ODP: AW: PCB making tips

R.G. Keen keen at austin.ibm.com
Wed Sep 2 16:36:18 CEST 1998

At 05:29 PM 1/09/98 +0200, Roman Sowa  wrote:
>>About laser printed PCB-film: i haven't tried this but few of my friends
>>For one it worked well (he printed on thin paper - not film), as for
>>I don't know if it depends on the printer, but they failed.
>I've seen a huge amt of correspondence on this...
>The ironing stage, and how clean the board is, are the crucial variables.
>NOBODY has ever claimed to get it right 'first time'...

Actually, many people do. I provide laser-printed PNP blue film patterns
for re-creating classic guitar effects to quite a number of people. I include
a few pages of instructions about how I do it. I regularly get email that
indicates that the first try produced a working board. I include two images
in case the first try is a disaster, as I thought most people would be 
first-timers with this. This has turned out to be unnecessary in most cases, 
as a significant fraction get two working boards this way. 

Over hundreds of toner images mailed out, only one person has emailed back that
they flubbed both images. 

Of course, that is the PNP blue material, not thin paper, clear mylar film, 
address label paper, or other stuff. I've used all of the various versions of
PCB film, special papers, clear mylar transparencies, TEC-2000, etc, and 
PNP Blue is by far the easiest and most forgiving. I highly recommend it to 
anyone who makes PCB's. The Dyna-Arts water soluble transfer paper is ideal
for making labeling decals, however. That's how I letter all my boxes.

There are some limitations of the toner transfer method:
1.There is some thermal expansion of the medium. A skilled worker can keep the
effects of this to a minimum, but it does limit the size of the precision 
alignment features to about 2" - that is, if you have a connector or IC socket
more than 2" long, the accumulated tolerance on thermal expansion of the film
will make fitting the real chip or socket to the pattern hard unless you 
precompensate for this by printing the pattern a few percent too small to start
with. This is easy with Easytrax. I don't know about other programs. Likewise, 
it's much easier if your boards are less than 6" by 6". I have done 8" boards,
but it's hard. The boards beginners do well on are about 3" by 3".

2.The melted toner expands to be slightly larger than the printed size. That is,
an error in ironing, makes the pads and traces too big. The commonest fault is
shorts where traces are closer together and the ironing spread it to touch. I 
deliberately make my beginner boards with wider margins between pads and traces 
to allow the first timer some slop. Pressing down on the iron exacrebates this, 
by the way, and makes the heating time more critical. You can use a sharp pointed
knife or pin to scratch away shorts after ironing very easily if you like. 

3. Opens result from insufficient heating on PNP blue, although this is by 
far less a problem than with other transfer media. You can easily touch up minor
opens with a "Sharpie" marker if you like. Large areas of non-transfer is an ironing
problem. I have experimented with the heated roller assembly from an IBM laser
page printer that I found in a junk store. I believe that this assembly could be 
used to make a perfect-first-time heater for PNP blue. Likewise, there are probably 
adjustments to office laminators that would make a perfect automated system. I 
have not messed with this because ironing is so easy. With other toner transfer 
systems, a modified laminator is probably perfect, as ironing is much more critical
with these mediums.

4. For beginners, I deliberately use lower resolutions: no lines between IC pads, 
0.030" traces on 0.075" centers, 0.070" minimum pads. For myself, I use 0.012" to 
0.015" traces on 0.025" to 0.050" centers, one line between pads on IC's, 0.050" 
or larger pads. The pad size is more a function of my failing middle-aged 
eyesight and hands than the ability to make the etched board. In practice, I get 
good enough layouts that the component packing more than makes up for any larger 

5. Double sided boards are hard: however, this is the same problem that all DIY
double sided techniques have, and is not unique.

First-time working results are composed of knowlege about the medium (which I provide
in the instructions) and a layout that is not too aggressive considering the probable
skill level of the user (which I provide in the layout). A first timer will not get 
0.012" traces right, and if they are expecting to simply take a layout that they 
have copped somewhere that uses thin traces and close spacing, along with 
not-educated ironing technique, they will not get the first one to work. 

Which does raise the question - what good is toner transfer if it can't do the
resolution that photo techniques can, doesn't have the accuracy, has built in 
problems with process control, etc?

My answer is - doggone, it's fast and easy if you recognize what it CAN do. Most
homemade synth PCB's don't really need to be as dense and fine as commercial boards.

More information about the Synth-diy mailing list