Printing 1:1 Artwork.
Tony Williams
I posted a request here recently requesting advice on the best
device for doing 1:1 artwork.
Many thanks to all who kindly replied, the Net does seem to be
a nice club.
I decided and have placed an order for the H-P Laserjet 6P with
2meg standard memory. The reasons for this choice are:-
1. It's 600x600 dpi and if your plot exceeds the memory it just
goes down in resolution. This is unlike some other printers
where I was told that if you exceed the memory they will plot
as much as they can on the first page and the rest on the
next page, and so on. ( What a bummer for Artwork!)
2. You only need to plug in standard SIMMs to increase memory.
3. It is supposed to print nice deep black in filled areas.
I also have some layout software, Easy-PC Pro, that allows you
to do a test plot, measure the results, and feedback the
neccessary X-Y corrections for fine tuning of the final plot.
Here's hoping and thanks again.
--
[Tony Williams, Ledbury, Herefordshire, UK.---Pagewidth=64-----]
Robert B. Hoffmann
Great choice! Might I also recommend checking out some of the higher
quality papers. We did this type of thing years ago and found that
the better "Laser" papers were definitely worth the money for fine
line designs. Nekoosa LASER 1000 was the paper of choice then. We
were able to use paper 1:1 check plots which were sent Fed-Ex to
overseas vendors while data was electronically sent. We had been
sending film, but for some reason film was delayed through customs where
the paper plots were not.
--
Robert B. Hoffmann IT Engineer
Hewlett Packard Company Boise Site DTC
Boise, Id 83714 Bob_Ho...@hp.com
Tony Williams
In article <3267BA...@hp.com>, Robert B. Hoffmann
<URL:mailto:Bob_Ho...@hp.com> wrote:
>
> Tony Williams wrote:
[snip]
> > I decided and have placed an order for the H-P Laserjet 6P with
> > 2meg standard memory. The reasons for this choice are:-
> >
> Great choice! Might I also recommend checking out some of the higher
> quality papers. We did this type of thing years ago and found that
> the better "Laser" papers were definitely worth the money for fine
> line designs. Nekoosa LASER 1000 was the paper of choice then. We
> were able to use paper 1:1 check plots which were sent Fed-Ex to
> overseas vendors while data was electronically sent. We had been
> sending film, but for some reason film was delayed through customs where
> the paper plots were not.
>
Nope, I'm going for film in order to try and get directly usable
artwork. For prototype pcbs I need to break these upfront costs
for Photoplotting+Tooling.
Bearing in mind my Public H-P Plug and your obvious approval,
do you think a Lifetime Free Service Contract could be in the
post very soon? It could also be useful gag-insurance in case
my brand_new_and_virgin H-P 6P turns out to be a real dog.
Well, it's worth a try.
Vincent Himpe
Tony Williams wrote:
>
> I posted a request here recently requesting advice on the best
> device for doing 1:1 artwork.
>
> Many thanks to all who kindly replied, the Net does seem to be
> a nice club.
>
> 2meg standard memory. The reasons for this choice are:-
>
> goes down in resolution. This is unlike some other printers
> where I was told that if you exceed the memory they will plot
> as much as they can on the first page and the rest on the
> next page, and so on. ( What a bummer for Artwork!)
>
> 2. You only need to plug in standard SIMMs to increase memory.
>
> 3. It is supposed to print nice deep black in filled areas.
>
> I also have some layout software, Easy-PC Pro, that allows you
> to do a test plot, measure the results, and feedback the
> neccessary X-Y corrections for fine tuning of the final plot.
>
> Here's hoping and thanks again.
> [Tony Williams, Ledbury, Herefordshire, UK.---Pagewidth=64-----]
I tested 15 different printers about one year ago for exactly the same
purpose (making pcb artwork. Laserprinters are crap !. I have found none
that can deliver the quality i need. (i did not test the hp 6P however)
the main problem of a laser is distortion. the paper stretches in the
length , or the laser is not correctly adjusted so the width is not
correct. Another problem is the drum in the laserprinter. somethimes
this is not equally round but a little bit oval , leading to distortions
in the length.
My ultimate choice was a hp 860 inkjet printer. It delivers 600 dpi
artwork.
When i print on the hp inkjetpaper i get the best quality.
the hp transparancies delivered for this printer can be used to make
your pcb directly using uv light.
The kind of things i can do : make pcbs with plcc84 bodies and tracks
between the pins , etch pcb with pqfp 208 pins components.
No problemo . and all from the output of the hp.
what you should do for a test is print a page of pads (diameter 60 mils)
and a distance of 50 mils apart. fill up the complete page with it.
Put a ruler on the paper and check the spacings of the holes. You will
be surprised !.
hope this helps
regard
Vincent Himpe
--
-------------------------------------------------------------------------------
The above information was provided by :
Vincen...@ping.be
http://www.ping.be/~ping0751
http://www.Geocities.com/SiliconValley/6731
** Disclaimer ** The nonsense above may be totally
wrong.
Use on your own
responsibility.
In case you should find typing or spelling errors: you can keep them.
--------------------------------------------------------------------------------
Ross LaGue
In article <ant26083...@ledelec.demon.co.uk>, Tony Williams
<to...@ledelec.demon.co.uk> wrote:
>
> The 6P is not yet connected up to the computer, but a few
> prints were done with the self-test. Paper printing is
> as good as you would expect from a laser, crisp blacks, etc.
>
> Film printing is good and I expect to be able to plot down to
> trackwidths of nearly 10thou. There is an almost straight
> path for the film and it is not excessively hot as it exits.
>
> I held the plot up to the light and there is some sign of
> light speckling, ie pinholes in areas of black...that's not
> so good news on a brand new printer but probably workable.
> Held up to light the black is not 100% but very good and
> fiddling with the exposure time will be minor, if at all.
> The black is relatively easy to scratch off so direct use
> for pcb manufacture is limited.
>
What type of film are you using?
--
Ross LaGue < r...@erinet.com > Dayton, Ohio
Tony Williams
> Tony Williams wrote:
>
> I posted a request here recently requesting advice on the best
> device for doing 1:1 artwork.
>
> Many thanks to all who kindly replied, the Net does seem to be
> a nice club.
>
> I decided and have placed an order for the H-P Laserjet 6P (etc).
My 6P arrived yesterday, looks a nice piece of kit, built like
the proverbial battleship. It came with a collection of floppies
with various drivers for the PC and a printer lead was thrown in.
The introductory manual is ok, with lots of pickies for us
simpletons. Regrettably that's the limit of documentation that
comes with it. I would have expected more for this money, at
least something on PCL5.
The 6P is not yet connected up to the computer, but a few
prints were done with the self-test. Paper printing is
as good as you would expect from a laser, crisp blacks, etc.
Film printing is good and I expect to be able to plot down to
trackwidths of nearly 10thou. There is an almost straight
path for the film and it is not excessively hot as it exits.
I held the plot up to the light and there is some sign of
light speckling, ie pinholes in areas of black...that's not
so good news on a brand new printer but probably workable.
Held up to light the black is not 100% but very good and
fiddling with the exposure time will be minor, if at all.
The black is relatively easy to scratch off so direct use
for pcb manufacture is limited.
I tend to take secondary masters as working positives and
correction for these minor probs are easy at that stage.
As a little experiment I fed the same film back through again.
Results were remarkably good, less than 5thou misalignment of
the overprint.
Suggestion to Mr H-P, if you are reading this.... For us Users
who have to print artwork there is a good case for printers to
have the facility to print ** Mirror Images **. Since that
could be done in the printer software it would be a nice cheap
extra feature for you to implement.
Dave Baldwin
Tony Williams (to...@ledelec.demon.co.uk) wrote:
: The introductory manual is ok, with lots of pickies for us
: simpletons. Regrettably that's the limit of documentation that
: comes with it. I would have expected more for this money, at
: least something on PCL5.
The PCL manual is about two inches thick.
: I held the plot up to the light and there is some sign of
: light speckling, ie pinholes in areas of black...that's not
: so good news on a brand new printer but probably workable.
: Held up to light the black is not 100% but very good and
: fiddling with the exposure time will be minor, if at all.
The 'standard' method with photocopies is to make two copies and overlay
them because the pinholes almost never line up.
: Suggestion to Mr H-P, if you are reading this.... For us Users
: who have to print artwork there is a good case for printers to
: have the facility to print ** Mirror Images **. Since that
: could be done in the printer software it would be a nice cheap
: extra feature for you to implement.
My PCB software does that (Tango for DOS) along with scaling and X/Y
compensation. For scale checking, a line grid of minimum width lines one
inch apart will allow you to check the print accuracy. You might also
check to see what happens when you print two or three copies in a row.
Since the printer heats up, the second two copies might align better than
the first two.
--
-=-=-=-=-=-=-=-=-=-=-=- Check out DIBs and TCJ -=-=-=-=-=-=-=-=-=-=-=-=-
Dave Baldwin: dib...@netcom.com | The Computer Journal 1(800)424-8825
DIBs Electronic Design | Home page "http://www.psyber.com/~tcj/"
Voice : (916) 722-3877 | Hands-on hardware and software
TCJ/DIBs BBS: (916) 722-5799 | TCJ/DIBs FAX: (916) 722-7480
-=-=-=-=-=-=- @#$%^&* I can't even quote myself! Oh,well. -=-=-=-=-=-=-
Pascal Dornier
Vincent Himpe wrote:
> I tested 15 different printers about one year ago for exactly the same
> purpose (making pcb artwork. Laserprinters are crap !. I have found none
> that can deliver the quality i need. (i did not test the hp 6P however)
>
> the main problem of a laser is distortion. the paper stretches in the
> length , or the laser is not correctly adjusted so the width is not
> correct. Another problem is the drum in the laserprinter. somethimes
> this is not equally round but a little bit oval , leading to distortions
> in the length.
> My ultimate choice was a hp 860 inkjet printer. It delivers 600 dpi
> artwork.
> When i print on the hp inkjetpaper i get the best quality.
> the hp transparancies delivered for this printer can be used to make
> your pcb directly using uv light.
> The kind of things i can do : make pcbs with plcc84 bodies and tracks
> between the pins , etch pcb with pqfp 208 pins components.
> No problemo . and all from the output of the hp.
I use laser printouts for checking. Not perfect, but in most cases I
can see whether footprint has the correct spacing etc.
In my opinion, the best process (but somewhat expensive) could be a
modified thermal transfer printer (normally used for printing barcode
labels etc). Either print directly to the PCB, or print _negative_ to
paper, and iron the remaining ink from the ribbon tape to the PCB.
I never tried this, but I think it should work.
--------------------------------------------------------------------
Pascal Dornier pdor...@best.com http://www.best.com/~pdornier
PC Engines PC hardware + embedded PC consulting
--------------------------------------------------------------------
Terry Hicks
I agree. The only observation I would make is that the Epson color stylus (at
least the original one) works well also. An additional check I would make is for
consistency of opaqueness. The inkjets have a constant level of opaqueness. Also
important is the degree of stretch. It must be consistent. The laser printer
stretches in a non-consistent manner. As a result the laser printer is acceptable for
single sided boards with narrow traces (up to about 0.050" maximum). This is true for
small I.C.s only. Note that laser printers stretch almost solely along the paper
path. The inkjet is a non-thermal process and thus minimizes most of these problems.
It also appears to me that I am talking about positive type PCB processes. For
negative processes the narrow trace limitation might be resolved by flowing a ground
plane accross the board.
Don Yuniskis
In article <dibaldDz...@netcom.com>,
Dave Baldwin <dib...@netcom.com> wrote:
>Tony Williams (to...@ledelec.demon.co.uk) wrote:
>
>: I held the plot up to the light and there is some sign of
>: light speckling, ie pinholes in areas of black...that's not
>: so good news on a brand new printer but probably workable.
>: Held up to light the black is not 100% but very good and
>: fiddling with the exposure time will be minor, if at all.
>
>The 'standard' method with photocopies is to make two copies and overlay
>them because the pinholes almost never line up.
Or, use thermal dye transfer technology -- black is BLACK!
>: Suggestion to Mr H-P, if you are reading this.... For us Users
>: who have to print artwork there is a good case for printers to
>: have the facility to print ** Mirror Images **. Since that
>: could be done in the printer software it would be a nice cheap
>: extra feature for you to implement.
This isn't as simple as it sounds! Depending on the technology
employed in the print engine, simply inverting the "black" and "white"
signals often results in output that is NOT exactly the same as
a true "mirror image"! Some print engines "write white", others
"write black". Some technologies dots "grow", etc. For traditional
applications (text, graphics, etc.) the results are usually not
visible (though there are some acid tests for these things). But,
when you're printing tiny traces (even at 2:1 !) this can be a
significant factor (underwidth or overwidth traces, etc.) -- consider
that traces don't always "line up" with the elements in the print
head so that can cost you as much as one "dot width". And, a 10 mil
trace is already only 3 dots at 300dpi, etc.
Something with more "global" knowledge of the page's contents is more
appropriate -- and, models for each targeted printer, etc.
>My PCB software does that (Tango for DOS) along with scaling and X/Y
>compensation. For scale checking, a line grid of minimum width lines one
>inch apart will allow you to check the print accuracy. You might also
>check to see what happens when you print two or three copies in a row.
>Since the printer heats up, the second two copies might align better than
>the first two.
You may also find that large areas of black cause the printer to have
trouble delivering enough toner and "fusing" heat, etc.
--don
Tony Williams
Thanks for this excellent feedback on this topic of 1:1 printing.
It is nice to see other opinions and/or bitter experiences.
Various replies are as follows:-
Vincent Himpe suggested avoiding Lasers in favour of Inkjet
and suggested a serious check on the film distortion....
Too late now, I've spent the money on the 6P. I am coming from
an inkjet, Stylus800, and did not have good experiences with
Artwork. Mine may have had an evil streak but it was noticeable
that when anything serious needed to be plotted that was the
time to act the goat... generally a blocked jet, etc. My feeling
is that the Stylus is ok when brand new but does age quickly.
Non-waterproof ink is a bummer as well.
----------------------------------------------------------------
Ross LaGue queried the film used....
I tried everything to hand, Inkjet, OHP, and even Mylar dwg film.
This is going to be a matter of experiment for least distortion
but the exit temperature does not seem that great.
----------------------------------------------------------------
Dave Balwin commented that the PCL Manual is 2" thick and the
fact that Tango does mirror-imaging within itself.
Manual first. This is something that is starting to irritate me.
Years ago, when you bought any piece of kit you always expected
and got decent technical documentation. Today, all you get is
a thin little idiot's intro, at best. For example, when trying
to optimise the Stylus performance I have no decent Stylus Manual
and have to work from an ancient Panasonic printer manual, a
nice 0.75" thick manual that just came with it, with everything
you need to know, aah......those were the days.
Ok, so the PCL Manual is 2" thick but H-P supplies 5/6 Floppies
with drivers on them, so another Floppy with the Manual on would
not be too much to ask. Alternatively do what US Robotics do,
allow full technical information to be downloaded from their
bulletin board. Either of these options would not cost the
supplier very much and puts the printing cost into the hands
of whoever does *need* to have a Manual. That would seem fair.
Now to mirror-imaging. The need for it is obvious to anyone who
has to take contact prints downstream and the requirement to
mirror depends on things like top/bottom of a pcb or whether
you are going towards silk-screening or not, etc.
My Artwork will come from a pcb-package, a dwg-package or even
a desktop publisher and only one of these has a mirror facility.
I do understand that the problem really lies with the software.
If you earn your living off CAD then directly useable hard-copy
is the name of the game. Having to get up to antics to get your
invoiceable product destroys the Work Amplifier concept. I wish
some of these software writers would understand that and not add
the plot/print function as a limited-facility afterthought.
So I simply remarked that, because mirroring is one of those
things that is really useful when needed, then having it in
the printer would seem a good idea.
--------------------------------------------------------------
Mind you, whilst on the subject of absurd lust-lists, a facility
for "reverse-video" printing might not come amiss also.
Ross LaGue
In article <ant27102...@ledelec.demon.co.uk>, Tony Williams
<to...@ledelec.demon.co.uk> wrote:
>
> Ross LaGue queried the film used....
>
> I tried everything to hand, Inkjet, OHP, and even Mylar dwg film.
> This is going to be a matter of experiment for least distortion
> but the exit temperature does not seem that great.
We have found that a thin mylar drawing film works better than
the OHP transparency film. Some touch-up is always necessary but
it is satisfactory for small PCB's. The dimensional stability is
not good but adequate for a cheap prototype...
Tony Williams
In article <54uqn8$7...@baygull.rtd.com>, Don Yuniskis
> This isn't as simple as it sounds! Depending on the technology
> employed in the print engine, simply inverting the "black" and "white"
> signals often results in output that is NOT exactly the same as
> a true "mirror image"! Some print engines "write white", others
> "write black". Some technologies dots "grow", etc. For traditional
> applications (text, graphics, etc.) the results are usually not
> visible (though there are some acid tests for these things). But,
> when you're printing tiny traces (even at 2:1 !) this can be a
> significant factor (underwidth or overwidth traces, etc.) -- consider
> that traces don't always "line up" with the elements in the print
> head so that can cost you as much as one "dot width". And, a 10 mil
> trace is already only 3 dots at 300dpi, etc.
>
> don
I think this refers to the "Reverse Video" comment.
The term "Mirror Imaging" referred to being able to print, on
film say, such that the text would read correctly when you
looked at it through the film.
A typical application of this would be one side of a pcb and we
used to produce "weatherproof" labels photographically by
printing on the emulsion side and laminating it down onto a
white backing. Very effective, and if a printer could do this
it would be a nice cheap way of doing such labels.
I thought that mirroring in software was simply a case of
Transforming all X or Y coordinates.
Or, in a raster device, you just swap lines end_for_end.
Is it any more complicated than that?
Tony Williams
Hello all,
The Laserjet 6P got connected to the computer today and some
real experimenting done, as follows:-
First, Units... 1mm = 0.040", = "40thou". ("mils" confuse me)
Two test dwgs were produced in a dwg-package, each dwg consisted
of an X-Y line-grid at 1" intervals and in the squares a symbol
consisting of concentric circles, 10thou linewidth and 10thou
between lines Each symbol was about 0.5" overall dia.
Dwg.1 had 88x such symbols, Dwg.2 had only 9x.
Test plots were done on paper, inkjet film, and mylar plotting
film. Just what was to hand really.
A pcb was developed and etched from the best plot.
Results:-
Dwg.1 clearly exceeded the 2meg RAM in the 6P and it defaulted
downwards in resolution. There were clear jaggies in the
quadrants of all circles. The resolution of the output was
good enough for only crude pcbs, probably 15/20thou min tracks.
Therefore all tests continued with Dwg.2 only.
Results on Inkjet film were extremely good, the tracks and spaces
were about equal size and I could detect no instance of a broken
track or shorted space. Distortion was... in the long direction
11" drawn was 10.94" measured and in the short direction 5" drawn
was 5" measured, more or less.
Results on the mylar were slightly worse, there were a few points
where broken tracks could be seen and the mylar crinkled badly.
Distortion was... in the long direction 11" drawn was 10.9"
measured and 5" dropped to about 4.98.
The inkjet film was used for a direct contact print onto a scrap
piece of photosensitive pcb, at the normal exposure time of
2.5 minutes, then developed and etched.
Results were very good. First there was no problem with the
normal exposure time, showing that the black has good uv-stopping.
Second, *every* 10thou circle was complete and with no breaks.
Note here that positive-working pcb and acid-undercut reduced
every 10thou track to about 7.xthou, with a corresponding
increase in the width of the original 10thou gaps.
Being a contact print there were no change in the dimensions.
Conclusions:-
For the first time in about 10 years I possess a method of
reliably and easily getting hard-copy directly out of a
CAD-package that could be classed as camera-ready artwork.
For prototype pcbs the print is directly useable and can be
relied on for tracks down in the 10/15thou region.
Dimensional distortions exist, but they are minor and simply
a matter of more experiments and fine-tuning.
Have to save my pennies and get more RAM in the 6P to avoid
that drop in resolution. Low priority though.
I started this thread with the question as to advice on the best
printer/plotter for 1:1 artwork. The private and public replies
got me to this result and for that, *grateful thanks to all*.
PS. I know the side topic of Mirror Image seems boring so I
have decided to not mention Mirror Image again, except of course
if someone else decides that Mirror Image is a useful facility
to have and takes up the Mirror Image cudgels.
Regards,