Homebrew PCBs

Hello,
I was interested in testing performance of various inkjet 
printers and transparency print media. Since I have a HP 692, 
Cannon S330, and Epson Stylus 660 inkjet printers plus various 
brands of transparencies I should be able to make a start on 
comparative testing.

 From what I've learned so far is that a 60X microscope view of 
the print through a incandescent back-light gives very good 
details of the print, and clearly distinguishes differences 
between printers (fitted only with genuine manufacture ink 
cartridges), driver settings, and print media. I have been taking 
digital photos through the microscope eye piece and the image 
quality seems plenty high enough for this purpose. If anyone is 
interested I will put some examples available online.

Now there is the question can a simple back-light images show how 
well it can work as a photomask in UV photoresist exposure? An 
improvement might be to replace the white light with a BLACKLIGHT 
(long wave UV tube with visible light filter). Long wave UV is 
the wavelengths photoresists are most sensitive too. anyone know 
if digital cameras respond to these wavelengths ?  If they don't 
then a phosphor screen or translucent fluorescent colored film 
could be placed directly on top of the print being tested. UV 
light that hits the phosphor or fluorescent pigment will emit 
visible light very closely as "seen" by photoresists during 
exposure. Anyone know where to get some translucent fluorescent 
colored film ?

On Mon, 11 Aug 2003 11:29:21 +1000, you wrote:

>Hello,
>I was interested in testing performance of various inkjet 
>printers and transparency print media. Since I have a HP 692, 
>Cannon S330, and Epson Stylus 660 inkjet printers plus various 
>brands of transparencies I should be able to make a start on 
>comparative testing.
>
> From what I've learned so far is that a 60X microscope view of 
>the print through a incandescent back-light gives very good 
>details of the print, and clearly distinguishes differences 
>between printers (fitted only with genuine manufacture ink 
>cartridges), driver settings, and print media. I have been taking 
>digital photos through the microscope eye piece and the image 
>quality seems plenty high enough for this purpose. If anyone is 
>interested I will put some examples available online.
[question sniped]

I can't help with your question, but I want you to know that I'm very
interested in the results of your tests. Compareing the results of an
Epson Stylus C62 and a HP-DeskJet 900CGi the Epson won hands down. I
figure it must be the ink and hence I think your attempt to research
this with a more sientific aproach could help in sheeding light on
this mystrey.

Hmm, just an idea, but if you manage to describe your test procedure
in a simple enough fashion so as others could repeat the test we all
could help in building up a database or such?

Markus

Adam Seychell wrote:
> Hello,
> I was interested in testing performance of various inkjet 
> printers and transparency print media...
...
> Now there is the question can a simple back-light images show how 
> well it can work as a photomask in UV photoresist exposure?

Maybe you could print a black area a few cm square then
detect how much uv gets thru with a uv photodiode.

http://www.google.com.au/search?q=uv+photodiode&ie=UTF-8&oe=UTF-8&hl=en&btnG=Google+Search&meta=

Some washing powder has a phosphor in it that converts
uv to visible, making white look "whiter".

A photomultiplier tube would be sensitive and has a sharp
cut-off for longer wavelengths.

hi

good idea this testing....
i too do not know if digital cams (ccd or cmos?) can see uv.
they definitely can see IR (good for testing remotes, night vision).

but i fear that they can't see uv. also any plastic protective covering (on 
the ccd chip?) or plastic
lens may shield it....

i have a uv eprom eraser here and a bad webcam at the pc...
i will try ...


looks bad.... the eraser has a uncoated fluorescent tube in it.
so it still produces some visible light...
and on the picture there is not more change on a illuminated white surface 
than visible with the eye...
i don't know if ccd/cmos but very cheap and very bad quality..(but lens is 
glass i think)



i read once some kinds of paper have chemicals in them which convert uv to 
visible.
i read this on a "homebrew co2 laser" page so the light may have to be 
intense to see converted white...
they wrote about the very bright white paper...



please check the net with google for ccd and uv... maybe there is 
something...


stuff that converts uv to visible is enough out there but you won't get a 
clear picture even if you make
some sort of screen....




stefan

Its difficult finding a UV photodiode under AU$100.
There are blue enhanced photodiodes go down to about 350nm with 
%10 sensitivity relative to peak sensitivity.
Farnell have the BPW21R for AU$20 which I can live with.
http://www.vishay.com/document/81519/81519.pdf

Its the 365nm peak emitted from mercury vapor that is most 
important for photoresists. This is just within the passband 
range for standard glass, and the BPW21R photodiode. The inkjet 
transparency film with mylar base and its ink receptive coating 
no doubt has its own influence on the UV.

NEC blacklight tubes (BL-B type) are widely available so thats 
not a problem, http://www.nelt.co.jp/nhe_hp/special/special.htm


I might not be hugely important to get the camera taking images 
from a UV backlighted photomask. More important is knowing the 
average UV absorption from a solid black area of print. This 
absorption figure can be measured with a standard BL-B tube and a 
photodiode like the BPW21R. I'm guessing the relative absorption 
will be measured by taking intensity reading with solid black and 
another reading without print.

A = log(Io/Ii)

where Io= photodiode current without ink
       Ii= photodiode current with ink
       A = relative absorption

I can measure and tabulate values of A for various inkjet 
transparencies / inkjet printers.




Russell Shaw wrote:

> Adam Seychell wrote:
> 
>>Hello,
>>I was interested in testing performance of various inkjet 
>>printers and transparency print media...
> 
> ...
> 
>>Now there is the question can a simple back-light images show how 
>>well it can work as a photomask in UV photoresist exposure?
> 
> 
> Maybe you could print a black area a few cm square then
> detect how much uv gets thru with a uv photodiode.
> 
> http://www.google.com.au/search?q=uv+photodiode&ie=UTF-8&oe=UTF-8&hl=en&btnG=Google+Search&meta=
> 
> Some washing powder has a phosphor in it that converts
> uv to visible, making white look "whiter".
> 
> A photomultiplier tube would be sensitive and has a sharp
> cut-off for longer wavelengths.
> 
> 
> 
> Be sure to visit the group home and check for new Bookmarks and files:
> http://groups.yahoo.com/group/Homebrew_PCBs
> 
> To unsubscribe from this group, send an email to:
> Homebrew_PCBs-unsubscribe@yahoogroups.com
> 
>  
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 
> 
>

Adam Seychell wrote:
> Its difficult finding a UV photodiode under AU$100.
> There are blue enhanced photodiodes go down to about 350nm with 
> %10 sensitivity relative to peak sensitivity.
> Farnell have the BPW21R for AU$20 which I can live with.
> http://www.vishay.com/document/81519/81519.pdf
> 
> Its the 365nm peak emitted from mercury vapor that is most 
> important for photoresists. This is just within the passband 
> range for standard glass, and the BPW21R photodiode. The inkjet 
> transparency film with mylar base and its ink receptive coating 
> no doubt has its own influence on the UV.
> 
> NEC blacklight tubes (BL-B type) are widely available so thats 
> not a problem, http://www.nelt.co.jp/nhe_hp/special/special.htm
> 
> I might not be hugely important to get the camera taking images 
> from a UV backlighted photomask. More important is knowing the 
> average UV absorption from a solid black area of print. This 
> absorption figure can be measured with a standard BL-B tube and a 
> photodiode like the BPW21R. I'm guessing the relative absorption 
> will be measured by taking intensity reading with solid black and 
> another reading without print.
> 
> A = log(Io/Ii)
> 
> where Io= photodiode current without ink
>        Ii= photodiode current with ink
>        A = relative absorption
> 
> I can measure and tabulate values of A for various inkjet 
> transparencies / inkjet printers.

Beware that if a uv photodiode is "slightly" sensitive to IR, it
may still dominate the measurement because atleast one transparency
(epson ones) have low attenuation to IR. It can be tested easily
by biasing an IR photo-interruptor LED so that its detector is in the
linear region, then measuring the output change when some film is
inserted.

I use a HP DeskJet 855C with good results but would like better as I 
do get a few pin holes if exposed too long,  using the HP 
Transparency, set paper type for HP Transparency, Best Printing and 
set to Manual and Intensity to Darkest setting.  I have to print 
from my old machine running Win 95 because my newer machine with XP 
the print driver will not let me do the settings listed above.

Going to be trying a HP 1220 that is at the office and see what it 
does as I think it has better resolution.

Using other brands of transparencies did not work as well on the HP 
855. Get best results with the HP Transparencies.  From what I have 
seen is that you need a transparency that is made for the type of 
ink being used.
A transparency that works well in one brand printer may not work as 
good in another brand of printer.


Ben

Ben H. Lanmon wrote:
> I use a HP DeskJet 855C with good results but would like better as I 
> do get a few pin holes if exposed too long,  using the HP 
> Transparency, set paper type for HP Transparency, Best Printing and 
> set to Manual and Intensity to Darkest setting.  I have to print 
> from my old machine running Win 95 because my newer machine with XP 
> the print driver will not let me do the settings listed above.
> 
> Going to be trying a HP 1220 that is at the office and see what it 
> does as I think it has better resolution.
> 
> Using other brands of transparencies did not work as well on the HP 
> 855. Get best results with the HP Transparencies.  From what I have 
> seen is that you need a transparency that is made for the type of 
> ink being used.
> A transparency that works well in one brand printer may not work as 
> good in another brand of printer.

I've not got around to trying this:

Most inkjets (except epson) use rough transparencies for ink adhesion.
The ones i've looked at were actually smooth film with "sand" like
particles on it. This is the stuff that always gets pinholes, and
i'd guess it happens on the smooth surface between particles.

I've noticed that acetone is a bad thing to clean plastics with,
and that it makes certain smooth clear plastics go rough and milky
white. How about making a "better" transparency? Get some smooth
uncoated stuff and treat it with acetone. The solution might need
to be diluted so you don't end up with home-made glue;)

thanks for the reply,
You are absolutely right about transparency making a difference. 
 From the 8 brands I've tried, each one behaves different. And 
yes, the best transparencies are the ones the manufactures 
recommended. So it not just all to do with marketing and sales.

It seems obvious manufactures will optimize their ink and media 
compatibility. I'm will still be testing non OEM brands.

However , what I have found is the recommended printer setting 
for transparencies is not the best choice for PCB photomasks.
For the Epson Stylus; "Photo Paper", at 720 dpi
The Cannon seemed to print best on "High Resolution Paper" mode.

Th. Epson stuff still wins with darkest and no pin holes.

Adam

> Using other brands of transparencies did not work as well on the HP 
> 855. Get best results with the HP Transparencies.  From what I have 
> seen is that you need a transparency that is made for the type of 
> ink being used.
> A transparency that works well in one brand printer may not work as 
> good in another brand of printer.
> 
> 
> Ben
> 
>

i am not too convinced that the "darkness" is most important to be 
measured...

do you really have problems with the whole printout being not dark enough?

i would be more interested in the quality of the printed tracks (the edges) 
and the equal distribution of the opaqe ink (if it is good too at the 
borders...)

pinholes are interesting too...


and i don't think one can get a smooth surface with acetone treatment.. but 
try....
maybe better then are inks based on solvents themselves...


regards
stefan

Stefan Trethan wrote:
> i am not too convinced that the "darkness" is most important to be 
> measured...
> 
> do you really have problems with the whole printout being not dark enough?

It makes exposure time less critical if the ink blocks better.

> i would be more interested in the quality of the printed tracks (the edges) 
> and the equal distribution of the opaqe ink (if it is good too at the 
> borders...)
> 
> pinholes are interesting too...
> 
> and i don't think one can get a smooth surface with acetone treatment.. but 
> try....

The idea is to *not* have a smooth surface.

> maybe better then are inks based on solvents themselves...

They'd probably clog less and dry quicker.

What I have used other than the HP transparencies which as I said 
seem to work the best in a HP machine.

Used Xerox did not work well in my HP 855, might get by putting two 
copies together but that is a pain to get lined up.

Used Apollo think they have several types, used the one listed for 
HP inkjets, did not work as well as the HP's did but better I 
beleive than the Xerox did.  The Apollo's most likely would need to 
used two copies together.  

The HP's I do with just one copy, but do have to watch the exposure 
time.  I might try some of the other print settings, right now as I 
said I use the HP Transparency setting and set to manual for 
darkest. 

These are the only brands I found at the local office supply, the 
HP's worked so I have stayed with them for now.

The HP the matte finish on the printed side is kind of like it has a 
clear overspray on it, the Xerox was similar but lighter coat and 
feels smoother, the Apollo was more like the HP but think it had a 
rougher finish don't recall for sure.

Anyone know about the HP print driver for XP, as I said it does not 
give me manual control so I print from my old machine on Win 95.  
Seems the HP driver that comes with XP is the only choice available.

Ben

Just ran some test prints on my HP Deskjet 855 using different 
settings, this was using the HP Transparencies.

The best copy was from the HP Transparency, manual with intensity 
set to darkest. Which is what I have been using. Might not pass for 
some people but works for me but would like alittle better, which is 
why I want to try the HP 1220 at the office as it has a higher 
resolution.  Just thought I'd try as I don't think I tried the other 
paper type settings before.

Ben

Stefan Trethan wrote:
> i am not too convinced that the "darkness" is most important to be 
> measured...
> 
> do you really have problems with the whole printout being not dark enough?

Not really.
I've mainly used epson stuff which is very dark, so I've never 
had problems with overexposure. I've also made a couple of boards 
using the HP stuff, which isn't as dark (sorry I can't give you 
relative absorption numbers :)) and also didn't have problems 
with overexposure.

I agree that line edge jaggedness and pin holes are more 
important than overall darkness since many of the inkjets I've 
tested seem to look dark enough so that exposure times can be 
resonably flexable (+-30%). However the most importantly defect 
is banding due to blocked or partly blocked jets. This can easily 
create a discontinuity in a track. Other problems are ink 
splattering which I have seen with some refillable inks on my 
Epson. If the splattering is severe enough it can actually create 
a short between closly spaces tracks.

Another thing I've found is that line edges are a lot more jagged 
on edges that are perpendicular to the travel direction of the 
inkjet head. This effect was noticed on all printers I've tested. 
Some of printing tests were more pronounced than others.


Adam Seychell wrote:

> 
> Stefan Trethan wrote:
> 
>>i am not too convinced that the "darkness" is most important to be 
>>measured...
>>
>>do you really have problems with the whole printout being not dark enough?
> 
> 
> Not really.
> I've mainly used epson stuff which is very dark, so I've never 
> had problems with overexposure. I've also made a couple of boards 
> using the HP stuff, which isn't as dark (sorry I can't give you 
> relative absorption numbers :)) and also didn't have problems 
> with overexposure.
> 
> I agree that line edge jaggedness and pin holes are more 
> important than overall darkness since many of the inkjets I've 
> tested seem to look dark enough so that exposure times can be 
> resonably flexable (+-30%). However the most importantly defect 
> is banding due to blocked or partly blocked jets. This can easily 
> create a discontinuity in a track. Other problems are ink 
> splattering which I have seen with some refillable inks on my 
> Epson. If the splattering is severe enough it can actually create 
> a short between closly spaces tracks.
> 
> 
> 
> 
> Be sure to visit the group home and check for new Bookmarks and files:
> http://groups.yahoo.com/group/Homebrew_PCBs
> 
> To unsubscribe from this group, send an email to:
> Homebrew_PCBs-unsubscribe@yahoogroups.com
> 
>  
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 
> 
>