Searching \ for '[EE] Make Your Own PCB' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/pcbs.htm?key=pcb
Search entire site for: 'Make Your Own PCB'.

Exact match. Not showing close matches.
PICList Thread
'[EE] Make Your Own PCB'
2008\12\06@005835 by solarwind

picon face
By using techniques at home to make your own PCB, is it practical and
possible to make a board capable of mounting very fine pitch SOIC
chips (such as the PIC32, which is 0.4 mm pitch)?

--
..::[ solarwind ]::..

2008\12\06@023505 by Jinx

face picon face
> possible to make a board capable of mounting very fine pitch SOIC
> chips (such as the PIC32, which is 0.4 mm pitch)?

FWIW I was considering making my own for the PICs that came with
the USB kit. Having seen them, I've possibly changed my mind. 1mm I
can do reasonably competently, 0.4mm is another story

It's not so much the IC footprint, although that's going to require patience.
The problem is getting the tracks away from the pins

*IF* I was to try this, I'd make a solid block of etch resist and then scratch
out the gaps under good lighting and magnification. It's what I'm doing right
now for a one-off personal job that has a few TSOP logic. Attempting to
draw tracks directly close to the IC will end in tears of frustration

The scratcher I use is a needle epoxied into a pen, works well

http://home.clear.net.nz/pages/joecolquitt/0makepcb.html

Last year I had to make a pad for a 60-pin modem connector that was
0.4mm pitch. It did finally come out OK, maybe my experience with that
makes me cautious now. Although one big problem with that project was
that the plastic used for the socket was very heat sensitive and distorted,
unlike an IC

So, yes, it is possible. For a one-off needed now I'd give it a go, for a few
now or to put away, I'd have them made, no question



2008\12\06@025141 by Vitaliy

flavicon
face
solarwind wrote:
> By using techniques at home to make your own PCB, is it practical and
> possible to make a board capable of mounting very fine pitch SOIC
> chips (such as the PIC32, which is 0.4 mm pitch)?

If you use the photo method (using presensitized PCBs) may give you enough
accuracy. Laser printer transfer method, probably not practical. I've done
both in the past, but not for 0.4 mm pitch devices.

Unless you are on a very tight budget, I would recommend ordering your PCB
from a professional board house. Advanced Circuits (4pcb.com) let you buy a
two-layer PCB for $33, as long as you ship it to a school address. Then
there's Sparkfun's BatchPCB.com: http://www.batchpcb.com/

Vitaliy

2008\12\06@060256 by Mike Harrison

flavicon
face
On Sat, 6 Dec 2008 00:58:14 -0500, you wrote:

>By using techniques at home to make your own PCB, is it practical and
>possible to make a board capable of mounting very fine pitch SOIC
>chips (such as the PIC32, which is 0.4 mm pitch)?
>

I'd say .4mm is just doable given the right technique & a good transparency, but only just
see  http://www.electricstuff.co.uk/pcbs.html  for info on making good homemade PCBs



2008\12\06@145851 by William \Chops\ Westfield

face picon face

On Dec 5, 2008, at 9:58 PM, solarwind wrote:

> By using techniques at home to make your own PCB, is it practical and
> possible to make a board capable of mounting very fine pitch SOIC
> chips (such as the PIC32, which is 0.4 mm pitch)?

No, not really.  You may be able to achieve that sort of line pitch,  
but by the time you get down there, there are a bunch of other things  
your PCB is expected to have (soldermask, plated-through holes and  
vias, for example) that are hard to do at home that you will really  
miss...

BillW

2008\12\06@153710 by solarwind

picon face
On Sat, Dec 6, 2008 at 2:58 PM, William Chops Westfield <spam_OUTwestfwTakeThisOuTspammac.com> wrote:
> No, not really.  You may be able to achieve that sort of line pitch,
> but by the time you get down there, there are a bunch of other things
> your PCB is expected to have (soldermask, plated-through holes and
> vias, for example) that are hard to do at home that you will really
> miss...

So you recommend just getting it done by some company?

If I want just one or two boards, whats the best place to get them made?


--
..::[ solarwind ]::..

2008\12\06@181644 by microsoftwarecontrol

flavicon
face
I make my own pcb, easily with 10mil,  TSSOP footprint, double sided. The
quality is acceptible and normally taking 2 hour for go through all
steps: diagram, pcb, pcb output, printing, cleaning board, tranfering to
board,
cleaning again, etching, cleaning, drilling, make via. Solding SMT parts is
very
fast and reliable process. My etching tank is 100mm * 150 mm.




----- Original Message -----
From: "solarwind" <.....x.solarwind.xKILLspamspam@spam@gmail.com>
To: "Microcontroller discussion list - Public." <piclistspamKILLspammit.edu>
Sent: Saturday, December 06, 2008 12:58 AM
Subject: [EE] Make Your Own PCB


> By using techniques at home to make your own PCB, is it practical and
> possible to make a board capable of mounting very fine pitch SOIC
> chips (such as the PIC32, which is 0.4 mm pitch)?
>
> --
> ..::[ solarwind ]::..
> --

2008\12\06@182418 by sergio masci

flavicon
face


On Sat, 6 Dec 2008, Mike Harrison wrote:

> On Sat, 6 Dec 2008 00:58:14 -0500, you wrote:
>
> >By using techniques at home to make your own PCB, is it practical and
> >possible to make a board capable of mounting very fine pitch SOIC
> >chips (such as the PIC32, which is 0.4 mm pitch)?
> >
>
> I'd say .4mm is just doable given the right technique & a good transparency, but only just
> see  http://www.electricstuff.co.uk/pcbs.html  for info on making good homemade PCBs

Getting a GOOD UV transparency is a big hurdle. Check the web and you'll
find that not all laser printers give the same good results. Some people
have had better luck with inkjets, others use two layers of tracing paper
per side.

I have been toying with the idea of simply printing an enlarged image of
the tracks to white glossy paper then using this to expose the photoresist
via a "camera obscura". One could use a normal light source to set up the
artwork and PCB inside the camera then switch to UV to expose the PCB (yes
I know the wavelength of the light used is important when determining the
position of the focal plane).

Regards
Sergio Masci

2008\12\06@182437 by solarwind

picon face
On Sat, Dec 6, 2008 at 6:15 PM, microsoftwarecontrol
<.....microsoftwarecontrolKILLspamspam.....yahoo.ca> wrote:
> I make my own pcb, easily with 10mil,  TSSOP footprint, double sided. The
> quality is acceptible and normally taking 2 hour for go through all
> steps: diagram, pcb, pcb output, printing, cleaning board, tranfering to
> board,
> cleaning again, etching, cleaning, drilling, make via. Solding SMT parts is
> very
> fast and reliable process. My etching tank is 100mm * 150 mm.

What is "10mil"? What unit is that?

--
..::[ solarwind ]::..

2008\12\06@190756 by Jinx

face picon face
> What is "10mil"? What unit is that?

1mil = 1/1000th inch, or 0.0254mm

Many PCB board houses would probably recommend a lower limit of
8mil track width and spacing, but they can go lower

0.4mm is about 16mil, to give you an idea of scale

0402 (40mil x 20mil, 1mm x 0.5mm) components are probably at the limit
of hand-soldering, if you want to do the job in a reasonable time. I've used
them when I had to, but would rather not

http://articulationllc.home.comcast.net/~articulationllc/sm0402.htm

Unless you're really tight on board space or trying to impress your mother,
0805 or 1206 are much more practical





2008\12\06@191259 by Tamas Rudnai

face picon face
> What is "10mil"? What unit is that?

http://www.google.ie/search?hl=en&q=10+mil+in+inch&btnG=Search&meta=

Tamas



On Sat, Dec 6, 2008 at 11:24 PM, solarwind <EraseMEx.solarwind.xspam_OUTspamTakeThisOuTgmail.com> wrote:

{Quote hidden}

> -

2008\12\06@191401 by peter green

flavicon
face

> What is "10mil"? What unit is that?
>  
When discussing pcb specs the term mil typically reffers to a thousanth
of an inch.

This can be very confusing to people who take the term mil to mean
millemeter.

2008\12\06@191743 by apptech

face
flavicon
face
"microsoftwarecontrol" Said

> steps: diagram, pcb, pcb output, printing, cleaning board, tranfering to
> board,

What do you do for "printing"?
There are many processes that would fit under this name.
It sounds like you have better than usual success so I'm sure others would
be interested in your methods. I would be.


   Russell

2008\12\06@192628 by solarwind

picon face
Unfortunately, the PIC32 I ordered has 0.4 mm pitch which is pretty
tight, but I did buy this for it:

http://futurlec.com/SMD_Adapters.shtml (last item at the bottom of the page).

It's gonna be a blast to solder. Lots of liquid flux and a roll of
copper wick should do the trick.

--
..::[ solarwind ]::..

2008\12\06@192919 by Funny NYPD

picon face
mil is mili-inch, 1/1000 of a inch.
It is used in some PCB layout software.

Funny N.
Au Group Electronics, http://www.AuElectronics.com




________________________________
From: Tamas Rudnai <@spam@tamas.rudnaiKILLspamspamgmail.com>
To: Microcontroller discussion list - Public. <KILLspampiclistKILLspamspammit.edu>
Sent: Saturday, December 6, 2008 7:12:34 PM
Subject: Re: [EE] Make Your Own PCB

> What is "10mil"? What unit is that?

http://www.google.ie/search?hl=en&q=10+mil+in+inch&btnG=Search&meta=

Tamas



On Sat, Dec 6, 2008 at 11:24 PM, solarwind <RemoveMEx.solarwind.xTakeThisOuTspamgmail.com> wrote:

{Quote hidden}

> -

2008\12\06@193922 by Funny NYPD

picon face
well, if you can solder the chip on the adapter board, what has prevent you from directly solder the chip to the actual board?

Funny N.
Au Group Electronics, http://www.AuElectronics.com




________________________________
From: solarwind <TakeThisOuTx.solarwind.xEraseMEspamspam_OUTgmail.com>
To: Microcontroller discussion list - Public. <RemoveMEpiclistspamTakeThisOuTmit.edu>
Sent: Saturday, December 6, 2008 7:26:01 PM
Subject: Re: [EE] Make Your Own PCB

Unfortunately, the PIC32 I ordered has 0.4 mm pitch which is pretty
tight, but I did buy this for it:

http://futurlec.com/SMD_Adapters.shtml (last item at the bottom of the page).

It's gonna be a blast to solder. Lots of liquid flux and a roll of
copper wick should do the trick.

--
..::[ solarwind ]::..

2008\12\06@194554 by Jinx

face picon face

> Unfortunately, the PIC32 I ordered has 0.4 mm pitch

> It's gonna be a blast to solder. Lots of liquid flux and a roll of
> copper wick should do the trick.

With experience and the right bit you can reflow using just the iron. The
trick is to make sure everything is clean and pre-tinned and use the very
tip of the bit

2008\12\06@194755 by Joseph Bento

face
flavicon
face

On Dec 6, 2008, at 5:06 PM, Jinx wrote:
>
> 0402 (40mil x 20mil, 1mm x 0.5mm) components are probably at the limit
> of hand-soldering, if you want to do the job in a reasonable time.  
> I've used
> them when I had to, but would rather not
>
> http://articulationllc.home.comcast.net/~articulationllc/sm0402.htm
>
> Unless you're really tight on board space or trying to impress your  
> mother,
> 0805 or 1206 are much more practical


I'm inclined to agree.  0805 can be soldered with reasonably good  
eyesight without magnification.  0603 is easier with magnification,  
and 0402 it is a must - for my 45 year old eyes anyway.  I'd rather  
stay away from 0201 components, and fortunately the engineers at my  
company have obliged.

My normal construction method involves a hypodermic applicator for  
paste solder, tweezers, and a 3 diopter magnification lamp.  (I've  
found that magnification greater than 3 diopter will not allow the  
tweezers or soldering iron to fit between the magnifier and board).  
Soldering is done either by hand with a Hakko temperature controlled  
iron or hotplate as described below.

I'll generally apply the solder paste, and place all the components on  
the most populated side of the board (if double sided).  A hotplate,  
preheated to 200c is in wait, and the populated board is carefully  
placed on the hotplate.  It's quite amazing if one hasn't seen it  
before to watch the components flow perfectly in place and center  
themselves on the pads.  The board is carefully observed, and once all  
the paste has melted, I'll allow an additional 10 seconds or so, and  
then carefully slide the board off the hotplate onto an aluminum plate  
(I use the bottom cover from a Hammond chassis).  This will cool the  
board rather quickly, and then the board is inspected under the  
magnifier for any bridges on the chips, tombstoned components, etc.  
If the board is double-sided, the opposite side obviously must be  
entirely hand soldered.

Alas, this method is generally not acceptable to place BGAs.  I have  
never attempted to hand solder a BGA, though it is apparently possible.

My above method has been perfectly acceptable for engineering  
prototypes, and is certainly acceptable for a 1-off hobby project.  
Success rate has actually been close to 100%.  If we had more than  
five or so units to build, they'd usually be assembled at a board house.

One of the more fun things to do today (I say in jest) is to use  
perfboard or Veroboard to build a SMD prototype.  Kapton tape, super  
glue, and kynar wire are now your best friends!

Joe

2008\12\06@195733 by peter green

flavicon
face
Funny NYPD wrote:
> well, if you can solder the chip on the adapter board, what has prevent you from directly solder the chip to the actual board?
>  
If it is a professionally made board then sure.

But soldering to a homemade board with bare copper tracks and no
soldermask is going to be tricker (not impossible just far more likely
to produce solder bridges etc) is going to be a lot trickier than
soldering to a professionally made apaptor.

2008\12\06@201243 by solarwind

picon face
On Sat, Dec 6, 2008 at 7:44 PM, Jinx <joecolquittEraseMEspam.....clear.net.nz> wrote:
> With experience and the right bit you can reflow using just the iron. The
> trick is to make sure everything is clean and pre-tinned and use the very
> tip of the bit

Reflow?

--
..::[ solarwind ]::..

2008\12\06@205936 by Jinx

face picon face
> Reflow?

Meaning the components and track already have enough solder to make
a joint. For example IC leads are already coated with solder (tinned). If
the track is also coated with solder then all that's needed is to put the two
together, ie put the IC on its pad and apply heat. Tinned component leads
would probably not have enough solder to flow onto a bare copper track,
even using flux to reduce surface tension, so you need to feed solder in as
you heat. How much and whether you have to depends on the components

2008\12\07@082835 by microsoftwarecontrol

flavicon
face
laser printing is good enough, far from perfect but acceptible.

----- Original Message -----
From: "apptech" <EraseMEapptechspamparadise.net.nz>
To: "Microcontroller discussion list - Public." <RemoveMEpiclistEraseMEspamEraseMEmit.edu>
Sent: Saturday, December 06, 2008 7:17 PM
Subject: Re: [EE] Make Your Own PCB


{Quote hidden}

> --

2008\12\07@083841 by microsoftwarecontrol

flavicon
face
I modified 0805, making it can pass 15 mil track. It is very helpful!
Did same thing to sot-23, sot-23 liked port.

all pic footprint, I can do it reliably.
http://www.piclist.com/techref/TMPRmod.htm is making of mine years ago





{Original Message removed}

2008\12\07@084437 by microsoftwarecontrol

flavicon
face
For printing, I expert pcb pattern to pdf file.
perfect double side matching! Now it is trouble for me to make single sided
and I need use
tape to cover unused side.


{Original Message removed}

2008\12\07@085252 by microsoftwarecontrol

flavicon
face
I found, solding paste and hot gun, is already very reliable and fast
process.
I replaced TSSOP chip many times. no problem at all.

some tiny chip, lm74 CAN5 footprint, 2*2 SON footprint, I found is even more
easier to do than long line of foot on 64 pins TQFP.
They don't have "foot" only tiny pad!  Believing is duo to getting heat
easilier!



{Original Message removed}

2008\12\07@102857 by olin piclist

face picon face
solarwind wrote:
> What is "10mil"? What unit is that?

This is getting rediculous.  We are here to help with PICs and electronics,
but not to answer stupid questions that can be answered in the obvious
places with a few seconds of effort.  I went to Google and entered the
search string "mil unit of measure".  The very first hit was a dictionary
entry directly and clearly answering your question.  Don't be so lazy.

Now if you do your homework and don't understand what some reference is
trying to tell you after at least a little effort, then come here and ask.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\12\07@104648 by olin piclist

face picon face
Jinx wrote:
> Unless you're really tight on board space or trying to impress your
> mother, 0805 or 1206 are much more practical

Yeah, I second that.  I consider 0805 the "normal" package unless there is a
reason I need to use something different.  On physically tight designs 0603
and even 0402 may be necessary.  I've used 1206 resistors a few times when I
needed more power dissipation than the 125mW most 0805 packages are rated
for.  Another solution if you need more power dissipation, especially for
hobby purposes, is to put two 0805 resistors of half the value in series.  I
actually did that on a commercial product once because the 1/2 value 0805
resistor was already used elsewhere on that design.

As for the soldering iron, get something temperature controlled.  The days
of open loop 12.5W or 20W or 25W are over.  When you were soldering leads
into eyelets or the bottom of tube sockets it didn't matter much if the iron
got a few 100F hotter than optimal.  However nowadays you're soldering right
up against sensitive ICs and on PC boards.  Excessive heat does bad things
to both.  For hobby purposes, probably any temperature controlled soldering
iron (sometimes called soldering pencil) with a reasonably fine point
conical tip will do.  As someone else suggested, look for some used ones,
but be prepared that you may need to get a new tip.

When you get one, take care of the tip properly.  That first and foremost
means to keep the temperature low unless you have a good reason to raise is.
I use the Weller WES50 and WES51s we have around here at 600F for normal
use, and sometimes crank it to 650-700F for desoldering.  Keep the tip
tinned regularly, and always tin the tip right before shutting the solder
station off.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\12\07@105455 by Byron Jeff

flavicon
face
On Sun, Dec 07, 2008 at 10:28:41AM -0500, Olin Lathrop wrote:
> solarwind wrote:
> > What is "10mil"? What unit is that?
>
> This is getting rediculous.  We are here to help with PICs and electronics,
> but not to answer stupid questions that can be answered in the obvious
> places with a few seconds of effort.  I went to Google and entered the
> search string "mil unit of measure".  The very first hit was a dictionary
> entry directly and clearly answering your question.  Don't be so lazy.
>
> Now if you do your homework and don't understand what some reference is
> trying to tell you after at least a little effort, then come here and ask.

Olin,

Do you think that maybe it's time to reference the excellent tutorial:

"How to ask questions the smart way." by Eric Raymond.

http://www.catb.org/~esr/faqs/smart-questions.html

It seems to laser in on exactly the issues you refer to above.

Solarwind, it may be worth taking the 20 minutes to read the document. It
offers great advise.

BAJ

2008\12\07@105550 by olin piclist

face picon face
solarwind wrote:
> Unfortunately, the PIC32 I ordered has 0.4 mm pitch

That's a rather strange choice of PIC.  You originally asked what PIC to
use.  Most people suggested a large 18F.  While most PIC32 should be well
capable of the task, the downside is that few people have experience with
them and it will be much harder to get help.  You're a lot more on your own
with a PIC32.  You definitely need to read the data sheet, carefully,
several times (not that wouldn't have to with a PIC 18, but now you really
really need to do your own homework).

> http://futurlec.com/SMD_Adapters.shtml (last item at the bottom of
> the page).

That looks like a good idea for one-off hobby use.

> It's gonna be a blast to solder. Lots of liquid flux and a roll of
> copper wick should do the trick.

I still don't understand why you need a package with so many pins.  Have you
figured out how many pins you really need for each function?  Wouldn't a 40
pin PIC do it?  Being able to make do with a 40 pin PIC is desirable because
that's the largest size that comes in DIP package.  For one off hobby use,
it may actually be better to use external multiplexers so that a 40 pin DIP
package can be used than to mess with a .4mm pitch SMD package.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\12\07@105718 by olin piclist

face picon face
Funny NYPD wrote:
> well, if you can solder the chip on the adapter board, what has
> prevent you from directly solder the chip to the actual board?

The fact that he would make the actual board, whereas the adapter was made
by a real PCB house that can easily handle such dimensions.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.

2008\12\07@112611 by solarwind

picon face
On Sun, Dec 7, 2008 at 10:55 AM, Olin Lathrop <RemoveMEolin_piclistspam_OUTspamKILLspamembedinc.com> wrote:
> That's a rather strange choice of PIC.  You originally asked what PIC to
> use.  Most people suggested a large 18F.  While most PIC32 should be well
> capable of the task, the downside is that few people have experience with
> them and it will be much harder to get help.

I have 10 PIC 12s, 16 PIC 16s, 1 PIC18 and 1 PIC32.

--
..::[ solarwind ]::..

2008\12\07@122133 by Wouter van Ooijen

face picon face
> If I want just one or two boards, whats the best place to get them made?

consider http://www.olimex.com, but do read their info carefully.


--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\07@140808 by Wouter van Ooijen

face picon face
> If the board is double-sided, the opposite side obviously must be  
> entirely hand soldered.

I used to do SMD (0805 and larger) by hand: patch of solder in one pad,
heat with tip of iron, put component in place with tweezer, solder the
other pad. No problems, but it takes time (but still less time than TH
components).

I just got a solder paste stencil, so I tried a PCB using that method.
Very easy, and a lot quicker. I used a paint stripper. Not ideal, but
certainly doable. (I will probably either build or buy an oven, or maybe
a hot-plate.) The flip side has some larger SMD components (SOIC28 and
some FETs), those were easy too. The main problem is how to use the
stencil when the bottom side is already populated, so it isn't an even
surface.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\07@145704 by Funny NYPD

picon face
You can use some water-soluble glue for two side SMD components.

Funny N.
Au Group Electronics, http://www.AuElectronics.com




________________________________
From: Wouter van Ooijen <RemoveMEwouterTakeThisOuTspamspamvoti.nl>
To: Microcontroller discussion list - Public. <EraseMEpiclistspamspamspamBeGonemit.edu>
Sent: Sunday, December 7, 2008 2:07:08 PM
Subject: Re: [EE] Make Your Own PCB

> If the board is double-sided, the opposite side obviously must be  
> entirely hand soldered.

I used to do SMD (0805 and larger) by hand: patch of solder in one pad,
heat with tip of iron, put component in place with tweezer, solder the
other pad. No problems, but it takes time (but still less time than TH
components).

I just got a solder paste stencil, so I tried a PCB using that method.
Very easy, and a lot quicker. I used a paint stripper. Not ideal, but
certainly doable. (I will probably either build or buy an oven, or maybe
a hot-plate.) The flip side has some larger SMD components (SOIC28 and
some FETs), those were easy too. The main problem is how to use the
stencil when the bottom side is already populated, so it isn't an even
surface.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\07@150231 by Joseph Bento

face
flavicon
face

On Dec 7, 2008, at 12:07 PM, Wouter van Ooijen wrote:

>>
> I just got a solder paste stencil, so I tried a PCB using that method.
> Very easy, and a lot quicker. I used a paint stripper. Not ideal, but
> certainly doable. (I will probably either build or buy an oven, or  
> maybe
> a hot-plate.) The flip side has some larger SMD components (SOIC28 and
> some FETs), those were easy too. The main problem is how to use the
> stencil when the bottom side is already populated, so it isn't an even
> surface.
>

When I saw a stencil used, the board (usually several boards prior to  
separating at the scribe line) was placed in a frame, a blob of paste  
was placed over the stencil, and a squeegee was passed several times  
across the stencil.  Following this, the board was placed in a pick  
and place machine, though I imagine parts can be placed by hand.

I have seen in some magazines (Elektor, I believe) where a modified  
toaster oven is used for SMD soldering.  The lower and upper heating  
elements are individually controlled (a PIC-based thermostat  
control).  This apparently allows enough temperature control to solder  
the opposite side of the board without components falling off the top  
side.

It would be interesting to try the toaster oven approach.  As a  
hobbyist, I couldn't afford nor justify any professional grade SMD  
equipment.

2008\12\07@162644 by Vitaliy

flavicon
face
sergio masci wrote:
> Getting a GOOD UV transparency is a big hurdle. Check the web and you'll
> find that not all laser printers give the same good results. Some people
> have had better luck with inkjets, others use two layers of tracing paper
> per side.

There are presensitized PCB blanks that can be developed using a normal 100W
bulb. I used the Philmore-Datak blanks, and they work great:

http://www.minute-man.com/acatalog/Online_Catalog_Pre_Sensitized_PC_Boards_240.html

What would be the advantage of using UV-sensitive photoresist?

Vitaliy

2008\12\07@163317 by Wouter van Ooijen

face picon face
> It would be interesting to try the toaster oven approach.  As a  
> hobbyist, I couldn't afford nor justify any professional grade SMD  
> equipment.

I ordered a stencil from http://www.smtstencil.co.uk, used a piece of a CD box
as scraper, a paint stripper for heating, and the metal core of a roll
of wire (happened to be in my dustbin) as a standoff between the PCB and
my (wooden) desk. Definitely not a professional setup, but enough for a
proof of concept :)

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\07@163449 by Wouter van Ooijen

face picon face
Funny NYPD wrote:
> You can use some water-soluble glue for two side SMD components.

As I said, no need for that when using a paint stripper. The main
problem is stenciling the paste on the top side when the bottom side
already has components.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\07@173819 by Funny NYPD

picon face
but when you reflow the board, will the components on the bottom side fall off at high temp?

Funny N.
Au Group Electronics, http://www.AuElectronics.com




________________________________
From: Wouter van Ooijen <RemoveMEwouterKILLspamspamvoti.nl>
To: Microcontroller discussion list - Public. <piclistSTOPspamspamspam_OUTmit.edu>
Sent: Sunday, December 7, 2008 4:33:50 PM
Subject: Re: [EE] Make Your Own PCB

Funny NYPD wrote:
> You can use some water-soluble glue for two side SMD components.

As I said, no need for that when using a paint stripper. The main
problem is stenciling the paste on the top side when the bottom side
already has components.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\07@193151 by Joseph Bento

face
flavicon
face

On Dec 7, 2008, at 3:37 PM, Funny NYPD wrote:

> but when you reflow the board, will the components on the bottom  
> side fall off at high temp?
>

I think you can localize the heat well enough with the paint stripper  
to minimize that, though falling components is certainly a possibility.

Joe

2008\12\07@210235 by apptech

face
flavicon
face
>> What is "10mil"? What unit is that?

> When discussing pcb specs the term mil typically reffers to a thousanth
> of an inch.
>
> This can be very confusing to people who take the term mil to mean
> millemeter.

It can. But using mil to mean 0.001 inch is far more established not only in
PCB uage but in industry in general than using it to mean mm. The former is
established "jargon" from time immemorial and the latter is less common and
(arguably, of course) sloppy. The former takes a longish phrase and replaces
it with a short contraction while the latter shortens analready short
phrtase slightly

viz

   mils        ... thousandths of an inch ...
   mils        ... millimeters ...



YOMV of course :-)



           Russell

2008\12\07@211711 by Kenneth Lumia

picon face

> but when you reflow the board, will the components on
> the bottom side fall off at high temp?
>

Depends on the mass of the components.  Typically,
small components such as 0805 resistors and caps stay
put due to the surface tension of the solder.  You just
have to make sure when you layout the PWB that you only
put small parts on the bottom.

You can reflow both sides of the PWB at the same time.
Solder paste is tacky enough to hold the small parts
upside down in the oven during reflow.  Just don't
jostle the board too much during insertion and removal
from the oven.


Ken

2008\12\08@013209 by Wouter van Ooijen

face picon face
Funny NYPD wrote:
> but when you reflow the board, will the components on the bottom side fall off at high temp?

Dunno what reflow is, I used a paint stripper. No, the components (1206,
0805, and some small diodes) did not fall off.

My main problem now is to get the stenciling right: the paste creeps
under the edges of the cutouts. Cleaning the stencil after each PCB
would waste a lot of paste.


--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\08@045737 by Danny Miller

flavicon
face
Well the simplest thing to do is just have only one component side.  Or
hand solder the bottom side and try to limit how many you put on there.
You can also try to do one side then lay out the bottom components and
reflow them with hot air, trying hard not to melt the bottom solder at
the same time.

There is a glue used to tack down components on the bottom side during
reflow.  You can do that.

Small components may or may not stay in place.  Larger ones I expect
have a good chance of falling off.  I think a hotplate process would
knock them off for sure.

Danny

Wouter van Ooijen wrote:
{Quote hidden}

2008\12\08@060236 by sergio masci

flavicon
face


On Sun, 7 Dec 2008, microsoftwarecontrol wrote:

> For printing, I expert pcb pattern to pdf file.
> perfect double side matching! Now it is trouble for me to make single sided
> and I need use
> tape to cover unused side.

I have no problem printing two sides that align correctly or are the
correct dimensions. The problem I have is that the mask is not dense
enough. I have tried printing to different types of media, using different
printers (and types of printers) and even photocopying onto different
media.

The densest masks I have achived are by using 60g/m^2 tracing paper with
an inkjet. I make two masks per side and overlay them. The UV exposure
time is increased from the recomended 3 minutes (with drafting film) to 4
minutes. After exposing and developing, the resulting etch mask is much
better than I have been able to produce with any other method.

Yes the tracing paper will crinkle a little where a lot of ink is applied
but the crinkling can be reduced by printing in "glossy mode" which
causes the ink to be deposited much more slowly. In any case the crinkling
does not seem to have a noticeable effect on the resulting etch mask.

Regards
Sergio Masci

2008\12\08@062204 by Alan B. Pearce

face picon face
>> What is "10mil"? What unit is that?
>>
>When discussing pcb specs the term mil typically reffers
>to a thousanth of an inch.
>
>This can be very confusing to people who take the term
>mil to mean millemeter.

As one who deals in both imperial and metric units, for exactly this very
reason, I refer to thousandths of an inch by the alternate to 'mil' i.e.
'thou', so as not to cause confusion with millimeters.

I would seriously suggest that anyone involved with a multinational list,
such as this, do similar. Yes I know it means breaking the habits of a
lifetime, but I managed to do exactly that, so anyone else should be able to
... Remember its the clarity of what you are saying that counts ...

2008\12\08@062242 by Alan B. Pearce

face picon face
> Unless you're really tight on board space or trying
> to impress your mother,
> 0805 or 1206 are much more practical

>From past experience, unless a PCB is extremely squeezed for space, I have
now resolved to use 1206 components, with a standard exception being 0603
for 100nF caps under large ICs like microprocessors and FPGAs, where you
really need to have the smaller components to get sensible short tracks on
power and ground.

This has been brought about by some sad experiences attempting to use small
components for prototyping.

2008\12\08@062413 by Mike Harrison

flavicon
face
On Mon, 8 Dec 2008 13:35:13 +0000 (GMT), you wrote:

{Quote hidden}

60gsm tracing paper is pretty flimsy, and generally won't go through a laser without crinkling - the
90GSM stuff is much better.

2008\12\08@062827 by sergio masci

flavicon
face


On Sun, 7 Dec 2008, Vitaliy wrote:

{Quote hidden}

I have only been playing with this for a few weeks (personal interest
reasons) and I have only had experience of the UV sensitive boards so I am
not in good position to answer this. However if I had to speculate I would
guess that UV sensitive boards are easier to prepare in normal room
lighting conditions.

Do you need to use a red light or dimmly light room when setting up your
Philmore-Datak blanks?

Regards
Sergio Masci

2008\12\08@063554 by sergio masci

flavicon
face


On Mon, 8 Dec 2008, Mike Harrison wrote:

{Quote hidden}

I haven't tried 60gsm through a laser and I haven't tried 90gsm tracing
paper at all. I will have to get some and experiment.

Regards
Sergio Masci

2008\12\08@064115 by Daniel Dourneau

flavicon
face
Selon "Alan B. Pearce" <spamBeGoneAlan.B.PearceSTOPspamspamEraseMEstfc.ac.uk>:

> >> What is "10mil"? What unit is that?
> >>
> >When discussing pcb specs the term mil typically reffers
> >to a thousanth of an inch.
> >
> >This can be very confusing to people who take the term
> >mil to mean millemeter.

the abbreviation for millimeter should be "mm"

2008\12\08@064236 by sergio masci

flavicon
face


On Mon, 8 Dec 2008, Alan B. Pearce wrote:

> ... Remember its the clarity of what you are saying that counts ...

Bravo!

Regards
Sergio Masci

2008\12\08@074314 by apptech

face
flavicon
face
> As one who deals in both imperial and metric units, for exactly this very
> reason, I refer to thousandths of an inch by the alternate to 'mil' i.e.
> 'thou', so as not to cause confusion with millimeters.

Wouldn't that run the risk of being a holier than thou approach? -
especially if drill sizes were being discussed?*


 Russell


* Sorry. Best I could manage at this time of night with the material
available. Feel free to improve.


2008\12\08@082053 by Alan B. Pearce

face picon face
>> As one who deals in both imperial and metric units, for exactly
>> this very reason, I refer to thousandths of an inch by the
> alternate to 'mil' i.e. 'thou', so as not to cause confusion
>> with millimeters.
>
>Wouldn't that run the risk of being a holier than thou approach? -
>especially if drill sizes were being discussed?*
>
>  Russell
>
>* Sorry. Best I could manage at this time of night with the material
>available. Feel free to improve.

<VBG> The bit I didn't say is that the 'th' is said to sound like the 'th'
in 'thought' or 'thousand', not the 'th' in 'the'. Produces a rather
different sound that doesn't make it like it is 'holier than thou' ... ;)))

2008\12\08@131340 by Danny Miller

flavicon
face
"mil" is widely used with PCB mfgs and machinists as
thousands-of-an-inch.  Dumb but... it's there.

Danny

Daniel Dourneau wrote:
{Quote hidden}

2008\12\08@132907 by Danny Miller

flavicon
face
In my experience, the inkjet printed a useless transparency.  I actually
have a photometer and put it between the photometer and the
blacklights.  I don't recall the exact number but the transmittance was
pretty high, like 15%-20%.  When I did this, I'd have to print on paper
and take it to Kinko's to have them print out transparencies for me.

The best thing I did was print on a laser printer TWICE, carefully align
the top and bottom then tape them together.  The way to do it is cut off
the corners of the top one then put a piece of tape on the top one
running over into the bottom one in the area you cut off.

The doubled-up transparency is very VERY black.

If your transparency is not black enough, you will either partially
harden the areas that need to be removed or weaken the areas which need
to stay depending on whether you have a positive or negative resist
process.  This gets worse with poorly controlled exposure times, since
the safe thing to do when you're not sure is to expose it longer.

Any sunlight coming through windows will expose a board.  Human eyes are
bad at gaging light levels due to their ability to compensate, you may
think that a room looks like a similar light level when lit by bulbs at
night but in reality it's dimmer by 10x or 100x.

A light bulb some distance away is probably not going to be a problem.  
Working under a desk lamp *may* be a problem.  Halogen lights generate
UV, though they're supposed to have a UV filter.  I wouldn't have them
around.  I'd avoid a brightly lit room even without windows.

Danny

sergio masci wrote:
{Quote hidden}

2008\12\08@135143 by Wouter van Ooijen

face picon face
> In my experience, the inkjet printed a useless transparency.  

I read somewhere that the results when using ink jet vary from useless
to pretty good. My own experience (years ago) was that original HP ink
did not work, but the cheap replacement ink was very good.

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\08@140240 by Picbits Sales

flavicon
face
> In my experience, the inkjet printed a useless transparency.  I actually
> have a photometer and put it between the photometer and the
> blacklights.  I don't recall the exact number but the transmittance was
> pretty high, like 15%-20%.  When I did this, I'd have to print on paper
> and take it to Kinko's to have them print out transparencies for me.
>

I use an old Epson printer with water based inks. I also use decent coated
transparencies and use the glossy paper setting.

The ink takes around half an hour to dry when taped to a 2kw fan heater -
any less and it sticks to the PCB.

I have never had a bad result using this method. As my old printer has been
on its last legs for the past 7 years I decided to buy a new Epson printer.
What a mistake that was. The ink is as far as I can tell solvent based,
scratches off the transparencies very easily and the light transmittance is
much greater than that of the old inkjet printer. I use el cheapo
replacement cartridges on my old Epson and have never had a problem.

I did try using my new Colour Laser printer with some proper colour laser
transparencies. It works after a fashion but has lots of pot marks in the
print. Okay for quick and nasty PCBs but nowhere near the quality I need to
making semi-production quantities and quality boards. Someone on
http://www.electro-tech-online.com has said he had similar problems and overcame
them by using dry wipe board markers. He prints off first on laser then goes
over the whole lot with the dry wipe marker. He then leaves it to dry then
gently wipes off the excess. This makes for a much darker denser print.

I've yet to try this as I keep forgetting to buy the marker pens but it
seems worth a go.

Dom

2008\12\08@150830 by Vitaliy

flavicon
face
sergio masci wrote:
> However if I had to speculate I would
> guess that UV sensitive boards are easier to prepare in normal room
> lighting conditions.
>
> Do you need to use a red light or dimmly light room when setting up your
> Philmore-Datak blanks?

The blanks are less sensitive to light, than photo paper. The manual
recommends using a yellow "bug light", but a small lamp pointed at the
opposite corner of the room did just as well for me -- provided plenty of
light to set everything up, without exposing the blank.

It sounds like having the right kind of transparency/ink combination are a
big problem for UV. For blanks that are sensitive to normal light, it is
WYSIWYG -- I printed PCB layouts on inkjet transparencies, using an Epson
printer.

The only downside that I can think of, is the longer exposure time: about 10
minutes with a 100W bulb.

Vitaliy

2008\12\08@155847 by sergio masci

flavicon
face


On Mon, 8 Dec 2008, Danny Miller wrote:

> In my experience, the inkjet printed a useless transparency.  I actually
> have a photometer and put it between the photometer and the
> blacklights.  I don't recall the exact number but the transmittance was
> pretty high, like 15%-20%.  When I did this, I'd have to print on paper
> and take it to Kinko's to have them print out transparencies for me.

I tried something similar. I printed the artwork onto plain white paper
and took that to a printer and got them to photocopy it onto both
transparency (their stock) and laserstar film (from MEGA UK). It seemed NO
better than I could produce myself on tracing paper on my trusty old
HP500C inkjet.

BTW I've also tried the jetstar film (from MEGA UK) in my inkjet. It is a
specially treated transparency. It seemed no better than the tracing paper
I tried. The one advantage that the tracing paper has however is that it
is much thinner and so when I doubled up the layers, the print is closer
together (the amount of ink was doubled for a much smaller increase in
media thickness compared to the special file).

After a lot of reading (of web material) it seems that artwork produced on
laser printers varies considerably between different models of printer.
Many modern laser printers also incoporate toner reduction logic in an
attempt to improve the life of the toner cartridge. Basically, they reduce
the amount of black printed in large black areas.

>
> The best thing I did was print on a laser printer TWICE, carefully align
> the top and bottom then tape them together.  The way to do it is cut off
> the corners of the top one then put a piece of tape on the top one
> running over into the bottom one in the area you cut off.
>
> The doubled-up transparency is very VERY black.

Yes I have also resorted to using two layers for each side of the PCB.
This does seem to produce a much better PCB.

>
> If your transparency is not black enough, you will either partially
> harden the areas that need to be removed or weaken the areas which need
> to stay depending on whether you have a positive or negative resist
> process.  This gets worse with poorly controlled exposure times, since
> the safe thing to do when you're not sure is to expose it longer.

Yes I understand this. The temprature of the developer and etchant are
also important and tend to vary between different boards I make.

I was thinking of using sodium metasilicate as the developer as other
people claim they have had good success with it and it is very forgiving
of over developing. In other words, this MIGHT compensate for the weak
black in the artwork.

{Quote hidden}

Yes you're preaching to the chior here :-)

Thanks for sharing.

Regards
Sergio Masci

2008\12\08@160247 by sergio masci

flavicon
face


On Mon, 8 Dec 2008, Wouter van Ooijen wrote:

> > In my experience, the inkjet printed a useless transparency.  
>
> I read somewhere that the results when using ink jet vary from useless
> to pretty good. My own experience (years ago) was that original HP ink
> did not work, but the cheap replacement ink was very good.

It seems that there are two basic types of ink - dye and pigment. I would
guess that the dye based ink needs a dye friendly media in order to
change it's colour, whereas the pigment based ink just sits on the
surface.

Regards
Sergio Masci

2008\12\08@172843 by sergio masci

flavicon
face


On Mon, 8 Dec 2008, Picbits Sales wrote:

{Quote hidden}

I've tried this and it was a failure. I guess the make of pens,
transparency and toner may all contribute to its success. I used "staples"
branded pens in case you're interested.

Regards
Sergio Masci

2008\12\08@174057 by Danny Miller

flavicon
face
Sharpie pen works great.  Don't go over the whole thing!  Just the defects.

Pinholes/pockmarks, DON'T use the sharpie.  Print two and do the
doubling up I described.  Sharpie on the whole thing is unnecessary,
could take a REALLY long time, and if you slip you're drawing in all new
errors.

Doubling up works because pinholes are inconsistent errors that will
basically never be present in both copies so it will definitely be covered.

Streaks due to bad transfer drums MAY be repeatable errors in both
copies unless you try some tricks like printing the second copy rotated
180 or something, but then you may or may not experience subtle
distortions that ruin your alignment.  The repeatable error will show up
in both transparencies and that's when you need a sharpie marker to
patch it up.

Danny

Picbits Sales wrote:
{Quote hidden}

2008\12\08@174248 by Danny Miller

flavicon
face
sergio masci wrote:
>
> I tried something similar. I printed the artwork onto plain white paper
> and took that to a printer and got them to photocopy it onto both
> transparency (their stock) and laserstar film (from MEGA UK). It seemed NO
> better than I could produce myself on tracing paper on my trusty old
> HP500C inkjet.
>
> BTW I've also tried the jetstar film (from MEGA UK) in my inkjet. It is a
> specially treated transparency. It seemed no better than the tracing paper
> I tried. The one advantage that the tracing paper has however is that it
> is much thinner and so when I doubled up the layers, the print is closer
> together (the amount of ink was doubled for a much smaller increase in
> media thickness compared to the special file).
>  
I wouldn't expect it to be.  It won't cause the printer to expel more
ink/toner.  At best it could promise a more even spread or something,
but really the stock transparency stuff is already just fine as far as
the film is concerned.

Inkjets do leave horizontal lines that follow the rows left by the print
head, and the thinnest point is always the problem.  When laser printers
(or regular copiers) go bad they leave vertical stripes and it can only
be fixed by replacing the expensive transfer drum, but when it's working
correctly there's no lines.  The sucky part is that when doubling up on
transparencies the inkjet row or laser printer stripe defects will
usually line up between both of them.  At one point when I was using an
old copier I tried flipping the artwork 180 deg around for the second
pass in an attempt to move the defects so they wouldn't line up.  
Unfortunately the copier also created some image distortion that became
significant when spread across the image.
> After a lot of reading (of web material) it seems that artwork produced on
> laser printers varies considerably between different models of printer.
> Many modern laser printers also incoporate toner reduction logic in an
> attempt to improve the life of the toner cartridge. Basically, they reduce
> the amount of black printed in large black areas.
>  
Well the toners do vary.  You want to go into the Printer Config and set
toner to Max.  Toner Saver off.  Whatever option is there, make it dark.
This should tell the printer to bypass the saver.

> Yes I understand this. The temprature of the developer and etchant are
> also important and tend to vary between different boards I make.
>
> I was thinking of using sodium metasilicate as the developer as other
> people claim they have had good success with it and it is very forgiving
> of over developing. In other words, this MIGHT compensate for the weak
> black in the artwork.
>  
Also flatness is an issue.  With thin traces, if the toner layer is
offset above the resist then light will spread in around the toner and
half-expose some of the border.
So, it's best to print so that the toner side will go DOWN against the
resist.
And basically you're never going to get perfect results without a vacuum
table.  Thicker glass does help, the thinner glass panes do bend when
clamped at the edges.



2008\12\08@175419 by sergio masci

flavicon
face


On Mon, 8 Dec 2008, Vitaliy wrote:

{Quote hidden}

I think the transparency/ink combination problem exists whether it's UV or
visible light. The problem is getting a dense black to block the light (UV
or visible).

This is why I was toying with the idea of trying reflected light from
paper via a camera obscura (basically a pin hole camera with the PCB as
the film)

>
> The only downside that I can think of, is the longer exposure time: about 10
> minutes with a 100W bulb.

I guess longer exposure time gives a greater margin for error but the same
is true of short exposure time and a more accurate timer :-)

I wonder if Philmore-Datak blanks are sensitive to red or green laser
pointers :-)

Regards
Sergio Masci

2008\12\08@195941 by Danny Miller

flavicon
face
That's getting pretty far into unnecessary.
2 transparencies overlaid for redundancy and you can do very fine lines
reliably as long as the glass sandwich is tight.

There's a lot to be said for toner transfer with magazine paper too.  I
did that with a heat gun- no laminator- and did some very fine lines,
although looking at the 7mil stuff under a microscope it was kinda
spotty.  Had some broken traces.

Danny

sergio masci wrote:
{Quote hidden}

2008\12\09@043348 by Vitaliy

flavicon
face
sergio masci wrote:
> I think the transparency/ink combination problem exists whether it's UV or
> visible light. The problem is getting a dense black to block the light (UV
> or visible).

I doubt the problem is as severe with normal light sensitive boards. I got
perfect results on the first try, and every time after that (also when I was
helping my cousin -- he had an HP printer, IIRC).

IANAE, but isn't it possible that an object that is opaque to visible light,
could be (semi-)transparent to UV?


> I guess longer exposure time gives a greater margin for error but the same
> is true of short exposure time and a more accurate timer :-)

I am wondering if maybe you guys are over-exposing your boards? In this
sense, the Philmore-Datak boards are very forgiving: an extra minute or two
under the bulb has no noticeable effect.


> I wonder if Philmore-Datak blanks are sensitive to red or green laser
> pointers :-)

No idea. :) The photo emulsion layer is green.


2008\12\09@045741 by Wouter van Ooijen

face picon face
> IANAE, but isn't it possible that an object that is opaque to visible light,
> could be (semi-)transparent to UV?

I don't see why not. The IR-passing housing of IR receivers is very dark
blue, almost black, to my eyes. And the black foam I use for packing ICs
reflects IR perfectly. (That's all IR, but why shouldn't that be the
case for UV).

--

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2008\12\09@072759 by sergio masci

flavicon
face


On Mon, 8 Dec 2008, Danny Miller wrote:

{Quote hidden}

I agree it wouldn't lay down more ink BUT the ratio of ink to media goes
up because the media is thinner. i.e. the media becomes more transparent
while the ink blocks the same as it did for the thicker media. Result:
shorter exposure time to break down non sheilded photoresist and hence
less light for the ink to block.

>  At best it could promise a more even spread or something,
> but really the stock transparency stuff is already just fine as far as
> the film is concerned.
>
> Inkjets do leave horizontal lines that follow the rows left by the print
> head, and the thinnest point is always the problem.

I am sure you are seeing the effect you describe with the inkjets you have
used but I have not. I have uploaded an image I scanned of an artwork I
produced on 60gsm tracing paper so that you can see for yourself. The scan
is in 600 dpi greyscale. You can see this image at
http://www.xcprod.com/scanned_circuit_artwork.jpg

{Quote hidden}

I've tried doing this but it makes no difference.

{Quote hidden}

Yes I understand this. Using two overlaid artworks for each side is also
going to produce a a less distinct edge because there will be a gap
between the two artworks.

> And basically you're never going to get perfect results without a vacuum
> table.  Thicker glass does help, the thinner glass panes do bend when
> clamped at the edges.

So using a system which does not involve tracing paper, transparencies or
drafting film (i.e. projecting the focused image of the artwork directly
onto the sesitised PCB) will give the best results.

Regards
Sergio Masci

2008\12\09@142833 by Danny Miller

flavicon
face
Sure there is.  It's called Wood's Glass:
http://en.wikipedia.org/wiki/Wood%27s_glass

That's the purple glass used on blacklight tubes.

Danny

Wouter van Ooijen wrote:
{Quote hidden}

More... (looser matching)
- Last day of these posts
- In 2008 , 2009 only
- Today
- New search...