A subject dear to my heart!

I have spent the last six months experimenting with direct laser
printing of PCBs, and I am now able to (fairly) reliably produce
boards using this method. If anyone is interested I can share a LOT
of details, but the Brother is NOT the way to go. I spent a month or
so working with the HL-2170, as well as a few other false starts
before settling on the Lexmark E260 as the printer of choice. The
"straightness" of the path is relative, and all of the laser printers
have to be modified to pass pcbs, although you can run the thin,
somewhat flexible (eg 8 mil) boards through some of them. You cannot
use the internal fuser.

When I described some of my earlier efforts at direct laser pcb,
there was some concern that the drum is too easily damaged. In the
past months I have made several hundred "runs" through my modified
printer, many of which were experimental, without proper smoothing
and protecting of edges, and I am still using the same drum. I am
still not sure of the limits of the process, but I have produced 5
mil traces without problems.

The steps are:
1 - Produce the artwork - I use Eagle.
2 - Print the pcb.
3 - Place board in oven to "fuse".
4 - Etch.

One real advantage of direct laser pcb is that you can see the
results instantly, and if the "image" is not perfect, you can just
brush the toner off, wipe the board with acetone, and do another run!

I was waiting till I had the process perfected before "publishing" on
this forum, but if anyone is interested, I can provide details. You will need:

1 - A Laser Printer - I strongly recommend the Lexmark E260, on sale
at Staples for $169 (?) with an additional $50 off if you bring in
any dead printer for "recycling".
2 - A high voltage power supply to provide a variable Transfer
Voltage. A 3KV (variable) at 100 microamps will do - Ebay is your friend.
3 - A high voltage relay to switch between the internal transfer
voltage and the external supply is useful but not strictly necessary.
4 - A simple microprocessor board to spoof the sensors that need to
be emulated and control the printer solenoids and such.

Mark

I don't want a flame war here, but I've got a board in my hand,
direct from my modified laser printer, that says you are wrong. When
the boards come out of my laser printer the toner is quite solidly
attached to the copper. You can turn the board over and tap it or
blow on it strongly without disturbing the toner.

Interestingly, double clad boards are easier to do than single sided
ones. It seems to require a much higher voltage to charge a board
that does not have copper touching the transfer roller.

Mark

I'll answer some of these questions now, but followup with pictures
and a detailed description when I have time.

In measuring the voltage from the internal power supply (IPS) to the
transfer roller (TR), I found that the voltage starts at about -400
vdc when the printer prepares to print, then changes to +1400 vdc
(approx) when printing, then reverts back to the negative voltage
after the image is printed. If you just apply a positive external
voltage to the TR, it works, but the drum never really gets cleaned
off between runs. You can connect the IPS, run a dummy print (a
period on a blank sheet) to clean the drum, then disconnect the IPS
and attach the external power supply (XPS), and it works ok. I found
(ebay) a hv relay that I programmed to use the IPS till the actual
printing is about to start, then switches to the XPS for printing,
then back to the IPS to clean up the drum.

The IPS is attached to the TR via a spring under the left (as you
face it) TR bushing. I need to do more experimentation as to what's
optimal, but for now I removed the TR, cut away the rubber roller on
each end, leaving about 4 inches of rubber in the middle. This allows
a 4" wide board to be printed, which is fine for me, but I'm sure it
can be much wider. I then soldered some high voltage wire to 2
bushings and put them on the bare steel rod on both sides to the
rubber roller. This is where I charge the board.

I should mention that the voltage required seems to change depending
on things like humidity and temperature, but once you get familiar
with the system you can zero in on the correct voltage in a few quick
runs. Since a run takes only a few seconds, this isn't much of a
problem. One of the things I want to do is measure the humidity and
temperature and see if there is a consistent relationship.

I am currently using a garolite (FR4 is garolite??) carrier 6 inches
wide with a 5 inch wide opening to carry the board. The board is
suspended by masking tape in the cutout so that when the carrier runs
through the TR, only the board hits the TR.

To feed the board I removed a lot of the stuff at the front of the
printer and added a spring loaded roller so that the carrier is
pushed through the printing area. This is MUCH easier than you might
think because there is a slot on each side of the printer that is
perfect for the purpose. I am reusing the centering mechanism, so the
board does feed right up the middle. Again, I'm not sure what is
optimum but I can run .062 board through , though I usually use .032
board. The .062 is too thick for the gear on the TR to mesh with the
drum gear, but it doesn't seem to matter because the board is being
pushed through via the front rollers and there is enough friction via
the drum to push the carrier out the back. I have experimented with
pulling the board through using the drum/TR vs pushing it through
from the rear, but have reached no firm conclusion as to which is better.

Right now I am timing the whole thing from when the solenoid pulls
in, but a better way is to use an optical sensor to start the timing
when the board reaches a fixed position. In theory you can use the
centering mechanism to set the X axis and the sensor to set the Y
axis, enabling registration for double sided boards.

Mark

Craig,

That's a good question. I'm having trouble with the fusing process.
When I heat to say 350 F, the toner comes off in some areas when I
etch. If I really heat the board to say 450F, the toner melts and
spreads some, but really adheres well. I bought a stage micrometer
and have been trying to measure trace widths, but since I haven't
stabilized my process, I'm not sure exactly what I have, though I do
think there is some spread. I would be interested in any results from
the sprinkled toner on ink people - what temperature do you heat the
board at, and for how long? Is there any spread?

The toner transfer people tell me there is no spread with tt. Perhaps
I should heat the board at a lower temperature to lightly fuse the
toner, then run it through a laminator??

I'm also still experimenting with etching, using peroxide/HCl, but
etching has been uneven with the areas around traces seeming to etch
way before the rest of the board.

All-in-all, there are a LOT of variables, but I have been making some progress.

Mark

I just posted a board with 3 mil traces. Using a laminator after
"setting" the toner in the oven make a BIG difference!

Mark