Homebrew PCBs

Looking for the cheapest, quickest and most environmentally friendly 
route to good PCB results (who isn't?), my research within this forum
and other web resources convinced me to use pre-cut, surplus PCB
material, glossy magazine covers as the toner transfer material and a
homebrew, replenishable HCL/H2O2 etchant.  Now, all I had to do was
get good, reproducible results using these materials.  In the past, I
have always experienced inconsistent results with toner transfer (TT)
PCBs.  In an effort to change that, I did some experimenting to
determine what I was doing wrong.  I don't doubt that this line of
investigation has been followed and documented before.  I had fun
doing it, anyway.

First, I needed to determine exactly why I wasn't getting good results
with my previous TT attempts using an iron.  I felt that understanding
that problem might lead me to some fundamental information that might
be useful elsewhere.  My past problems always involved imperfect
adhesion of the toner to the PCB.  So, suspecting a toner fusion
problem possibly related to low temperature, my first questions were,
"At what temperature does toner melt and how hot are my irons?"

I found a few web references to toner fusing/melting at 200C (392F). 
Since I own a Raytek MiniTemp for use in another hobby (RC cars and
planes), I could easily measure the surface temperature of my irons to
determine if they actually reached that temperature.  I was intrigued
by the results, all taken with the irons set at their highest
temperature settings:

1. Small travel iron (the one I'd used for all of my previous TT
efforts): 350F at its hottest spot (very small area), 210F over most
of its surface

2. Large household iron: 205F at its hottest spot (very small area),
175F over most of its surface

It appears that toner must only be tacky/plastic and not fully melted
under an iron at these temperatures.  Of course, when using a process
like ironing which involves rubbing and potentially high levels of
localized pressure, a fully melted toner might smear or spread beyond
its intended areas of adhesion.  However, toner that was only tacky
would explain the great importance of iron techniques involving
pressure and movement and the resulting variability in results if
those techniques are not adhered to consistently.

So, I concluded that I needed to somehow bring the PCB and transfer
medium close to or at the toner fusion temperature and apply uniform
pressure across the width of the board to reliably transfer the toner
while avoiding smearing or spreading.

The obvious solution would be to use a laminator, but to buy one that
was already known to work reliably with PCB toner transfers would
possibly violate my goal of a "cheap" route to reliable PCB
production.   I found the forum database of results using various
laminators for TT and noticed the Royal PL2000 (~$90 on eBay, $149
elsewhere and, apparently, discontinued by Royal).  I wondered how
well the newer Royal model PL2100 would do, since it was for sale at
Royal.com for $29.95.  So, I took a chance and ordered it, even though
Royal's claim of "Easy to use - no temperature controls to adjust"
made me suspicious. The forum database description of the Royal PL2000
mentioned a temperature control set to maximum.  Since the PL2100
would most likely be permanently preset to a median temperature, would
it be one of those inexpensive laminators that don't supply enough
heat to adhere the toner to the PCB?
  
I received the laminator before I received the .021" double-sided PCB
I planned to put through it.  So, I tested the laminator with one of
its included laminating bags to make certain that it wasn't DOA...
before I disassembled it.  Upon disassembly, I found a synchronous
motor to turn the two rollers, a half-wave rectifier circuit to supply
the two 60V, 110W, 9.75" wide flexible ceramic heating elements used
to heat the surface of one of the rollers and a Daewood (not Daewoo �
it's Chinese, not Korean) KSD602 auto-reset thermostat (with an
unknown temperature trip point).

Using my Raytek MiniTemp, I measured the temperature of the heating
elements at 390F (there's that temperature again - perhaps they're
using toner fusers intended for laser printers?) and the adjacent
metal structures at 150F.  Because of the way the heating elements are
arranged, the surface of one of the rollers is exposed to 390F radiant
heat over about 180 degrees of its rotation.  This made the odds of
actual toner melting and fusion look promising.  Just in case, I used
Teflon insulated wire to solder a miniature, momentary toggle switch
rated at 5A @ 250V across the thermostat (which was only activating
the heating elements on an approximately 30% duty cycle) and drilled a
hole for it in an open area of the case.  That way, when feeding the
board through the laminator, I'd make sure the heaters were on shortly
before and during the pass of the PCB through the laminator.  I DO NOT
recommend that anyone else do this since you could potentially damage
your laminator, your house and yourself if you aren't careful.  At
that point, the least of your problems will be that you also voided
your laminators warrantee.  Besides, I came up with a better backup
method that requires no modifications to the laminator and which is
probably applicable to anyone with any model of "wimpy" laminator. 

After receiving the .021" PCB material, I tested the Royal PL2100
laminator using Tom Gootee's proto26.bmp artwork as a test pattern
printed on glossy magazine-cover paper.  I made twelve passes of the
board through the laminator on the first test to give it every chance
at success.  I DID NOT use my thermostat bypass switch. 
Unfortunately, while the toner pattern separated easily from most but
not all areas of the paper after a soaking in warm water, about 50% of
the toner was not well adhered to the PCB and could be removed with
little effort.

SO, it was time to implement my "wimpy laminator" backup plan.  The
laminator was already in the kitchen on a counter next to the oven in
anticipation of this and the oven was already pre-heated to 390F. 
Since I knew from my previous measurement of the laminator heating
elements that the rollers of the laminator could be safely exposed to
390F temperatures, I knew that running a PCB _already_ at that
temperature thru the laminator would not harm its rollers in any way.
 Here's what I did at that point, step-by-step.  This sounds complex,
but it was really very easy and quick and could easily be scaled up to
process a large number of PCBs at the same time.  It's also a probable
fix for others with low-heat laminators:

1.  I ran the board with its glossy-paper/artwork through the
laminator twice to fully flatten the paper and tack it to the board
(perhaps only one pass would have been enough � I'll determine that
later).

2.  I placed the board in the 390F pre-heated oven and checked it with
the Raytek MiniTemp to determine when it reached the oven's ambient
temperature � it took less than two minutes to do so � BTW, to make it
easy to pick up, I put the board on a broiling pan so that one side of
the board stuck out beyond the pan � also, you don't need a Raytek
MiniTemp to do this right; just put your board in the oven for a few
minutes to reach the oven's ambient temperature.
 
3.  With a wool glove on one hand, I removed the board from the oven
and ran it through the laminator twice (once again, maybe once would
have been enough � I'll determine that later).

4.  I put the PCB on the counter and let it cool to "warm but
completely comfortably to handle with a bare hand" (forgot to measure
that with the Raytek MiniTemp) � that only took a minute or so.

5.  I put the board in a pan of warm water and let it soak.

6.  Only a few minutes later, one corner of the paper was already
floating free � I pulled the paper back diagonally until I met the
slightest resistance, waited for about 10 seconds, pulled again and
repeated the process � in less than two minutes, the paper was off and
ALL of the toner was bonded to the PCB.

7.  Just for good measure, I rinsed off the board under a stream of
warm tap water and then put it back into the oven for a few minutes to
let it heat to 390F again.  That may have been completely unnecessary.
 I will determine that later with additional tests.

8.  I let the PCB cool to room temperature, immersed it in warm water
and used a Quickie plastic mesh dish scrubber (from Wal-Mart � see
photos at link below to see what it looks like) to remove the tiny
amount of lingering paper residue.  This took less than a minute to
complete with very gentle scrubbing (not that the toner couldn't have
withstood even extremely hard scrubbing, it's just that only very
light scrubbing was needed).

What I ended up with was the best toner transfer I've ever managed, by
far.  It's absolutely perfect.

Here are a few photos including a close-up of the toner pattern on the
board (the board's orange color in a few of the photos is an artifact
of the lighting used):

http://ph.groups.yahoo.com/group/Homebrew_PCBs/photos/browse/843d

If anyone has artwork (max 6"x6", the size of my test board) that
would provide a more challenging test for this method, please email it
to me, post it in the photos section or point me to it. 

Thanks,
Bill