Homebrew PCBs

Hi

What's the method to find out the track width (in
mils) that a laser printer can print reliably from
it's resolution in dpi ?

Stefan told me that a 600dpi can di upto 2mil ?

I forgot to ask him how he did that ?

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

--- In Homebrew_PCBs@yahoogroups.com, Kamal Shankar 
<kbshankar2000@...> wrote:
>
> Hi
> 
> What's the method to find out the track width (in
> mils) that a laser printer can print reliably from
> it's resolution in dpi ?
> 
> Stefan told me that a 600dpi can di upto 2mil ?
> 
> I forgot to ask him how he did that ?
> 
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around 
> http://mail.yahoo.com
>


My guess, ( the way I  calculate it ) would be 
 
1 inch  divided by the DPI (Dots Per Inch),  then round off

i.e.   1 
      / 600
    -----------
    0.0016666666666666666666666666666667
rounded to 0.002

contrary to what is implied in the term "mil",
1 mil = 0.001 inch

This might not be the way Stefan works his 
calculations

> -----Message d'origine-----
> De : Homebrew_PCBs@yahoogroups.com 
> [mailto:Homebrew_PCBs@yahoogroups.com] De la part de scratch_6057
> Envoyé : juin 22 2006 07:16
> À : Homebrew_PCBs@yahoogroups.com
> Objet : [Homebrew_PCBs] Re: How to calculate effective mil 
> values from printer dpi
> 
> 
...
> 
> contrary to what is implied in the term "mil",
> 1 mil = 0.001 inch
...


Actually, the term "mil" _does_ make sense.  Americans assume it's supposed
to make sense in English;  "mille" means thousand in French.  But I wouldn't
be surprised if machinists used that term for some other reason too.

Robert
:)

"Mil" isn't used by machinists (here in the U.S.).  Our basis is the 
thousandth of an inch .001 and of course the Inch unit.

So, our references follow those two terms.  For example, the distance 
of 12 13/16"

A machinist might say that as 12 inches and 812 and a half thousandths.

12 13/16 = 12.81235

The 12 inches is obvious
the 812 refers to the number of thousandths
the 1/2 refers to 1/2 of a thousandth = .0005

If you hear a machinist say something like "It 3 tenths off".  He 
isn't referrening to tenths of an inch, it's tenths of a thousandth 
= .0001"

Very few machine shops have the real ability to work to any greater 
precision than a tenth of a thousandth .0001".  However, there are 
rare cases when you get into the millionths of an inch.  Tolerances in 
this range are usually found in electronics and optics.  I personally 
have only had exposure to millionths of an inch when working with a 
company that produces optics for spy cameras (satellites).  But again, 
we would say a "Mil" in that reference, it would be said as for 
exampe: Twenty millionths of an inch

The crazy terminology of each industry, it keeps everyone confused ;-)

Chris




> > 
> > contrary to what is implied in the term "mil",
> > 1 mil = 0.001 inch
> ...
> 
> 
> Actually, the term "mil" _does_ make sense.  Americans assume it's 
supposed
> to make sense in English;  "mille" means thousand in French.  But I 
wouldn't

> be surprised if machinists used that term for some other reason too.
> 
> Robert
> :)
>

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

From: "lcdpublishing" <lcdpublishing@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Thursday, June 22, 2006 2:21 PM
Subject: RE : [Homebrew_PCBs] Re: How to calculate effective mil values from 
printer dpi


> "Mil" isn't used by machinists (here in the U.S.).  Our basis is the
> thousandth of an inch .001 and of course the Inch unit.

We call 0.001" a 'thou' in the UK. I use mils for PCBs, though.

Leon

Yup, same here too - just didn't want to inject too much slang into 
the subject ;-)

Chris

> We call 0.001" a 'thou' in the UK. I use mils for PCBs, though.
> 
> Leon
>

>Stefan told me that a 600dpi can di upto 2mil ?

Probably useful if all you're interested in printing is isolated 
straight lines. Typically, the toner "spill" on paper for even a 1200 
DPI laser extends beyond .002". For a HP printer using micro-fine toner 
it can be more than .005" on paper. I brought this up with HP once and 
their response was that their print quality does not address anything 
that cannot be seen with the naked eye.

--- In Homebrew_PCBs@yahoogroups.com, Kamal Shankar <kbshankar2000@...> 
wrote:
>

On Thu, 22 Jun 2006 03:45:27 +0200, Kamal Shankar  
<kbshankar2000@...m> wrote:

> Hi
> What's the method to find out the track width (in
> mils) that a laser printer can print reliably from
> it's resolution in dpi ?
> Stefan told me that a 600dpi can di upto 2mil ?
> I forgot to ask him how he did that ?


What it can print reliably might be quite different from how wide a dot  
is, theoretically.
If you make a layout you'll want to control how many dots the printer lays  
down beside each other, so you'll want to enter the calculated width.  
However this is probably not the final trace width, since it will spread  
out a little, underetch a little, etc...

You will not get a reliable trace at 1.66mil. I didn't get a reliable one  
at 3.33mil either, using toner transfer. 6.66 seems to work well however.  
This is a test i made some time ago:
<http://trethan.at.tf/pub/img1/PCB2.JPG>


I think the limit is underetching here, not the printer. If i remember  
right the 3.33mil line looked fine before the etch.

ST

>6.66 seems to work well however

What was the intended spacing between the two 6.66 lines in the artwork?

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

On Thu, 22 Jun 2006 21:41:57 +0200, derekhawkins <eldata@...> wrote:

> What was the intended spacing between the two 6.66 lines in the artwork?


On the right you see single lines of the width specified in the middle.
On the left you see two lines with the gap specified in the middle.

I hope this answers your question.

ST

> I think the limit is underetching here, not the
> printer. If i remember  
> right the 3.33mil line looked fine before the etch.

Stefan - my "limit" currently is my procedure for
removing the paper from the copper clad after doing
the transfer.

The method I use serviced me well till now (when I
started to do < 10mil), as currently all thin traces
(<10mil) just feel bad about themselves and say bye
bye alongwith the soggy paper.

So right now, I have to use a Sharpie to correct those
tracks :(

Hmmm... maybe I should learn the correct way to remove
the paper...

Currently, after doing the transfer I immerse the clad
in cold water and then lukewarm water for 5 mins (till
the paper becomes soggy)

Then I gently ruboff the paper in running water, and
right now all thin traces (and some parts of ground
plane) are saying bye bye ....

What method do you use ? Will soaking it overnight
work ?

I heard that soaking it overnight may even lead to the
paper falling off by itself ?

The funny thing is, the toner which sticks to the clad
takes hell lot of effort to scrub off (I have since
resorted to using acetone - Thanks mycroft!)

But the 10mil traces and random parts of ground plane
seem to be not so stubborn :(


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

> contrary to what is implied in the term "mil",
> 1 mil = 0.001 inch

The term has it's root in the Latin word - m\ufffdll\ufffdsimus,
or 1000 as we all know.

I guess, as standard for all milling machines are in
inches (anyone remember the Harrison lathes ?) -
British makes were highly respected for their quality
;)


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

--- Kamal Shankar <kbshankar2000@...> wrote:

> Hmmm... maybe I should learn the correct way to remove
> the paper...

or use another paper
 
> Currently, after doing the transfer I immerse the clad
> in cold water and then lukewarm water for 5 mins (till
> the paper becomes soggy)

throw the hot board into a cold (warm) water bowl

> Then I gently ruboff the paper in running water, and
> right now all thin traces (and some parts of ground
> plane) are saying bye bye ....

paper should not be 'rubbed' off, it must 'peel' gently or even float off by
itself. may take more than 5 minutes...
if you have to rub --> wrong paper!! 
it's ok to brush out some residues between tight tracks and holes.

> But the 10mil traces and random parts of ground plane
> seem to be not so stubborn :(

are you using an iron? hot enough?
beware of the steam holes, they play dirty tricks!
do you use a paper towel 'buffer' between the iron and the TT paper back?, up
to 4 ply can be very useful.



__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

On Fri, 23 Jun 2006 08:20:50 +0200, Herbert E. Plett <cachureos@...>  
wrote:

>
> paper should not be 'rubbed' off, it must 'peel' gently or even float  
> off by
> itself. may take more than 5 minutes...
> if you have to rub --> wrong paper!!
> it's ok to brush out some residues between tight tracks and holes.


I don't agree, i rub quite a lot with some papers and it isn't a problem.
It's more work, yes, but it does not harm the toner. If it can't survive  
rubbing then it is not adhered properly.

Use a sponge for rubbing instead of your fingers if you make many boards  
at once.

ST

On Fri, 23 Jun 2006 02:56:28 +0200, Kamal Shankar  
<kbshankar2000@...m> wrote:

>
> Then I gently ruboff the paper in running water, and
> right now all thin traces (and some parts of ground
> plane) are saying bye bye ....
> What method do you use ? Will soaking it overnight
> work ?

I don't soak it long, no. I have never tried so i dunno if a paper will  
come off overnight.
Your method sounds fine to me, i believe the toner does not adhere well  
enough.

What board preparation/cleaning do you use? What method to iron/fuse?

> The funny thing is, the toner which sticks to the clad
> takes hell lot of effort to scrub off (I have since
> resorted to using acetone - Thanks mycroft!)

You can use a hard edge, like a steel scraper, and scrape it off. If you  
don't scratch with the corner it will not damage the traces.
I then use acetone to clean the board and remove any tiny pieces left, but  
the scraping makes quick work of the bulk of it.

ST

> do you use a paper towel 'buffer' between the iron
> and the TT paper back?, up
> to 4 ply can be very useful.

Hmm... is using a "buffer" between the iron (set at
cotton) and TT paper back that important ?

Tilll today, I just apply the iron directly onto the
back of the 65gsm glossy paper.

Will using a "buffer" actually improve anything ?

Also, will applying too much pressure on the paper
actually destroy the TT ?

I have noticed this leads it tracks falling off as
easily as if proper heat was not applied.


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

--- Kamal Shankar <kbshankar2000@...> wrote:

> > do you use a paper towel 'buffer' between the iron
> > and the TT paper back?, up
> > to 4 ply can be very useful.
> 
> Hmm... is using a "buffer" between the iron (set at
> cotton) and TT paper back that important ?
> 
> Tilll today, I just apply the iron directly onto the
> back of the 65gsm glossy paper.
> 
> Will using a "buffer" actually improve anything ?

surely it does, just try yourself when, how and how much. Direct ironing (with
the edge or tip) is sometimes necessary to insure proper application, mostly on
edges or large patches. I, myself, go over all the board with the tip, just in
case...

> 
> Also, will applying too much pressure on the paper
> actually destroy the TT ?

not I am aware of. If you use the buffer you don't risk to squeeze the traces
at all, then be prudent with retouching the critical areas with the tip of the
iron.

 
> I have noticed this leads it tracks falling off as
> easily as if proper heat was not applied.

I see two reasons:
- iron is not hot enough. should be, at least, some 235C (460F). buffer should
yellow in less than 5 minutes.
- improper board prep. scrub the board with that (kitchen) 'abrasive' cream
until towel (not fingers) gets black, top down and then left right. Rinse under
running water and wipe dry with towel. NO FINGERS, please. Board will be shiny
and only very slightly scratched, any dull part is wrong. Sandpapering or deep
scratches are not good, avoid them.

If you have to re-do a TT then remove the failed print with acetone (that
smelly nail polish remover) and RE-SCRUB the board. The heat, water and acetone
residue will stain the copper enough to make it not useable as is.



__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com