Something Every Epson 9900 Owner Needs To Know About Banding

[John Trevino]

Hello everyone,

I recently had a banding issue that I got resolved through an onsite service call and I think everyone should be aware of this solution...
Like many of the other people I have read about in various forums, I had a banding issue that I was finding no easy solution for.  Using every possible troubleshooting suggestion listed in the manual produced no results.  Reading about other user experiences validated some of my issues, but I also found lots of speculation, over analysis, and frustration, with little information that made my issue go away.

So let me explain what did the trick...

When the tech arrived, the first thing he did was open the cover and examine something he called the "Timing Fence" with a small flashlight.  This is a translucent black tinted strip, about as wide as a piece of scotch tape, that runs parallel to the print head's belt about an inch above it, but sits close to the metal housing of the printer's interior (see attached pics, or links below for non-forum members).  

Within moments, he pointed exactly to where he thought the banding was occurring, which in fact, did line up with what I was experiencing on my prints.  He explained that the Timing Fence (also referred to in Epson's part list as the CR Encoder Scale, part no. EN1894) is etched with tiny lines closely spaced together and that there's a sensor on the print head which uses these lines to keep track of each row of ink being laid down.  If this strip has any dirt, smudges, dried ink, etc.. on it, then the sensor will be unable to read this, causing a jitter or imperfection of some sort in the print.  If anyone is familiar with woodworking, it's like using a router with a template that has a notch or burr on it.  That bump would be transferred to the finished work piece, which is the same thing that was happening in my prints.

SO HERE'S WHAT HE DID...
He took some Windex, sprayed it on a lint free paper towel (like the Rags In A Box product from Home Depot) and wiped it down real well.  Immediately, one of the areas of banding disappeared.  In the one area that remained we noticed that although it had been wiped down, the strip itself had popped out of its plastic holder and was sitting in front of it.  He took it and carefully seated it back into place behind the two end tabs and lined up the open slot area on the strip with the black plastic catch piece of the holder.  Problem solved completely.

I should say that before you attempt any of this yourself, this is a very fragile part.  Don't go after this if you are heavy handed because a crease or bend in this could mean having to get a new one and could be more time sidelined.  But if you're careful and work slowly are pay attention this will save you $275 easily.  

I know a lot of the manual's suggestions will probably work for most things, but this was just one of those things that seemed elusive to detection and as a result, incredibly frustrating.  Hopefully, this solution will become more widely recognized as another way of troubleshooting your printer banding issues.  

Good Luck!
John  


Link to pics for non-forum members:
http://flic.kr/p/auoPsL
http://flic.kr/p/auoPu3

[artbot]

but encoder strips do nothing to the x direction of the printer.  they simply read back an y position of the head carriage to the carriage board. also the encoder strip reads back a universal number.  it has no way to affect a specific color.  it gives a single number to the carriage that affects the firing of all the colors. the only thing a dirty vs. clean vs. damaged encoder strip can do is cause stair stepping of the print left or right, staggered head firing causing a murky overspray in certain places.  the encoder strip can not contribute or eliminate to a horizontal line/gap spanning the length of the print.  

[Wayne Fox]

but encoder strips do nothing to the x direction of the printer.  they simply read back an y position of the head carriage to the carriage board. also the encoder strip reads back a universal number.  it has no way to affect a specific color.  it gives a single number to the carriage that affects the firing of all the colors. the only thing a dirty vs. clean vs. damaged encoder strip can do is cause stair stepping of the print left or right, staggered head firing causing a murky overspray in certain places.  the encoder strip can not contribute or eliminate to a horizontal line/gap spanning the length of the print.  

I agree, but I suppose the banding referred to by the OP could have been opposite of the horizontal banding most refer to.  He really didn't describe the original problem that necessitated the service call.

[John Trevino]

Hi guys,

The reason I didn't include any lengthy description of my initial banding issues was because I just wanted to provide one possible solution right up front for folks that may have run out of ideas and who are looking for answers .  So backing up now, here's a couple detail pics of what was happening on my prints:
http://www.flickr.com/photos/blurryanimal/6247700634/
http://www.flickr.com/photos/blurryanimal/6247178069/

In my case, I first noticed the issue when I was profiling some Museo Silver Rag paper with the i1Photo Pro. In the final test checker swatch sheet that you print out as part of this process, I kept seeing some banding in the last two grey swatches.  Thinking I had an issue specific to the black inks, I created a greyscale in Photoshop and used it to evaluate and troubleshoot what I was seeing.

I did manage to resolve the streaks and banding I was seeing at this stage by adjusting the Paper Feed parameter under the Custom Paper setting and making a custom paper in the printer's control panel for the Canson Photo Gloss I was using at this point. Because the blacks no longer exhibited problems, I began printing some actual work.  On closer inspection of this work, however, I noticed two distinct areas of banding on my print.  These are the areas that wound up being traced back to the Timing Fence I mentioned earlier.

In my case, these areas I mention extended across the entire print and were present both horizontally and vertically by nature of simply printing both landscape or portrait oriented images.  I understand as artbot points out that the Timing Fence would mainly be responsible for problems potentially occurring on the "Y" axis and that all kinds of other scenarios could be involved that would be at the root of banding that people may be experiencing.  I've also been referring to this issue I had as banding, but after looking at the linked images above, if people determine it's technically call it something else, then fine.  Suffice it to say that there were imperfections I was experiencing that were traced back to the Timing Fence/Encoder Strip.  So if anyone has images they're printing that are coming out with the striations like the ones I linked to above, and you've tried all the usual troubleshooting methods, you might wanna give this a shot to see if it'll work for you

[BobDavid]

Thanks for the tip and the photos that you included in your post. I've been having this problem with Ilford Gold Fiber Silk. Especially in dark areas where there is a lot of black with little shadow detail. I may muster up the courage this weekend and try to tackle this per your instructions.

[davidh202]

Before you mess with the strip make sure there is something wrong with it by inspection!
If it ain't broke don't fix it !
I'd believe the problem would appear on every paper, not just one.
It is in an area that's unlikely to get dirty.

That strip may in fact be photographically exposed negative film with an emulsion that could easily be damaged.!

I tested mine with a damp Q tip on the very left edge and it sure appears to swell up and lighten like a negative or slide emulsion would when whetted!!
David

[John Trevino]

Indeed, if it's not broke you would have no reason to be monkeying around in the first place.  This is not a case of "I'm going to make my already fine output even better by...," just a suggestion to consider for those who are having real trouble, have exhausted all the basic and more technical troubleshooting solutions in the manual, and are at the end of their rope.

As for the problem showing up on all types of paper and not just one, in fact in my experience it was showing up on papers that included surfaces that were luster on up to glossy, but not matte. 

As for handling procedures and that strip being easily damaged, I can only speak to what I observed in my situation with the tech I watched do this.  While he worked carefully, there was nothing overly paranoid about how he went about it, no latex gloves, anti-static anything, this was handled barehanded, but gingerly.  So use common sense  and everything should be fine.

[AlexThiery]

I also have problems with canson photo gloss can you help me with the  Paper Feed parameter what number solved your problem?

[John Trevino]

Hi Alex,

In my situation I used a +20 setting, then later used this as part of a custom paper set up for that paper.  This was made from the control panel after sending a sample image to print, then adjusting it each time based off of what I was seeing, higher values for dark banding, lower values for white banding.  I also didn't see any particular improvement the higher I went.   Though in my case, the paper feed proved to not be the sole culprit anyway.   So I don't know if that would change in your case, or not, but the sweet spot for me seemed to be around +20-+35.