Digital BW, The Print

I thought some of you might be interested in this post by Elena on the 
EpsonWideFormat list (copied with her permission).

There's more on that list.

Bob Frost.



"Dear all,
I would share this experience with you. It definitely looks like
compatible inks really can cause some level of degradation to the
printer head precision.

This experience concerns MIS inks, but there's no reason to think
it shouldn't affect other non-Epson inks in general!
This experience also concerns a 4800 but, again, there's no reason
to think it should affect it only, indeed!

The story:

My just bought, second hand (but like brand new) 4800 was still using
the ink remaining in its original epson carts. All nozzles was firing
perfectly straight. Btw, it's hard you can notice deflected nozzles by
mean of the standard nozzle test pattern offered by the printer panel or
driver: it just puts in evidence that a nozzle is firing or not, roughly.
Only a special, straight nozzle test (I can do that since I'm a
developer, or just try the ones available from the service mode) can
show how effectly good and straight the nozzles are firing.

Now after months of work, I finally accomplished the goal I had posed:
to totally unlock it, disabling ink monitor and mtnk monitor (and I
thus could notice that effectively there's still about half the ink
remaining in the carts when the printer judges it's out ;)

So I finally installed my CIS, which I filled with InkSupply's MIS K4 inks.
I usually go with them since I was always good with their inks, and
they have a quite good reputation and reasonable prices.
Once the printer was properly purged and primed with the new inks,
the first nozzle check started showing an high percentage of deflected
nozzles! At first I thought of some air, partial blockage or so on,
it's normal enough when doing this kind of jobs. But after some cleaning
cycles I noticed absolutely no changes to the results. Again, remember
that a normal nozzle test would never show the problem so well as a
"straight" nozzle test does:

 http://www.elenadomain.it/pub/NozzleCheckBad.png

[Note that deflection (horizontal darker or lighter lines) is most
 noticeable in darker inks but under scanner inspection it affects pretty
 all the channels - LLK is not present in the test because I actually use
 it for other purposes ]

Even if deflection effectively does cause some level of degradation in the
print quality, the plain users _might_ not perceive it so much because of
the stochastic screens used by the standard epson driver. But just use a
regular screen, or a microweaving mode previledging speed vs. quality,
and you will notice it well. And it is, definitely, not a good thing.

But I, as a crazy tech, hacker and programmer, planned and found a solution
to fix the problem!
I know that (at least) Epson heads are calibrated at factory level against
the little and unavoidable manufacturing defects. Calibration is obviously
made using the Epson inks intended for that head, and it comes out in the
form of a so called Head Rank ID - without going deep in details, those
well known codes you have to enter when you replace a printer head.
All stuff undocumented and kept well secret as usual - but the concepts
behind the code at least are documented in some patents assigned to Epson.

Compatible ink manufacturers try to make their liquids as compatible as
they can to the originals as for rheology (and visual aspects, of course)
but they will never be AS the originals. So, it's unavoidable that a
head calibrated with a particular ink won't perform as good with a
liquid having different characteristics. The proof that my problem was
not definitely caused by an instant nozzle fault or clog is simple:
I purged a line and filled it back with the previous Epson ink: all
deflections in that channel disappeared! Note well, this experience
does affect MIS inks and the 4800, but I really have no reason to
think it shouldn't be a quite general rule! May be that inks from
another manufacturer, perhaps more expensive, are less prone to this issue.

Then, the solution was obvious: tweaking the head rank id to match the
caractheristics of the new inks. How to do that is perhaps over the
scope of this posting, which is intended to be a FACT TO SPREAD and
NOT A TUTORIAL. And it's unfortunately not a job for any people, it's
strong, heavy technical stuff.

After some attempts and trials and errors I could identify what seems
to be the parameter for the correction strenght. I brought it to an
higher value and here's the result:

 http://www.elenadomain.it/pub/NozzleCheckGood.png

nozzles fire perfectly straight again, now!

I hope my story have been of some interest.
Bye
/&"