Homebrew PCBs

Hi Folks

I was guided here from the 1-Wire weather list[1], so I can stop posting 
off-topic stuff there about board fabrication ;-)

Just by way of introduction, I'm a freelance web applications developer 
(Perl, MySQL, a little PHP, somewhat less C) and general technology 
consultant, having spent many years in IT/network/comms management in an 
engineering multinational.

I have done hobby electronics on and off, in a rather half-hearted way, 
since my early teens.  My earliest days were spent working with 
Veroboard which, in hindsight, seems to require more effort to plan, 
assemble and debug than does making up a PCB.  First attempts at PCB 
fabrication were using a rub-down 'Letraset' system back in the early 
1990s.  I lost interest in this (too much like Hard Work) and only again 
started on PCB fabrication about 7 years ago, when I moved to Australia.

This time around, I had the freeware version of Cadsoft's Eagle, running 
on my Linux laptop.  I experimented with printing using both laser and 
inkjet printers, onto OHP transparency film.  (I actually found that the 
laser printer with the inkjet film seemed to produce the best results.)

I built myself a UV exposure box using an old file card drawer lined 
with aluminium foil, a pair of UVB tubes and ballast, covered by a sheet 
of thin glass, which sits on top of the rails that used to hold the file 
dividers.  This assembly is controlled by the digital timer from a 
deceased microwave oven.  (It even says "enjoy your meal" at the end of 
the exposure, which I still find amusing, in a sad way.)

Transparencies are contact printed with the above onto Kinsten 
pre-sensitised board and developed using sodium metasilicate.  (I gave 
up with caustic soda - often as not it would remove ALL the etch-resist, 
exposed or not.)

I do my etching using a rocker dish (actually an old Ferrero Rocher box) 
and ammonium persulphate.  Having to use thrice-loathsome ferric 
chloride near put me off for life - ammonium persulphate has been my 
saviour.

After three or four years of being too busy/sick/otherwise engaged, I 
find myself with a few projects requiring PCBs[2], so I am getting ready 
to go again.  I was going to treat myself to a bubble etch tank but it 
seems that both my sources (at least sources charging sensible prices) 
are out of stock for quite some time.  I was rather looking forward to 
the heated tank so that I wouldn't have to fish the PCB out half-way and 
put the solution in the microwave - but looks like I'll just have to 
carry on like that for a while longer.

The one thing that I have acquired this time round is the not-for-profit 
'Standard' version of Eagle.  Getting that extra board space is well 
worth it, in my mind.  Still waiting for the manuals, which were out of 
stock when I ordered.

Holes are something I hate; I have been trying to use carbide drills in 
a cheap-and-nasty, full-sized drill press.  All I will say is that it is 
very expensive in drills.  Last week, I took delivery of a Dremel and 
press - hopefully this process should get a bit easier for me.  (I still 
prefer larger SMDs though - anything that saves drilling is worth it!)

Issue that I will be needing to address:

* I'm still a little shaky when it comes to double-side boards; I have 
yet to perfect a method of aligning top and bottom films in a way that 
suits a) my UV exposure box and b) my shaky hands.  If I can get away 
with single-sided and a few jumpers, I generally will.

* Tinning boards.  Tinning solution is just too expensive for what I am 
doing.  Hand-tinning with solder doesn't seem to suit smaller components.

* Making vias.

* Mounting SMDs with central pads (in addition to the side pads). 
Haven't actually tried this yet, will probably just run a big pad end-to 
end of the device, poking out either end, and put the iron on one of the 
ends.

* Mounting SMDs with lots of small pins.  At the moment, I'm just trying 
to avoid them altogether.  I've read about the toaster oven technique - 
still sounds pretty tricky to me.

The above are challenges that I will be facing when I get back into it, 
but am not worrying about unduly at the moment.  When it comes to it, I 
will ferret through stuff on the Group and tap into your combined wisdom...

Cheers

M


References
----------

1 - <http://www.buoy.com/mailman/listinfo/weather>

2 - 1-Wire barometer, Nixie clock, isolation/level conversion boards for 
GPS modules, (wired) remote control board for outside lighting, modified 
quartz clocks that will read Tropical and Lunar years - and various 
other stuff.

-- 
Matthew Smith
Smiffytech - Technology Consulting & Web Application Development
Business: http://www.smiffytech.com/
Personal: http://www.smiffysplace.com/
LinkedIn: http://www.linkedin.com/in/smiffy

On Tue, 25 Mar 2008 08:41:56 +1030, you wrote:

>Hi Folks

Double sided boards.  If you're doing photo, then the positives should
line up properly.  Putting them together, on top of a board, then
drilling the board and positives together gives you matched holes
which can then be aligned.

If you decide to do toner transfer, there's other methods.

I'm somewhat avoiding SMD's and definitely avoiding BGA's.  Will
probably have to go to another SMD method, and center pads are not in
the works.

Vias:  I've tried both the pins and eyelets, and somewhat prefer
eyelets.  Those need to be soldered on both sides, though.  I've
gotten designs where the via does not have a parts lead going through
it, which facilitates using parts that do not enable me to solder the
top layer.

Once you do get the design done, though, any one thing on a top layer
becomes much nicer than jumpers.  Even eyelets are better, IMHO.

The Dremel is better (often) than the full sized press.  Even better
would be a high speed spindle.

Harvey

>
>I was guided here from the 1-Wire weather list[1], so I can stop posting 
>off-topic stuff there about board fabrication ;-)
>
>Just by way of introduction, I'm a freelance web applications developer 
>(Perl, MySQL, a little PHP, somewhat less C) and general technology 
>consultant, having spent many years in IT/network/comms management in an 
>engineering multinational.
>
>I have done hobby electronics on and off, in a rather half-hearted way, 
>since my early teens.  My earliest days were spent working with 
>Veroboard which, in hindsight, seems to require more effort to plan, 
>assemble and debug than does making up a PCB.  First attempts at PCB 
>fabrication were using a rub-down 'Letraset' system back in the early 
>1990s.  I lost interest in this (too much like Hard Work) and only again 
>started on PCB fabrication about 7 years ago, when I moved to Australia.
>
>This time around, I had the freeware version of Cadsoft's Eagle, running 
>on my Linux laptop.  I experimented with printing using both laser and 
>inkjet printers, onto OHP transparency film.  (I actually found that the 
>laser printer with the inkjet film seemed to produce the best results.)
>
>I built myself a UV exposure box using an old file card drawer lined 
>with aluminium foil, a pair of UVB tubes and ballast, covered by a sheet 
>of thin glass, which sits on top of the rails that used to hold the file 
>dividers.  This assembly is controlled by the digital timer from a 
>deceased microwave oven.  (It even says "enjoy your meal" at the end of 
>the exposure, which I still find amusing, in a sad way.)
>
>Transparencies are contact printed with the above onto Kinsten 
>pre-sensitised board and developed using sodium metasilicate.  (I gave 
>up with caustic soda - often as not it would remove ALL the etch-resist, 
>exposed or not.)
>
>I do my etching using a rocker dish (actually an old Ferrero Rocher box) 
>and ammonium persulphate.  Having to use thrice-loathsome ferric 
>chloride near put me off for life - ammonium persulphate has been my 
>saviour.
>
>After three or four years of being too busy/sick/otherwise engaged, I 
>find myself with a few projects requiring PCBs[2], so I am getting ready 
>to go again.  I was going to treat myself to a bubble etch tank but it 
>seems that both my sources (at least sources charging sensible prices) 
>are out of stock for quite some time.  I was rather looking forward to 
>the heated tank so that I wouldn't have to fish the PCB out half-way and 
>put the solution in the microwave - but looks like I'll just have to 
>carry on like that for a while longer.
>
>The one thing that I have acquired this time round is the not-for-profit 
>'Standard' version of Eagle.  Getting that extra board space is well 
>worth it, in my mind.  Still waiting for the manuals, which were out of 
>stock when I ordered.
>
>Holes are something I hate; I have been trying to use carbide drills in 
>a cheap-and-nasty, full-sized drill press.  All I will say is that it is 
>very expensive in drills.  Last week, I took delivery of a Dremel and 
>press - hopefully this process should get a bit easier for me.  (I still 
>prefer larger SMDs though - anything that saves drilling is worth it!)
>
>Issue that I will be needing to address:
>
>* I'm still a little shaky when it comes to double-side boards; I have 
>yet to perfect a method of aligning top and bottom films in a way that 
>suits a) my UV exposure box and b) my shaky hands.  If I can get away 
>with single-sided and a few jumpers, I generally will.
>
>* Tinning boards.  Tinning solution is just too expensive for what I am 
>doing.  Hand-tinning with solder doesn't seem to suit smaller components.
>
>* Making vias.
>
>* Mounting SMDs with central pads (in addition to the side pads). 
>Haven't actually tried this yet, will probably just run a big pad end-to 
>end of the device, poking out either end, and put the iron on one of the 
>ends.
>
>* Mounting SMDs with lots of small pins.  At the moment, I'm just trying 
>to avoid them altogether.  I've read about the toaster oven technique - 
>still sounds pretty tricky to me.
>
>The above are challenges that I will be facing when I get back into it, 
>but am not worrying about unduly at the moment.  When it comes to it, I 
>will ferret through stuff on the Group and tap into your combined wisdom...
>
>Cheers
>
>M
>
>
>References
>----------
>
>1 - <http://www.buoy.com/mailman/listinfo/weather>
>
>2 - 1-Wire barometer, Nixie clock, isolation/level conversion boards for 
>GPS modules, (wired) remote control board for outside lighting, modified 
>quartz clocks that will read Tropical and Lunar years - and various 
>other stuff.

----- Original Message ----- 

From: "Matthew Smith" <matt@...>
To: "Homebrew PCBs" <Homebrew_PCBs@yahoogroups.com>
Sent: Monday, March 24, 2008 10:11 PM
Subject: [Homebrew_PCBs] New Member


>
> Transparencies are contact printed with the above onto Kinsten
> pre-sensitised board and developed using sodium metasilicate.  (I gave
> up with caustic soda - often as not it would remove ALL the etch-resist,
> exposed or not.)

I don't have any problems with NaOH, you just need the right concentration 
(about 12 g/litre). You do have a lot more latitude with sodium 
metasilicate, though.

>
> Holes are something I hate; I have been trying to use carbide drills in
> a cheap-and-nasty, full-sized drill press.  All I will say is that it is
> very expensive in drills.  Last week, I took delivery of a Dremel and
> press - hopefully this process should get a bit easier for me.  (I still
> prefer larger SMDs though - anything that saves drilling is worth it!)
>
> Issue that I will be needing to address:
>
> * I'm still a little shaky when it comes to double-side boards; I have
> yet to perfect a method of aligning top and bottom films in a way that
> suits a) my UV exposure box and b) my shaky hands.  If I can get away
> with single-sided and a few jumpers, I generally will.

I just use jumpers.

>
> * Tinning boards.  Tinning solution is just too expensive for what I am
> doing.  Hand-tinning with solder doesn't seem to suit smaller components.

I don't bother, I remove the resist by rubbing with kitchen paper moistened 
with IPA, and it leaves a small amount of resist on the board which stops 
them oxidising. I don't have any soldering problems, although it would be a 
bit easier with tinned boards.

>
> * Making vias.
>
> * Mounting SMDs with central pads (in addition to the side pads).
> Haven't actually tried this yet, will probably just run a big pad end-to
> end of the device, poking out either end, and put the iron on one of the
> ends.
>
> * Mounting SMDs with lots of small pins.  At the moment, I'm just trying
> to avoid them altogether.  I've read about the toaster oven technique -
> still sounds pretty tricky to me.

I don't have any problems with that, using drag-soldering with plenty of gel 
flux.

Leon
--
Leon Heller
Amateur radio call-sign  G1HSM
Yaesu FT-817ND transceiver
Suzuki SV1000S motorcycle
leon355@...
http://www.geocities.com/leon_heller

On Tue, 25 Mar 2008, Matthew Smith wrote:
Hi,

> The one thing that I have acquired this time round is the not-for-profit
> 'Standard' version of Eagle.

Unfortunately, possibly too late for you now since you've paid, and
probably therefore will want to stick with Eagle... but if you're running
Linux, then PCB is a great PCB layout package:

http://pcb.sourceforge.net

with no arbitrary restriction on size or numbers of layers, and it's Free
as in freedom. I've done some fairly complex layouts using this software,
including a Z80 single board computer (Eurocard sized, double sided), a
remote SPI driven nixie display (again, Eurocard, but single sided), and
I'm just finishing off a rather densely populated 4 layer board (which I'm
going to send off to get fabricated...although some on the list have done
4 layer homebrew PCBs).

> * I'm still a little shaky when it comes to double-side boards; I have
> yet to perfect a method of aligning top and bottom films in a way that
> suits a) my UV exposure box and b) my shaky hands.  If I can get away
> with single-sided and a few jumpers, I generally will.

You might find toner transfer is easier for double sided. My second ever
board was double sided, and I've never had a 'failed' board yet using
toner transfer.

Basically, for registration, I use a bright light source (sunlight works
well) and put the transfer sheets face to face, then use the light to get
them lined up. PCB prints bold alignment lines to get it close - to get it
perfect, I just line up the pins and vias. Then I tape the sheets together
at one end to form a hinge, re-check alignment, and slip the board between
the sheets and get ironing. Usually, registration is absolutely spot on
using this method.

Dylan Smith wrote:
> Basically, for registration, I use a bright light source (sunlight works
> well) and put the transfer sheets face to face, then use the light to get
> them lined up. PCB prints bold alignment lines to get it close - to get it
> perfect, I just line up the pins and vias. Then I tape the sheets together
> at one end to form a hinge, re-check alignment, and slip the board between
> the sheets and get ironing. Usually, registration is absolutely spot on
> using this method.

I take the extra step of sandwiching a small strip of pcb material between the
films using double sided tape. It gives a good edge for alignment and reduces
the errors from the film bending around the edge of the pcb. Then I tack the
film to the substrate using adhesive tape to stop anything moving while it's
exposed, inverted and exposed again.

Matthew Smith <matt@...> writes:
> Hi Folks

Hi!

> The one thing that I have acquired this time round is the not-for-profit 
> 'Standard' version of Eagle.  Getting that extra board space is well 
> worth it, in my mind.  Still waiting for the manuals, which were out of 
> stock when I ordered.

Beware that Eagle has some nasty DRM built-in to it, some folks have
found themselves locked out of their own designs due to it.  Be *very*
careful about any library you use other than that which you create
yourself.  I use gEDA and PCB (and help maintain PCB), which have no
limits, no cost, and use text file formats.

> Last week, I took delivery of a Dremel and press - hopefully this
> process should get a bit easier for me.

I use a dremel in a home-made press.  Since switching to my own press,
I haven't broken a drill bit (except when I drop them on the floor)
and I use mostly 13 mil (vias).
http://www.delorie.com/pcb/dremel-stand/

> (I still prefer larger SMDs though - anything that saves drilling is
> worth it!)

I use mostly 0.5mm pitch and 0603 discretes.  But then again, others
have commented on my small-part-insanity.
http://www.delorie.com/pcb/smd-challenge/

> * I'm still a little shaky when it comes to double-side boards; I have 
> yet to perfect a method of aligning top and bottom films in a way that 
> suits a) my UV exposure box and b) my shaky hands.  If I can get away 
> with single-sided and a few jumpers, I generally will.

My technique (although for toner) is to tape on a paper pattern and
drill through some alignment holes just outside the board outline
(i.e. in the "waste" board).  Then I poke pinholes in the toner sheets
so I can align each side to the holes.  I can get within a few mil
this way.

> * Tinning boards.  Tinning solution is just too expensive for what I am 
> doing.  Hand-tinning with solder doesn't seem to suit smaller components.

I use MG Chemical's "Liquid Tin".  Just soak for a few minutes and rinse.
http://www.delorie.com/pcb/liquidtin/

> * Making vias.

I use 13 mil holes with a 28 mil pad, which is just enough to ensure
copper all the way around the hole.  I wrap some 20 mil solder around
the board to hold it up off my work surface, and cut a bunch of
lengths of wire wrap wire.  Each length is about 2-3 inches.  I flux
the wire, and, for each via, poke it through the hole, solder the top,
and cut it off.  This leaves all the vias with the wire sticking 20
mil out the other side, so just flip and solder them all.

> * Mounting SMDs with central pads (in addition to the side pads). 
> Haven't actually tried this yet, will probably just run a big pad end-to 
> end of the device, poking out either end, and put the iron on one of the 
> ends.

Hot plate and solder paste.  http://www.delorie.com/pcb/hotplate/
and http://geda.seul.org/projects/djs_pcbs/

Others have put large vias under the pad, so they can solder through
from the other side.

> * Mounting SMDs with lots of small pins.  At the moment, I'm just trying 
> to avoid them altogether.  I've read about the toaster oven technique - 
> still sounds pretty tricky to me.

Hot plate again.  It's *really* easy to fix shorts with copper braid,
so aim for slightly more than enough paste.

Dylan Smith <dyls@...> writes:
> Linux, then PCB is a great PCB layout package:
> 
> http://pcb.sourceforge.net

I agree, but then again, I'm biased :-)

PCB and gEDA run just fine on MacOS/X also, and mostly work on
Windows.

In theory, PCB can do boards up to a quarter mile per side, down to
1E-5 inch precision, with as many layers as you like (I've personally
tested 53 layers).  It's very difficult to use at that point, of
course.  IIRC it *has* been used for panel-sized 12 layer boards.  I
get hints and stories that I can't always share, since I'm one of the
maintainers.

> going to send off to get fabricated...although some on the list have
> done 4 layer homebrew PCBs).

I'm one, but again, there's that insanity peeking through.

--- In Homebrew_PCBs@yahoogroups.com, DJ Delorie <dj@...> wrote:
>
> 
> Matthew Smith <matt@...> writes:
> > Hi Folks
> 
> Hi!
> 
> > The one thing that I have acquired this time round is the
not-for-profit 
> > 'Standard' version of Eagle.  Getting that extra board space is well 
> > worth it, in my mind.  Still waiting for the manuals, which were
out of 
> > stock when I ordered.
> 
> Beware that Eagle has some nasty DRM built-in to it, some folks have
> found themselves locked out of their own designs due to it.  Be *very*
> careful about any library you use other than that which you create
> yourself.  I use gEDA and PCB (and help maintain PCB), which have no
> limits, no cost, and use text file formats.
> 

Interesting, I have used Eagle for many years and have not heard about
that. 
I know there are limits on board size and layer count for the free and
nonprofit versions. Do you have any more specifics or links concerning
DRM and eagle?

"javaguy11111" <javaguy11111@...> writes:
> Interesting, I have used Eagle for many years and have not heard about
> that. 

IIRC it's something new in the latest release; a lot of folks who've
been using Eagle for years upgrade and suddenly find their boards
locked because somewhere in the past they got tainted by a library or
part from a pirated copy of Eagle, passed from friend to friend.
Worse, the taint is contagious - it can get passed to a part that *is*
legal, just because something *else* therein came from a bad source.

So you could have a perfectly legal copy of some library part from a
legitimate source, and still get locked out of your design.

> Do you have any more specifics or links concerning DRM and eagle?

http://groups.google.com/group/sci.electronics.design/browse_frm/thread/3fd8ddbb525a110a/af8c28ff5ccb0e8f?lnk=st&q=eagle+drm#af8c28ff5ccb0e8f

--- In Homebrew_PCBs@yahoogroups.com, DJ Delorie <dj@...> wrote:
>
> 
> "javaguy11111" <javaguy11111@...> writes:
> > Interesting, I have used Eagle for many years and have not heard about
> > that. 
> 
> IIRC it's something new in the latest release; a lot of folks who've
> been using Eagle for years upgrade and suddenly find their boards
> locked because somewhere in the past they got tainted by a library or
> part from a pirated copy of Eagle, passed from friend to friend.
> Worse, the taint is contagious - it can get passed to a part that *is*
> legal, just because something *else* therein came from a bad source.
> 
> So you could have a perfectly legal copy of some library part from a
> legitimate source, and still get locked out of your design.
> 
> > Do you have any more specifics or links concerning DRM and eagle?
> 
>
http://groups.google.com/group/sci.electronics.design/browse_frm/thread/3fd8ddbb525a110a/af8c28ff5ccb0e8f?lnk=st&q=eagle+drm#af8c28ff5ccb0e8f
>

Thanks for the info. I was not aware of Eagle locking people out of
their designs like that. Especially the issue with parts libraries.

Mark Robinson wrote:
> I take the extra step of sandwiching a small strip of pcb material between the
> films using double sided tape. It gives a good edge for alignment and reduces
> the errors from the film bending around the edge of the pcb. Then I tack the
> film to the substrate using adhesive tape to stop anything moving while it's
> exposed, inverted and exposed again.

That sounds like it might be a good way to align toner-transfer transfers...

Take an L-shaped offcut of PCB material (of whatever thickness you're using) 
and transfer an alignment marker on one side.

Drill a hole in the centre of the alignment target, align it against the 
bottom transfer so the marks match up, then tape it down with double-sided tape.

Cut the area around the alignment mark out of the bottom transfer, then put 
the L-frame onto a lightbox.

Now take the top transfer, align it against the alignment marker, and stick it 
down.

Iron on as normal.


Does this sound plausible? Seems it might work a bit better than my 
tape-the-two-transfers-together method, which seems to fail dismally on 
anything PCB blank thicker than ~0.8mm.

-- 
Phil.                         |  (\_/)  This is Bunny. Copy and paste Bunny
ygroups@...         | (='.'=) into your signature to help him gain
http://www.philpem.me.uk/     | (")_(") world domination.

Philip Pemberton <ygroups@...> writes:
> Cut the area around the alignment mark out of the bottom transfer,
> then put the L-frame onto a lightbox.

The problem here is that, at least with the Pulsar paper, it's too
opaque to use on a light table.  I've tried, you just can't see
through it well enough to line up the toners.

--- Philip Pemberton <ygroups@...> wrote:


> 
> 
> Does this sound plausible? Seems it might work a bit
> better than my 
> tape-the-two-transfers-together method, which seems
> to fail dismally on 
> anything PCB blank thicker than ~0.8mm.
> 
I did some exposures using polyester paper, which
works really well.  I used push pins to get the
alignment (holes in the copper) and I put a few drops
of water on the board.  The surface tension holds the
polyester to the board during exposure.

Again, at that time I had the luxury of a calibrated
exposure system, but the exposure area was no more
than 4" x 4" at a time.

I had the luxury at one time to be able to "spin-coat"
small boards with resist, but drilling of the
alignment holes should be done after the resist is
applied otherwise the holes create undesired abnormalities.


      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ

DJ Delorie wrote:
> Philip Pemberton <ygroups@...> writes:
>> Cut the area around the alignment mark out of the bottom transfer,
>> then put the L-frame onto a lightbox.
> 
> The problem here is that, at least with the Pulsar paper, it's too
> opaque to use on a light table.  I've tried, you just can't see
> through it well enough to line up the toners.

Hm. The HP Everyday paper isn't too bad in this respect - nice and easy to see 
through (it's quite thin). The problem is the toner usually doesn't stick to 
it evenly -- there are grey patches all over the pattern. Needless to say, 
although it just floats off the board, it needs a lot of touch-up.

I can't say I've ever used the Pulsar paper, though I have used Press-n-Peel, 
which is pretty good (if a bit on the expensive side). Shame it's become 
somewhat hard to find in the UK, IIRC the UK distributor went bust. Techniks 
still list them on their website as a distributor though...

I suppose the challenge is to find a paper that's thin enough to let light 
pass through it relatively unimpeded, and also works well for TT.

-- 
Phil.                         |  (\_/)  This is Bunny. Copy and paste Bunny
ygroups@...         | (='.'=) into your signature to help him gain
http://www.philpem.me.uk/     | (")_(") world domination.

Actually this happened to me, and I was very upset back then especially 
because CadSoft support refused to help even though the case was pretty 
clear considering the fact that I have a legal licensed copy of eagle 
for years now and they easily could check this.

Anyways, the issue is no longer really such a problem cause thre are 
ULPs which export ANY design and library part into ascii format which 
then later on can be reimported thereby making a previousely non working 
design or part useable again withouth restriction. The good thing about 
this is that no single detail of information is lost during this 
operation except for the DRM stuff.

Markus

javaguy11111 schrieb:

>
> --- In Homebrew_PCBs@yahoogroups.com 
> <mailto:Homebrew_PCBs%40yahoogroups.com>, DJ Delorie <dj@...> wrote:
> >
> >
> > "javaguy11111" <javaguy11111@...> writes:
> > > Interesting, I have used Eagle for many years and have not heard about
> > > that.
> >
> > IIRC it's something new in the latest release; a lot of folks who've
> > been using Eagle for years upgrade and suddenly find their boards
> > locked because somewhere in the past they got tainted by a library or
> > part from a pirated copy of Eagle, passed from friend to friend.
> > Worse, the taint is contagious - it can get passed to a part that *is*
> > legal, just because something *else* therein came from a bad source.
> >
> > So you could have a perfectly legal copy of some library part from a
> > legitimate source, and still get locked out of your design.
> >
> > > Do you have any more specifics or links concerning DRM and eagle?
> >
> >
> http://groups.google.com/group/sci.electronics.design/browse_frm/thread/3fd8ddbb525a110a/af8c28ff5ccb0e8f?lnk=st&q=eagle+drm#af8c28ff5ccb0e8f 
> <http://groups.google.com/group/sci.electronics.design/browse_frm/thread/3fd8ddbb525a110a/af8c28ff5ccb0e8f?lnk=st&q=eagle+drm#af8c28ff5ccb0e8f>
> >
>
> Thanks for the info. I was not aware of Eagle locking people out of
> their designs like that. Especially the issue with parts libraries.
>
>

Markus Zingg <homebrew-pcb@...> writes:
> ULPs which export ANY design and library part into ascii format which 

Are these ascii formats documented?  I suppose they could be used to
convert to/from gEDA format, too.

I asked cadsoft support a question once (how you can add dimensions to
the outline of a PCB). The response i got was that it can't do that
and i should use a ULP. I tried the ULP, the result was useless to the
extent of being ridiculous. I was very disappointed with the support
and never tried again.

Compare to that the target support, they even called me because they
wanted to ask further questions, looked up the number in a phone book
from another country. Not only did they answer my requests as well as
they could every time, but they also implemented new features
following that request, if it was useful. They now keep a wish list
right in the software, where users can vote which features would be
most valued to them, so they can work on the most urgent ones first.

Now if you consider that i have only a single small license (worth
~100eur) with target, and my company has a multi user professional
license with eagle (worth i don't know how much) they really ought to
be ashamed about their support.

Your mileage may vary, but my experience is the company is not
helpful. That they would not help with the DRM issue seems very
fitting to me. Frankly i don't know what they are doing all day, the
software looks and works as if it hadn't changed a bit since MS-DOS
times. They seem to feel no need for improvement from their side,
instead expecting you to use or possibly even write ULPs. If i wanted
to write my own software i would not pay good money for theirs.

I really don't like to badmouth anyone, but those are honest
experiences, and i'm very disappointed.

ST

On 25 Mar 2008 15:18:14 -0400, DJ Delorie <dj@...> wrote:
>
> Markus Zingg <homebrew-pcb@...> writes:
> > ULPs which export ANY design and library part into ascii format which
>
> Are these ascii formats documented?  I suppose they could be used to
> convert to/from gEDA format, too.
>

DJ,

Basically eagles ULP language contains all information about the 
internal data structures of every existing object. All you need to write 
export scripts along with detailed structure documentation can be found 
in the online help. This is also true for the evaluation versions and 
even the most restrictive version can load the most complex design (just 
not change it) and run a ULP on it. So, it's relatively simple for 
anyone intersted and somewhat savey with writing software to write a ULP 
that even directly builds up an ascii (or even binary format for that 
matter) that directly matches what virtually any third party software 
would use. Btw, the ULP is VERY similar to standard 'C' programming 
language alas if you know 'C' your almost instantly up and running.

One can think of eagle what he wants, it's not THAT bad IMHO. It's of 
course not compareable against the big names in this field (PADS etc.) 
but one can do quite sophisticated stuff with it and add missing 
functionality with the help of writing ULPs. That's at the same time 
also the worest thing about it cause one can solve many many problems 
using ULPs which eventually kind of hide the issue/problem from CadSoft. 
I.e. I wrote ULPs to make eagle even out digital busses lenght wise, or 
to route impedance controlled signal pairs etc. The aproach is not as 
good as it would be if native support for stuff like that would be 
present, but at least if you are on a budget (like me) there is a way to 
help yourself out.

I later made quite good experience with CadSupport. It apears like they 
act just quite paranoid if they suspect that they might deal with 
someone useing illegal copies of their software and in my case they 
definately went beond any common sense back then...

With regard to STs comments on eagle - well, I once tried to use target 
and miserably failed cause I did not found it to be ergonomical for me. 
Aparently there are reasons why there are different packages on the 
market and to me it's fine if everyone finds a product that matches 
his/her needs. I respect other peoples opinions and preferences. I don't 
say I'm absolutely happy with eagle. I admit that I think I'm somewaht 
outgrown of it and if I would have the money I definately would consider 
something in the upper end (PADs, Allegro - you name it).

Markus

DJ Delorie schrieb:

>
>
> Markus Zingg <homebrew-pcb@... 
> <mailto:homebrew-pcb%40shdesign.info>> writes:
> > ULPs which export ANY design and library part into ascii format which
>
> Are these ascii formats documented? I suppose they could be used to
> convert to/from gEDA format, too.
>
>

Who or what is Target? I don't see a link in the Links section...
hint, hint.

Steve Greenfield

--- In Homebrew_PCBs@yahoogroups.com, "Stefan Trethan"
<stefan_trethan@...> wrote:

>
> I asked cadsoft support a question once (how you can add dimensions to
> the outline of a PCB). The response i got was that it can't do that
> and i should use a ULP. I tried the ULP, the result was useless to the
> extent of being ridiculous. I was very disappointed with the support
> and never tried again.
> 
> Compare to that the target support, they even called me because they
> wanted to ask further questions, looked up the number in a phone book
> from another country. Not only did they answer my requests as well as
> they could every time, but they also implemented new features
> following that request, if it was useful. They now keep a wish list
> right in the software, where users can vote which features would be
> most valued to them, so they can work on the most urgent ones first.
> 
> Now if you consider that i have only a single small license (worth
> ~100eur) with target, and my company has a multi user professional
> license with eagle (worth i don't know how much) they really ought to
> be ashamed about their support.
> 
> Your mileage may vary, but my experience is the company is not
> helpful. That they would not help with the DRM issue seems very
> fitting to me. Frankly i don't know what they are doing all day, the
> software looks and works as if it hadn't changed a bit since MS-DOS
> times. They seem to feel no need for improvement from their side,
> instead expecting you to use or possibly even write ULPs. If i wanted
> to write my own software i would not pay good money for theirs.
> 
> I really don't like to badmouth anyone, but those are honest
> experiences, and i'm very disappointed.
> 
> ST

----- Original Message ----- 

From: "Steve" <alienrelics@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Wednesday, March 26, 2008 12:07 AM
Subject: [Homebrew_PCBs] PCB Software, was Re: New Member


> Who or what is Target? I don't see a link in the Links section...
> hint, hint.

http://www.ibfriedrich.com/

PCB Pool supplies a free version of it.

Leon
--
Leon Heller
Amateur radio call-sign  G1HSM
Yaesu FT-817ND transceiver
Suzuki SV1000S motorcycle
leon355@...
http://www.geocities.com/leon_heller

On Tue, 25 Mar 2008, Stefan Trethan wrote:

[Eagle]
> fitting to me. Frankly i don't know what they are doing all day, the
> software looks and works as if it hadn't changed a bit since MS-DOS
> times. They seem to feel no need for improvement from their side,
> instead expecting you to use or possibly even write ULPs.

Someone in alt.sysadmin.recovery remarked on looking at a vendor - check
their laurels, and if they are flat, avoid that company because they've
been resting on their laurels far too long.

Sounds like Cadsoft have been doing just that.

----- Original Message ----- 

From: "Dylan Smith" <dyls@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Wednesday, March 26, 2008 10:00 AM
Subject: Re: [Homebrew_PCBs] Re: New Member


> On Tue, 25 Mar 2008, Stefan Trethan wrote:
>
> [Eagle]
>> fitting to me. Frankly i don't know what they are doing all day, the
>> software looks and works as if it hadn't changed a bit since MS-DOS
>> times. They seem to feel no need for improvement from their side,
>> instead expecting you to use or possibly even write ULPs.
>
> Someone in alt.sysadmin.recovery remarked on looking at a vendor - check
> their laurels, and if they are flat, avoid that company because they've
> been resting on their laurels far too long.
>
> Sounds like Cadsoft have been doing just that.

The Pulsonix software I use and the cheaper Easy-PC, from the same company, 
were both developed  using C++ and proper OOP techniques. Both are regularly 
upgraded with new versions every couple of years. The few bugs that users 
come across are fixed immediately if they are serious, or a workaround is 
provided, often in 24 hours.

Leon
--
Leon Heller
Amateur radio call-sign  G1HSM
Yaesu FT-817ND transceiver
Suzuki SV1000S motorcycle
leon355@...
http://www.geocities.com/leon_heller

When you use a laminator to fuse toner, do you use the carrier or 
feed the board in directly. Is there any trick to keep the toner from 
smudging as it melts?

Mark

I just feed the board and the paper, no carrier.  IMHO if the toner is
smudging, it's too hot.  There's a sweet spot between "solid" and
"liquid" where it's plasticy and sticky - that's what you want.

I would say if the toner is smudging the paper moves relative to the board?
If it spreads out it might be too hot, or too much pressure, or too much toner.
But a smudge requires some relative movement in my understanding.

ST

On 26 Mar 2008 12:26:20 -0400, DJ Delorie <dj@...> wrote:
>
> I just feed the board and the paper, no carrier.  IMHO if the toner is
> smudging, it's too hot.  There's a sweet spot between "solid" and
> "liquid" where it's plasticy and sticky - that's what you want.
>
>
> ------------------------------------
>
> Be sure to visit the group home and check for new Links, Files, and Photos:
> http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
>
>
>
>

Well, I'm not doing toner transfer, actually, I'm using a modified 
laser printer to print directly on the board - the toner is still in 
powder form, so it is relatively easy for it to smudge. I suspect 
I'll have to use a hot plate and directly compress the board between 
two plates using a teflon sheet between the toner and the top plate. 
The laminator caused only a little smudging, but it was my first 
attempt and I used the carrier. Perhaps just a teflon sheet will 
allow use of the laminator. We'll see.

Mark


At 12:34 PM 3/26/2008, you wrote:

>I would say if the toner is smudging the paper moves relative to the board?
>If it spreads out it might be too hot, or too much pressure, or too 
>much toner.
>But a smudge requires some relative movement in my understanding.
>
>ST
>
>On 26 Mar 2008 12:26:20 -0400, DJ Delorie <dj@...> wrote:
> >
> > I just feed the board and the paper, no carrier.  IMHO if the toner is
> > smudging, it's too hot.  There's a sweet spot between "solid" and
> > "liquid" where it's plasticy and sticky - that's what you want.
> >
> >
> > ------------------------------------
> >
> > Be sure to visit the group home and check for new Links, Files, and Photos:
> > http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
> >
> >
> >
> >
>
>------------------------------------
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
>
>
>

At 10:22 AM 3/26/2008, you wrote:
>When you use a laminator to fuse toner, do you use the carrier or
>feed the board in directly. Is there any trick to keep the toner from
>smudging as it melts?
>
>Mark

I run it through once with a carrier to tack the paper down. Then run 
it through a couple of more times without the carrier to completely 
melt the toner.
  Smudging has not been a problem. The paper itself seems to act as a 
buffer to keep the toner from smearing.

  Note that you need to use a carrier for very small boards, else 
they'll get stuck inside of laminators where the rollers are too far 
apart to pull the board completely through..




carl

--------------------------------------------------------
Henry Carl Ott   N2RVQ
http://users.rcn.com/carlott/
--------------------------------------------------------

Hello Mark,
                  I am an ex printer technician and I have been toying with 
the concept of direct printing with a laser for the last couple of days.

What concerned me is that the conductivity of the copper on the PCB was 
likely to cause the image on the PCB to be blurred or poorly focused along 
the Y axis.

In normal operation with paper the image is transferred from the drum to the 
page by a single line (wire) across the X axis. Manufacturers go to some 
trouble to make this wire as thin as possible and ground two parallel lines. 
This improves Y focus and I expected this would become a problem due the 
conductive copper coating.

I am very interested to hear what your experiments are revealing.

If you are using the image drum and you have the above problem then you need 
to provide shielding above and below the point on the image drum that is 
closest to the PCB.

Ideally the shielding voltage should be further away from the from PCB 
voltage than the image drum itself. The easiest way to achieve this is 
probably have a divider on the HT side and upping the actual HT.

I am a bit rusty on the theory so please let me know if this is the problem 
and I will do my revision and get back to you with a mode detailed 
information.

If this is simply a fusing problem, ie the image is fine on the PCB then 
there could be a number of causes.

Firstly there are two types of fuser today and the laminator may be the more 
modern type.

The older type used a heated roller and often this roller had a soft 
coating, this type is ideal for what you want as it will provide an even 
pressure to the toner. It will also work better at a lower temperature 
(desirable) as it will have a greater surface area of physical contact.

The older type that had a solid roller is less suitable as the two hard 
surfaces together will cause uneven pressure and the reduced surface contact 
will require more temperature which in turn increases the probability of 
blurring.

Fuser temperature needs to be accurate. All fusers have a temperature sensor 
and are closely regulated.

The older type where good but had two problems for normal printer use.
1) Heating a whole roller took a lot more energy.
2) The complete roller has far more thermal mass and it takes time to get 
the higher thermal mass to an accurate temperature. This caused long delays 
before the printer could print its first page after power on or standby and 
this delay is an important spec to manufacturers.

The newer type works completely differently. The roller is somewhat 
flattened on the side that contacts the paper and does not revolve. Around 
this roller is a film of thermally conductive material that can revolve 
around the roller to allow the paper free movement. There is thin strip 
heating element below the thermally conductive film that can change in 
temperature very rapidly doe to it's low thermal mass.

I doubt this newer type is any good for what you want for two reasons -
1) It is not very pliable and would probably cause uneven pressures.
2) It is not very robust and I would expect the edges of the PCB to wear 
into it very quickly.

The older type with the hard roller is probably not much good either due to 
uneven pressure and higher required temperatures.

The older type with the soft roller would be ideal.

I don't know what is in the laminator you have. Or even if that have any 
form of temperature regulation. Very crude regulation could be done with a 
light dimmer but two warnings on this. Don't even try it if you are not 
experienced with electrical things or you have no safety breakers on you 
power supply (RCD, CBR) etc. And The power requirements for the laminator is 
likely to be several hundred watts and most light dimmers do not go up that 
far.

Some other thoughts I have been thinking are to do away with the image 
transfer all together. Perhaps the image unit could still be used to provide 
an even coating of toner to the complete surface of the PCB. The new scanner 
would be installed with a high powered laser and an optical attenuator 
placed in the path for the syntonisation sensor to prevent damage to it by 
the higher power.

It idea is that the scanners laser provide the power to fuse the toner to 
the board and rest of the toner is just removed by blowing it a way.

I haven't the math for it yet so I don't know what power laser would be 
required.

The power to the toner at any point is the product of the laser power and 
the time. The faster the scanning the lower the effective power. However on 
the other hand is the scanners need for speed. The scanner accuracy is 
dependent on its inertia which is the product of rotational mass by the 
rotational speed squared. So obviously there are limits that will dictate 
the minium laser power required.

Thanks.

----- Original Message ----- 
From: "Mark Lerman"

Well, I'm not doing toner transfer, actually, I'm using a modified
laser printer to print directly on the board - the toner is still in
powder form, so it is relatively easy for it to smudge. I suspect
I'll have to use a hot plate and directly compress the board between
two plates using a teflon sheet between the toner and the top plate.
The laminator caused only a little smudging, but it was my first
attempt and I used the carrier. Perhaps just a teflon sheet will
allow use of the laminator. We'll see.

Mark

I've spent the last month or so looking at all the discussions re direct laser printing on boards. There are all sorts of warnings, nay sayers, etc who say it can't work. Then I saw the post and site of laserpcb, where Dave says (he has a patent)that he can print on copper foil using his special "gel" then glue it to FR4 to make a PCB. So, if it is possible to put toner on copper foil, it should be possible to put it on boards directly.

I took a Samsung ML-2510 and modified the paper path to be straight and flat, and moved the feed sensor (also inverted it) then I simply ran a board through. No problem. A few sparks, but the image seems very good, no distortion, very crisp. I tried Dave's Gel (sorry Dave), but the image seems identical with and without the gel. 

I haven't finished experimenting, which is why I wanted the test pattern. I seem to easily be able to print 1 mil traces, but I'm not sure yet of the spacing I can use, though I would guess 5-8 mil. 

The board comes out of the printer with the toner as powder electrostatically stuck to the board, so it has to be fused. I am reluctant to use the fuser on the printer for several reasons. One is that it is a pain to move it to a flat path - I would have to move a gear, among other things. Secondly, the speed might easily be too fast to fuse the toner. Also, if the toner is smudged or not perfect, it is very easy to wipe and reprint the board if it isn't fused. 

An advantage of fusing it is that it (hopefully) will be stuck in place, even if it isn't fused completely. Obviously there is a lot of work to do here.

I think I have demonstrated that direct laser printing of pcb's is possible. If anyone wants further information, I will be happy to share.

Mark


-----Original Message-----

>From: listgroups08@...
>Sent: Mar 26, 2008 7:30 PM
>To: Homebrew_PCBs@yahoogroups.com
>Subject: Re: [Homebrew_PCBs] Laminator Questions
>
>Hello Mark,
>                  I am an ex printer technician and I have been toying with 
>the concept of direct printing with a laser for the last couple of days.
>
>What concerned me is that the conductivity of the copper on the PCB was 
>likely to cause the image on the PCB to be blurred or poorly focused along 
>the Y axis.
>
>In normal operation with paper the image is transferred from the drum to the 
>page by a single line (wire) across the X axis. Manufacturers go to some 
>trouble to make this wire as thin as possible and ground two parallel lines. 
>This improves Y focus and I expected this would become a problem due the 
>conductive copper coating.
>
>I am very interested to hear what your experiments are revealing.
>
>If you are using the image drum and you have the above problem then you need 
>to provide shielding above and below the point on the image drum that is 
>closest to the PCB.
>
>Ideally the shielding voltage should be further away from the from PCB 
>voltage than the image drum itself. The easiest way to achieve this is 
>probably have a divider on the HT side and upping the actual HT.
>
>I am a bit rusty on the theory so please let me know if this is the problem 
>and I will do my revision and get back to you with a mode detailed 
>information.
>
>If this is simply a fusing problem, ie the image is fine on the PCB then 
>there could be a number of causes.
>
>Firstly there are two types of fuser today and the laminator may be the more 
>modern type.
>
>The older type used a heated roller and often this roller had a soft 
>coating, this type is ideal for what you want as it will provide an even 
>pressure to the toner. It will also work better at a lower temperature 
>(desirable) as it will have a greater surface area of physical contact.
>
>The older type that had a solid roller is less suitable as the two hard 
>surfaces together will cause uneven pressure and the reduced surface contact 
>will require more temperature which in turn increases the probability of 
>blurring.
>
>Fuser temperature needs to be accurate. All fusers have a temperature sensor 
>and are closely regulated.
>
>The older type where good but had two problems for normal printer use.
>1) Heating a whole roller took a lot more energy.
>2) The complete roller has far more thermal mass and it takes time to get 
>the higher thermal mass to an accurate temperature. This caused long delays 
>before the printer could print its first page after power on or standby and 
>this delay is an important spec to manufacturers.
>
>The newer type works completely differently. The roller is somewhat 
>flattened on the side that contacts the paper and does not revolve. Around 
>this roller is a film of thermally conductive material that can revolve 
>around the roller to allow the paper free movement. There is thin strip 
>heating element below the thermally conductive film that can change in 
>temperature very rapidly doe to it's low thermal mass.
>
>I doubt this newer type is any good for what you want for two reasons -
>1) It is not very pliable and would probably cause uneven pressures.
>2) It is not very robust and I would expect the edges of the PCB to wear 
>into it very quickly.
>
>The older type with the hard roller is probably not much good either due to 
>uneven pressure and higher required temperatures.
>
>The older type with the soft roller would be ideal.
>
>I don't know what is in the laminator you have. Or even if that have any 
>form of temperature regulation. Very crude regulation could be done with a 
>light dimmer but two warnings on this. Don't even try it if you are not 
>experienced with electrical things or you have no safety breakers on you 
>power supply (RCD, CBR) etc. And The power requirements for the laminator is 
>likely to be several hundred watts and most light dimmers do not go up that 
>far.
>
>Some other thoughts I have been thinking are to do away with the image 
>transfer all together. Perhaps the image unit could still be used to provide 
>an even coating of toner to the complete surface of the PCB. The new scanner 
>would be installed with a high powered laser and an optical attenuator 
>placed in the path for the syntonisation sensor to prevent damage to it by 
>the higher power.
>
>It idea is that the scanners laser provide the power to fuse the toner to 
>the board and rest of the toner is just removed by blowing it a way.
>
>I haven't the math for it yet so I don't know what power laser would be 
>required.
>
>The power to the toner at any point is the product of the laser power and 
>the time. The faster the scanning the lower the effective power. However on 
>the other hand is the scanners need for speed. The scanner accuracy is 
>dependent on its inertia which is the product of rotational mass by the 
>rotational speed squared. So obviously there are limits that will dictate 
>the minium laser power required.
>
>Thanks.
>
>----- Original Message ----- 
>From: "Mark Lerman"
>
>Well, I'm not doing toner transfer, actually, I'm using a modified
>laser printer to print directly on the board - the toner is still in
>powder form, so it is relatively easy for it to smudge. I suspect
>I'll have to use a hot plate and directly compress the board between
>two plates using a teflon sheet between the toner and the top plate.
>The laminator caused only a little smudging, but it was my first
>attempt and I used the carrier. Perhaps just a teflon sheet will
>allow use of the laminator. We'll see.
>
>Mark
>
>
>------------------------------------
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
>
>
>

As for laminators, I just bought a Staples machine - no adjustments. I also just ordered a pid temperature controller that I will use either with the laminator or with a "hot plate". Does anyone know the proper fuser temperature??

BTW, I'm not sure whether registration for double sided boards will work - it would require a precision optical sensor plus a rigid mechanical guide. The simplest way is probably to use two single sided boards of half the thiskness and use registration pins for alignment, then glue them together. Should work fine.

Mark

-----Original Message-----

>From: listgroups08@...
>Sent: Mar 26, 2008 7:30 PM
>To: Homebrew_PCBs@yahoogroups.com
>Subject: Re: [Homebrew_PCBs] Laminator Questions
>
>Hello Mark,
>                  I am an ex printer technician and I have been toying with 
>the concept of direct printing with a laser for the last couple of days.
>
>What concerned me is that the conductivity of the copper on the PCB was 
>likely to cause the image on the PCB to be blurred or poorly focused along 
>the Y axis.
>
>In normal operation with paper the image is transferred from the drum to the 
>page by a single line (wire) across the X axis. Manufacturers go to some 
>trouble to make this wire as thin as possible and ground two parallel lines. 
>This improves Y focus and I expected this would become a problem due the 
>conductive copper coating.
>
>I am very interested to hear what your experiments are revealing.
>
>If you are using the image drum and you have the above problem then you need 
>to provide shielding above and below the point on the image drum that is 
>closest to the PCB.
>
>Ideally the shielding voltage should be further away from the from PCB 
>voltage than the image drum itself. The easiest way to achieve this is 
>probably have a divider on the HT side and upping the actual HT.
>
>I am a bit rusty on the theory so please let me know if this is the problem 
>and I will do my revision and get back to you with a mode detailed 
>information.
>
>If this is simply a fusing problem, ie the image is fine on the PCB then 
>there could be a number of causes.
>
>Firstly there are two types of fuser today and the laminator may be the more 
>modern type.
>
>The older type used a heated roller and often this roller had a soft 
>coating, this type is ideal for what you want as it will provide an even 
>pressure to the toner. It will also work better at a lower temperature 
>(desirable) as it will have a greater surface area of physical contact.
>
>The older type that had a solid roller is less suitable as the two hard 
>surfaces together will cause uneven pressure and the reduced surface contact 
>will require more temperature which in turn increases the probability of 
>blurring.
>
>Fuser temperature needs to be accurate. All fusers have a temperature sensor 
>and are closely regulated.
>
>The older type where good but had two problems for normal printer use.
>1) Heating a whole roller took a lot more energy.
>2) The complete roller has far more thermal mass and it takes time to get 
>the higher thermal mass to an accurate temperature. This caused long delays 
>before the printer could print its first page after power on or standby and 
>this delay is an important spec to manufacturers.
>
>The newer type works completely differently. The roller is somewhat 
>flattened on the side that contacts the paper and does not revolve. Around 
>this roller is a film of thermally conductive material that can revolve 
>around the roller to allow the paper free movement. There is thin strip 
>heating element below the thermally conductive film that can change in 
>temperature very rapidly doe to it's low thermal mass.
>
>I doubt this newer type is any good for what you want for two reasons -
>1) It is not very pliable and would probably cause uneven pressures.
>2) It is not very robust and I would expect the edges of the PCB to wear 
>into it very quickly.
>
>The older type with the hard roller is probably not much good either due to 
>uneven pressure and higher required temperatures.
>
>The older type with the soft roller would be ideal.
>
>I don't know what is in the laminator you have. Or even if that have any 
>form of temperature regulation. Very crude regulation could be done with a 
>light dimmer but two warnings on this. Don't even try it if you are not 
>experienced with electrical things or you have no safety breakers on you 
>power supply (RCD, CBR) etc. And The power requirements for the laminator is 
>likely to be several hundred watts and most light dimmers do not go up that 
>far.
>
>Some other thoughts I have been thinking are to do away with the image 
>transfer all together. Perhaps the image unit could still be used to provide 
>an even coating of toner to the complete surface of the PCB. The new scanner 
>would be installed with a high powered laser and an optical attenuator 
>placed in the path for the syntonisation sensor to prevent damage to it by 
>the higher power.
>
>It idea is that the scanners laser provide the power to fuse the toner to 
>the board and rest of the toner is just removed by blowing it a way.
>
>I haven't the math for it yet so I don't know what power laser would be 
>required.
>
>The power to the toner at any point is the product of the laser power and 
>the time. The faster the scanning the lower the effective power. However on 
>the other hand is the scanners need for speed. The scanner accuracy is 
>dependent on its inertia which is the product of rotational mass by the 
>rotational speed squared. So obviously there are limits that will dictate 
>the minium laser power required.
>
>Thanks.
>
>----- Original Message ----- 
>From: "Mark Lerman"
>
>Well, I'm not doing toner transfer, actually, I'm using a modified
>laser printer to print directly on the board - the toner is still in
>powder form, so it is relatively easy for it to smudge. I suspect
>I'll have to use a hot plate and directly compress the board between
>two plates using a teflon sheet between the toner and the top plate.
>The laminator caused only a little smudging, but it was my first
>attempt and I used the carrier. Perhaps just a teflon sheet will
>allow use of the laminator. We'll see.
>
>Mark
>
>
>------------------------------------
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
>
>
>

I'm surprised to hear that printing directly on copper works without
any kind of coating.  I've tried that in many different laser printers
and it's never even come close to working without the gel.  I guess
the straight path you've set up combined with the flat board allows
the toner to sit undisturbed and therefore not smear.  With the copper
foil going through the various bends of the printer, the (dried) gel
allows the toner to stick to the copper.

Dave


--- In Homebrew_PCBs@yahoogroups.com, mlerman@... wrote:
>
> I've spent the last month or so looking at all the discussions re
direct laser printing on boards. There are all sorts of warnings, nay
sayers, etc who say it can't work. Then I saw the post and site of
laserpcb, where Dave says (he has a patent)that he can print on copper
foil using his special "gel" then glue it to FR4 to make a PCB. So, if
it is possible to put toner on copper foil, it should be possible to
put it on boards directly.
> 
> I took a Samsung ML-2510 and modified the paper path to be straight
and flat, and moved the feed sensor (also inverted it) then I simply
ran a board through. No problem. A few sparks, but the image seems
very good, no distortion, very crisp. I tried Dave's Gel (sorry Dave),
but the image seems identical with and without the gel. 
> 
> I haven't finished experimenting, which is why I wanted the test
pattern. I seem to easily be able to print 1 mil traces, but I'm not
sure yet of the spacing I can use, though I would guess 5-8 mil. 
> 
> The board comes out of the printer with the toner as powder
electrostatically stuck to the board, so it has to be fused. I am
reluctant to use the fuser on the printer for several reasons. One is
that it is a pain to move it to a flat path - I would have to move a
gear, among other things. Secondly, the speed might easily be too fast
to fuse the toner. Also, if the toner is smudged or not perfect, it is
very easy to wipe and reprint the board if it isn't fused. 
> 
> An advantage of fusing it is that it (hopefully) will be stuck in
place, even if it isn't fused completely. Obviously there is a lot of
work to do here.
> 
> I think I have demonstrated that direct laser printing of pcb's is
possible. If anyone wants further information, I will be happy to share.

> 
> Mark
>

Hello Mark,
                  Thank you for responding. Your advise is greatly 
appreciated.

I had a look at the web site and method that you mentioned. I see two 
distinct differences in your approach.

Firstly the web site shows a full contact method. The copper foil is coming 
into full contact with the image unit, fuser and other less significant 
parts.

I can see a number of problems with this -
1) The image unit is likely to be damaged, particularly where the edges of 
the copper foil contacts the image unit. The same applies for the fuser 
roller especially the more modern type.
2) There are safety issues as the foil is still in contact with the charged 
image unit at the time that it is being handled by the operator.
3) The jell is coming into full contact with the image unit and fuser.

I doubt the jell is able to act as an insulator so I suspect the jell is 
what is transferring the toner to the copper as the copper will be at the 
same electrical charge potential as the image drum. It may also may be 
increasing the thermal transfer from the fuser to the copper.

In your method you mentioned sparks so I assume the PCB has no physical or 
electrical contact with the image unit? Normally I would expect the PCB to 
float in voltage between the potential's of the image unit and charge wire. 
The existence of sparks indicates that either it is static discharge in 
which case there is not likely to be more than one, possibly two, sparks on 
any print. Or the sparks are a result of a discharge path meaning that for 
each obvious spark there is a corresponding spark from the copper to another 
part of the printer.

Ideally (from a image point of view and not safety) the copper should be 
electrically connected to the charge voltage or perhaps more appropriately a 
division of the charge voltage.

I can see that you know what you are doing to a fair degree. I do not know 
how experienced you are with electronics and more particular high voltage 
electrostatics. I can provide some information that my assist you but I warn 
that this information is provided in a context only suitable to qualified 
technician such as myself. You may need to seek independent advice for 
further explanation of safety procedures.

The HT generator in your printer is most likely a split rail generator. The 
ideal voltage for the copper on the PCB (to prevent electrostatic discharge) 
can be calculated this way -

((drum voltage)-(charge voltage))*((distance from the PCB to the image 
drum)/(distance from charge wire to drum))

On a single rail supply a divider can be made with resistors that or 
normally used in the focus electronics of a television. These are available 
for domestic electronic repair shops or a good electronics supplier. Normal 
resistors are not at all useful.

If you are a technician (or have a technician friend) and know how to 
measure the drum and charge voltages with respect the chassis ground then I 
can tell you how to build a simple circuit to fix this problem and the 
associated safety hazards. This however cannot be done with a simple 
multi-meter and in fact it is downright dangerous to attempt to use a common 
garden variety multi-meter without an external divider or HT probe.

I also need to know if there are two sparks for each flashover and where 
they are located or if the is just one and where it is located.

I may be able to get some specs for your printer online but it is unlikely 
with newer models.

 Getting back to the web site mentioned -

At least this shows that a normal fuser will work. While you mentioned that 
it is convenient to be able to blow away the toner on a poor copy, I expect 
in the long run you will have consistent results and it may be more 
convenient to use the fuser directly. If possible, this may not always be 
the case for different fuser types.

The better short term solution would be an oven with good temperature 
regulation. The closer you can get to the ideal (lower end) temperature the 
longer the fusing will take. This gives the greatest margin for error in the 
timing of the heating process. Obviously the board will have be handled very 
carefully as even a breath to close to the board will disturb the toner 
pattern.

I hope you don't feel that I am a "Nay sayer". I have confidence in your 
project. I am happy to help if my 'opinion' is useful.

If you are a technician or you find my comments irrelevant than I am quite 
happy not to bother you.

Thanks, Robert.

----- Original Message ----- 
From: Mark

I've spent the last month or so looking at all the discussions re direct 
laser printing on boards. There are all sorts of warnings, nay sayers, etc 
who say it can't work. Then I saw the post and site of laserpcb, where Dave 
says (he has a patent)that he can print on copper foil using his special 
"gel" then glue it to FR4 to make a PCB. So, if it is possible to put toner 
on copper foil, it should be possible to put it on boards directly.

I took a Samsung ML-2510 and modified the paper path to be straight and 
flat, and moved the feed sensor (also inverted it) then I simply ran a board 
through. No problem. A few sparks, but the image seems very good, no 
distortion, very crisp. I tried Dave's Gel (sorry Dave), but the image seems 
identical with and without the gel.

I haven't finished experimenting, which is why I wanted the test pattern. I 
seem to easily be able to print 1 mil traces, but I'm not sure yet of the 
spacing I can use, though I would guess 5-8 mil.

The board comes out of the printer with the toner as powder 
electrostatically stuck to the board, so it has to be fused. I am reluctant 
to use the fuser on the printer for several reasons. One is that it is a 
pain to move it to a flat path - I would have to move a gear, among other 
things. Secondly, the speed might easily be too fast to fuse the toner. 
Also, if the toner is smudged or not perfect, it is very easy to wipe and 
reprint the board if it isn't fused.

An advantage of fusing it is that it (hopefully) will be stuck in place, 
even if it isn't fused completely. Obviously there is a lot of work to do 
here.

I think I have demonstrated that direct laser printing of pcb's is possible. 
If anyone wants further information, I will be happy to share.

Mark

Mark, 

Intriguing conversion!

What happens if instead of using a fuser you carefully just heat the board
until the toner starts to melt and get sticky?

Bertho

From: Mark      Sent: Wednesday, March 26, 2008 22:40
--- > snip

 The board comes out of the printer with the toner as powder
electrostatically stuck to the board, so it has to be fused. I am
reluctant to use the fuser on the printer for several reasons. One is
that it is a pain to move it to a flat path - I would have to move a
gear, among other things. Secondly, the speed might easily be too fast
to fuse the toner. Also, if the toner is smudged or not perfect, it is
very easy to wipe and reprint the board if it isn't fused. 
> 
> An advantage of fusing it is that it (hopefully) will be stuck in
place, even if it isn't fused completely. Obviously there is a lot of
work to do here.
> 
> I think I have demonstrated that direct laser printing of pcb's is
possible. If anyone wants further information, I will be happy to share.
> 
> Mark





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

Hello Mark,
                  For the want of a simple answer :(

NORMAL fuser temperatures range from about 160^C to about 200^C. Some even 
venture outside this.

In reality it is not temperature that is important but (mili) watts per 
pixel. If you imagine the toner to be a bitmap then each pixel requires the 
same power. More pixel's means higher power. Thicker paper means more power. 
However the pixel's are normally more important than paper thickness.

Modern fusers are predictive. The CPU knows how many pixel's are across the 
(each) next line coming to the fuser and how slowly the fuser heats due to 
its thermal mass so it can correct for this by increasing the supply power 
timed in advance to provide the optimum power at the correct time.

Older hot roller fusers had too much thermal mass for such predictive 
calculations to have any effect. These at a guess would have been in the 
rang 165^C to 195^C.

Copper boards however are a different story as copper is one of the best 
thermal conductors. This introduces other considerations.

One approach is to use a higher temperature and a higher feed rate so that 
the toner dissipates most of the energy before it can transfer (in lager 
quantities as temperature) to the copper laminate. A slower feed will have 
the effect of preheating the toner before it gets to the heater roller which 
in turn causes a higher temperature when it does get there. Ironically a 
smaller heater roller that has a lower thermal mass, will loose temperature 
before getting to the preheated toner cancelling some or all of this effect. 
Keep in mind that a heat roller rotates several times over one print.

Another approach is to use an oven in which case it is best to preheat the 
oven to just below the liquid/soft plastic state of the toner. After some 
time it can be considered the thermal mass of the laminate has evened out 
with the toner. You could then increase the temperature at a set power rate 
and time the process or complete the process at a fixed temperature. On a 
lower powered oven it may be necessary to raise the temperature to a fixed 
maximum and leave it there for a fixed time.

I would pre heat an oven to about 120^C to 130^C and then place the board 
into it for 15 minutes or so. Then increase the temperature control to about 
165^C to 170^C and watch the toner carefully. A minute or so after the toner 
appears more glossy I would remove it. Some toners may not have the glossy 
property so it would need to be timed. The most critical issue is the 
maximum temperature. There is a temperature rage within which the toner 
melts slowly. The lower the temperature you can get it to do this, the 
better as timing then becomes less critical. The copper has a high thermal 
mass so if it is over heated it will stay to hot for a while. So if the 
toner gets to close to melting point then the toner pattern will most likely 
be ruined.

Thanks Rob.

----- Original Message ----- 
From: Mark

As for laminators, I just bought a Staples machine - no adjustments. I also 
just ordered a pid temperature controller that I will use either with the 
laminator or with a "hot plate". Does anyone know the proper fuser 
temperature??

BTW, I'm not sure whether registration for double sided boards will work - 
it would require a precision optical sensor plus a rigid mechanical guide. 
The simplest way is probably to use two single sided boards of half the 
thiskness and use registration pins for alignment, then glue them together. 
Should work fine.

Mark
---------------------------

Well, what did you expect when you buy a staple machine instead of a
laminator ;-)

Sorry, couldn't resist.

ST



On Thu, Mar 27, 2008 at 2:22 AM,  <mlerman@...> wrote:
> As for laminators, I just bought a Staples machine - no adjustments.

Thanks, Rob, for the detailed information - very informative. The 
spark seems to be a single spark, not always present, seen after the 
board leading edge leaves the printer. I've been assuming it's static 
electricity. I've been handling the boards as they leave the printer 
(careless??) with no problems, i.e. no shock.

The ML-2510 is an inexpensive Samsung printer in which the real 
action takes place as the paper rides up along the rear of the 
printer. There is a set of powered rollers just below the drum that 
helps push the board up through the drum mechanism. The drum has it's 
own geared rollers as well, but I think the push from the bottom 
really helps keep the speed constant. The board only travels about 2 
inches inside the printer, feeding in from the bottom, out from the top.

I have a service manual for the printer if you would like to take a 
look. Unfortunately the schematic pages are not there, even though 
the site I bought it from said they would be.

BTW, I get really nice prints on aluminum sheet, too.

Mark

--- Mark Lerman <mlerman@...> wrote:

> Thanks, Rob, for the detailed information - very
> informative. The 
> spark seems to be a single spark, not always
> present, seen after the 
> board leading edge leaves the printer. I've been
> assuming it's static 
> electricity. I've been handling the boards as they
> leave the printer 
> (careless??) with no problems, i.e. no shock.
> 
Try beveling the edges of the copper on the leading
edge.  High voltage doesn't like sharp edges.  Just
light sind or file the edges of the board.  See if
that helps.

A slight radius on the corners could help too.


      ____________________________________________________________________________________
Never miss a thing.  Make Yahoo your home page. 
http://www.yahoo.com/r/hs

I'm waiting for some 1/32 board to arrive, but meanwhile I'm using 
some very thin (.010) DS board and some 1/32 aluminum as well as 1/16 
pcb to experiment with. The toner seems pretty well fixed on the 
board - even a moderate puff will not dislodge it. I think it's going 
to take a fair amount of experimentation, but it should work.
I just ordered a simple pid controller that I'm going to try to hook 
to an electric skillet to fuse the toner. I'm also going to see if I 
can move the printer's fuser back the inch or so it needs to go to 
provide a straight path. I have to move a gear and remount the fuser, 
but it looks like there is enough room. Unfortunately, my machining 
skills leave much to be desired. If it works, it will pretty neat - 
just print the board, etch and drill!

Mark


At 10:16 AM 3/27/2008, you wrote:

>--- Mark Lerman <mlerman@...> wrote:
>
> > Thanks, Rob, for the detailed information - very
> > informative. The
> > spark seems to be a single spark, not always
> > present, seen after the
> > board leading edge leaves the printer. I've been
> > assuming it's static
> > electricity. I've been handling the boards as they
> > leave the printer
> > (careless??) with no problems, i.e. no shock.
> >
>Try beveling the edges of the copper on the leading
>edge.  High voltage doesn't like sharp edges.  Just
>light sind or file the edges of the board.  See if
>that helps.
>
>A slight radius on the corners could help too.
>
>
> 
>____________________________________________________________________________________
>Never miss a thing.  Make Yahoo your home page.
>http://www.yahoo.com/r/hs
>
>------------------------------------
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBsYahoo! Groups Links
>
>
>

I revisited Dave's gel, and it does seem to hold the toner better 
than without it. I printed a board using it, no fusing. I can turn 
the board on it's side and rap it against my desk without dislodging 
any of the toner.

Mark

Hi ,
I am a hobby electronic and look for a easy way to make my own pcb`s.

Thanks 
Hans




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

Hans:

Thanks for joining, it nice to have new members. You may want to review the
pages associated with this mailing list. There are numerous tutorial links
contained here if you search the list. My own views are that the toner
transfer method using cheap magazine paper is best for my budget and time.
Others would disagree, but the UV methods require some specialized
equipment, but would yield better overall results. I might have to include a
SSOP part soon, so I might be up against a hard limit of the method that I
use. I can't see getting a .6 mm lead spacing with the toner transfer
method.

Chris




On Mon, Feb 21, 2011 at 5:52 AM, <HaHoLang@...> wrote:

>
>
>
>
>
> Hi ,
> I am a hobby electronic and look for a easy way to make my own pcb`s.
>
> Thanks
> Hans
>
> [Non-text portions of this message have been removed]
>
>  
>


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

Chris Kleeschulte <laconia@...> writes:
> Others would disagree, but the UV methods require some specialized
> equipment, but would yield better overall results. I might have to
> include a SSOP part soon, so I might be up against a hard limit of the
> method that I use. I can't see getting a .6 mm lead spacing with the
> toner transfer method.

I wish I could get toner to be as reliable as UV, so that boards that
don't require small traces can get done faster and cheaper, but I've
found - in my case - I can *finish* a UV board faster because I know
it's going to "just work".  It seems I can never get a good-enough toner
transfer any more, there's always a broken trace somewhere.  When I get
some free time, I need to throw some science at it and figure out what's
going wrong.

I can easily do 8/8, and with care 5/5, with UV - but it's more steps
and more $$$ per board.

Chris,

.6mm lead spacing should be quite easy to do with TT, in my experience. 0.6mm = .024"; 10/10 (eg, 10 mils traces, 10 mils spaces, with 1 mil = .001") is my standard design rule with TT, and I can produce boards with this very reliably and consistently. The most recent board I did, using a good bit of SMD, used 8/8 design rules; it came out flawlessly. Either of these will work fine with your SSOP parts.

--- In Homebrew_PCBs@yahoogroups.com, Chris Kleeschulte <laconia@...> wrote:

>
> Hans:
> 
> Thanks for joining, it nice to have new members. You may want to review the
> pages associated with this mailing list. There are numerous tutorial links
> contained here if you search the list. My own views are that the toner
> transfer method using cheap magazine paper is best for my budget and time.
> Others would disagree, but the UV methods require some specialized
> equipment, but would yield better overall results. I might have to include a
> SSOP part soon, so I might be up against a hard limit of the method that I
> use. I can't see getting a .6 mm lead spacing with the toner transfer
> method.
> 
> Chris
> 
> 
> 
> 
> On Mon, Feb 21, 2011 at 5:52 AM, <HaHoLang@...> wrote:
> 
> >
> >
> >
> >
> >
> > Hi ,
> > I am a hobby electronic and look for a easy way to make my own pcb`s.
> >
> > Thanks
> > Hans
> >
> > [Non-text portions of this message have been removed]
> >
> >  
> >
> 
> 
> [Non-text portions of this message have been removed]
>

I do not have any sources of tin, but I suspect a web search would 
turn up some.

I do want to make you aware that there are sources for a tin plating 
solution that only requires that you dip the copper clad board in it 
for a few minutes at room temperature.

http://www.vetco.net/catalog/product_info.php?products_id=936&gclid=CNPC5-rNq7wCFeZj7AodPA8ALg

There are other sources and a web search will turn up many. I have 
used such solutions and they work quite well with no fuss.

Paul A.