clear like fresh uncoated FR4 (or -- as I've been told -- FR4 
that's been developed with HCl+H2O2), but not badly stained either.

Exposure was 75 seconds in a Mega AZ210. Image was printed on Stabilo 
inkjet OHP film using a Canon Pixma iP4600. Settings are on the Wiki 
(http://pcbwiki.philpem.me.uk/). Development was CPD5 developer at half 
strength (1 part dev to 9 parts water: 50ml dev, then add water to make 
up to 500ml). I'm not arguing about the reduced strength: it makes the 
developer even cheaper! :)

Even though Mega say using CPD5 with the Fotoboard isn't a good idea, it 
works pretty well. YMMV. From the MSDS, it looks like the Fotoboard 
developer is... well, a half-strength version of CPD5. Sodium 
metasilicate and water, essentially. I still have no idea why the Seno 
developer didn't work, though I might raise the issue with Mega at some 
point this week.

Also, CPD5 doesn't like being left out in the open air -- mix it as soon 
as possible before use, and discard it afterwards. It'll keep for about 
an hour if you're doing a couple of boards (say 3 or 4) but starts to 
lose its potency once you get past an hour or two after mixing. The 
concentrate doesn't seem to suffer from this effect (a bit like ID11 if 
you're into photography: stock solution keeps for a few months, mixed 
solution is 1-shot).

I should have mentioned how I adjusted the exposure... Here goes.

- Expose at 340 seconds and develop.
- Scan on scanner.
     (result: <http://img413.imageshack.us/img413/8161/xyzvm.jpg>)
- Find first patch that shows evidence of background not being 100%
   copper. In this case, that's patch 4 or 5.
- Calculate adjustment factor: sqrt(2)^numSteps. That is, 1.414 ^ 5,
   (where ^ = raise to the power of) or 1.414 ^ 4.
- Apply adjustment factor: 340 / adjFactor. Exposure time is between
   60 and 85 seconds (the decimal is rounded as normal).
- Because the "normal" exposure seems to be in between two steps,
     Exposure time = ((85-60)/2) + 60 = 72.5 seconds
       Rounded up = 73 seconds.]

I made a wild assumption that 75 seconds would probably be fine, and in 
any case it's a nice round number :)

Essentially you're adjusting to use the minimum exposure that makes all 
the exposed photoresist dissolve in the developer.

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

Leon Heller wrote:
> I don't have any problems using Fotoboard 2 with my home-made UV 
> exposure unit (two tubes about 15 cm from the glass.

So a fairly low powered unit then?

The AZ210 I'm using has four 15W Sylvania F15W/350BLB-T8 tubes on each 
side, spaced about 2" apart, with a rather odd metal reflector that 
looks like it's been hit repeatedly with a very small hammer. Distance 
from board to bulb is maybe 1.5" or so, through a sheet of what appears 
to be glass (though is probably quartz-glass or something similar that 
doesn't block UV).

 > I found the optimum
> exposure level (11 minutes) by using the old strip technique that used 
> to be common in photography.

You mean the "test strip" method? Cover 10%, expose for 30 seconds, 
cover another 10%, expose for another 30 seconds, repeat?

That would seem to be the best way to do it -- though the attraction of 
the step wedge (if you can get it working!) is that it gives you an 
exact adjustment and with 1 or 2 adjustments (one small offcut from a 
double-sided board) you can get an exact figure.

In this case it looks like the issue was a PEBKAC Fault on my part :)

 > Transparencies are printed on Mega's
> Jetstar Premium film using an HP 5940 printer (1200 dpi). Over exposure 
> doesn't seem to matter too much, I've missed the alarm before now and 
> probably left it for 20 minutes, on one or two occasions.

With a base exposure time of 11 minutes, you're still only talking about 
a 2x over-exposure. In my case I had nearly a 5x over-exposure...

What surprises me even more is that the CIF laminate worked under these 
conditions... I've pulled out my lab notebook and it's there in black 
and white: 340 seconds, AZ210 UV box, developed with Seno developer roller.

> I develop in NaOH (12 gm/litre). It takes about 20 seconds.

I'm not touching NaOH with a barge-pole... or at least a full PPE kit.

> I've found that the edges of the material sometimes don't work properly, 
> probably because it's been exposed to fluourecent lighting, which has 
> leaked under the protective coating.

It looks to me like the coating isn't going all the way to the edges. 
I've seen instances where the edge ~1cm is a shade lighter than the 
centre (which does appear to be light exposure) but the major defect 
with my boards is that the coating doesn't actually go all the way to 
the edge.

I haven't had any similar issues with the CIF board (though that was a 
0.8mm RF laminate) and it turns out what I thought was Microtrak 
actually wasn't... so I haven't tried that either. Hey-ho...

> It's important not to use outdated material, the resist seems to go off 
> after about six months.

Is that all? Huh.
Though I did have some Maplin photoresist board here that has utterly 
lost any function whatsoever. I ended up running the timer on the AZ210 
at 999 seconds (the maximum it'll go to), and even step 1 on the test 
wedge was only just going grey. That was with no artwork film...

I ended up stripping the photoresist and using it for toner transfer. 
I'm toying with the idea of buying some spray-on (or liquid) photoresist 
to re-coat a few boards, but it's not a high priority at the moment. I 
think I'd strap the board to a 120mm cooling fan, put the contraption in 
a shoebox, then set it spinning and spray it. In theory, that should 
produce a fairly even coating of photoresist. But again, it's not high 
on my agenda -- wet-laminating sheets of Riston onto blank boards seems 
like a more profitable venture (in terms of likelihood of success).

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

On 22/03/2010 20:02, Philip Pemberton wrote:
> Leon Heller wrote:
>> I orginally used a laser printer with Laserstar film, but get much
>> better results with an inkjet printer. No pin holes, for starters.
>
> What are you printing on with the inkjet? Tracing paper, JetStar
> Premium, JetStar, normal OHP film or something else?
>
> I find the Stabilo inkjet OHP film works pretty well but is a pain to
> find, and takes a bit of time to dry. I've been trying to get my mitts
> on a pack of 3M CG6000 Universal Inkjet/Laser OHP film, but %#$!ing
> Royal Mail have gone and "lost" it. Grrrr.
>
> I've used JetStar Premium as well, but the coating seems quite thin and
> easy to scratch -- sometimes the act of removing a sheet from the pack
> is enough to scratch off a piece of the coating. The Stabilo OHP film
> doesn't seem to scratch at all, and with LaserStar the toner tends to
> scratch off long before the coating.
>

I print on Jetstar Premium, and let it dry for an hour or so. I handle 
it carefully and don't have any problems with scratches.

Leon
-- 
Leon Heller
G1HSM

On 22/03/2010 20:32, Philip Pemberton wrote:
> Leon Heller wrote:
>> I don't have any problems using Fotoboard 2 with my home-made UV
>> exposure unit (two tubes about 15 cm from the glass.
>
> So a fairly low powered unit then?
>
> The AZ210 I'm using has four 15W Sylvania F15W/350BLB-T8 tubes on each
> side, spaced about 2" apart, with a rather odd metal reflector that
> looks like it's been hit repeatedly with a very small hammer. Distance
> from board to bulb is maybe 1.5" or so, through a sheet of what appears
> to be glass (though is probably quartz-glass or something similar that
> doesn't block UV).

Won't you get a lack of collimation with such a short distance?

Leon
-- 
Leon Heller
G1HSM

Leon Heller wrote:
> Won't you get a lack of collimation with such a short distance?

Apparently not... I haven't noticed any issues with poor line definition 
(which AIUI is the main effect of poor collimation). It certainly goes 
down to 10:10 track:spacing (mils), and one of my boards (a GPS demo 
board) had 8mil tracks.

The board I'm working on at the moment has 10:10 around an SMPSU 
controller, then 16 thru 24mil tracks with variable spacing everywhere 
else (I'm trying to keep track and spacing widths as high as possible).

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

OK, I think I may have a retraction to publish....

I've just finished re-doing the Fotoboard step-wedge tests using 
LaserStar film. Seems my laser printer *is* capable of producing an 
acceptably dark black, and that the issues with the Seno developer not 
working were entirely down to setting the timer on the UV box far too 
high and burning the heck out of the photoresist.

For anyone who's interested, the exposure times that worked for me are 
as follows:
   with Stabilo Inkjet OHP film: 75 seconds
   with LaserStar: 170 seconds

Inkjet printer is a Canon Pixma iP4600.
Laser printer is a Kyocera FS-C5200dn.

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

> 
> The AZ210 I'm using has four 15W Sylvania F15W/350BLB-T8 tubes on each 
> side, spaced about 2" apart, with a rather odd metal reflector that 
> looks like it's been hit repeatedly with a very small hammer. Distance 
> from board to bulb is maybe 1.5" or so, through a sheet of what appears 
> to be glass (though is probably quartz-glass or something similar that 
> doesn't block UV).
> 
>


Ordinary glass does not block much of the long wave UV used here, a quartz pane would be very expensive.

/BL tubes are a better choice than /BLB. The Woods glass used to block visible light in the /BLB also blocks some of the UV output.

James wrote:
> /BL tubes are a better choice than /BLB. The Woods glass used to block visible light in the /BLB also blocks some of the UV output.

These aren't Wood's Glass -- I've got an F4T5/BLB somewhere, which is a 
deep purple/black colour when switched off; these tubes are off-white 
and look like normal white-light fluorescent tubes. Well, aside from the 
blatantly obvious "WARNING: EMITS UVA LIGHT WHEN OPERATING -- AVOID 
PROLONGED EYE OR SKIN EXPOSURE" warning on one end of the tube.

According to the specsheet I've got, "350BLB" means it's a 350nm 
blacklight (near-UV / UVA) tube with a strong UV emission. Pretty much 
exactly what you'd want for photoresist. Sold for use in fly-killers, 
interestingly enough.

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