Epson XL10000

[dvhasker]

We have a Epson XL10000 scanner that has dull (out of focus) light line that runs the length of the scan. I removed the top cover and vacuumed out what I could but the line persists. I am hesitant to use forced air. I want to get to the actual sensor array.

Two Questions:

- It's a lighter line and not a darker line, so I'm unsure that it's dust. To me, if it were dust the line would be darker - Correct?
- is there a teardown or maintenance manual that would show me how to get into the actual sensor?

Thanks!

Duncan

 

[koraks]

- It's a lighter line and not a darker line, so I'm unsure that it's dust.

On a scan of a negative or a positive?

how to get into the actual sensor?

You don't want to and it's also virtually certain that you don't need to. In 99.9% of the cases, the kind of problem you're describing is down to fouling of the platen glass or the glass in the transparency lid. Note that when you're working with a flatbed scanner for transparencies, there are 4 glass surfaces that at some point need cleaning. This means in practice (partly) disassembling the unit and cleaning the glass. I've not worked on the 10000XL, but on the smaller Epson flatbeds it's a fairly simple matter of removing some screws and opening the housing (of either the box itself or the transparency lid) and then lifting out the glass. Be careful not to scratch it.

 

[dvhasker]

On a scan of a negative or a positive?



You don't want to and it's also virtually certain that you don't need to. In 99.9% of the cases, the kind of problem you're describing is down to fouling of the platen glass or the glass in the transparency lid. Note that when you're working with a flatbed scanner for transparencies, there are 4 glass surfaces that at some point need cleaning. This means in practice (partly) disassembling the unit and cleaning the glass. I've not worked on the 10000XL, but on the smaller Epson flatbeds it's a fairly simple matter of removing some screws and opening the housing (of either the box itself or the transparency lid) and then lifting out the glass. Be careful not to scratch it.

It's a reflective scanning. There are no marks on the glass itself as we had the cover off.

 

[koraks]

Ok, thanks for clarifying. You cleaned the glass top and bottom?

PS: by cleaning, I mean properly cleaning. Not vacuuming the unit out. This will likely not be sufficient.

 

[dvhasker]

Ok, thanks for clarifying. You cleaned the glass top and bottom?

PS: by cleaning, I mean properly cleaning. Not vacuuming the unit out. This will likely not be sufficient.

Yes, glass is clean. To add to this it is not a sharp or in-focus line so it's not sitting on the sensor and it's not on the glass. More like a smudge. wider and lighter with a non-defined edge.

 

[koraks]

Can you post an example?

One thing that comes to mind is something somehow settling on the mirror or the lens. But it's rare for this to happen.

 

[dvhasker]

Hi Koraks

This is a scan of a 4x6 print. I thought something on the mirror myself, but would that not be darker and not lighter?

 

[koraks]

Thanks, I see what you mean. It's subtle, but it does appear to be colored and not monochrome and it appears to be fairly sharply delineated as well. Could be dust on the sensor alright as you initially expected, but I'm not 100% certain.

 

[dvhasker]

Thanks, I see what you mean. It's subtle, but it does appear to be colored and not monochrome and it appears to be fairly sharply delineated as well. Could be dust on the sensor alright as you initially expected, but I'm not 100% certain.

Any idea why it would be lighter as opposed to darker? I'd love a tech manual for the scanner.

Thanks,

Duncan

 

[koraks]

One thing is what something like dust does to the actual sensor signal, another issue is what the scanner firmware/software does with an aberrant signal. Simply put, a speck of dust can cause a signal that doesn't behave like the firmware/software expects it, causing it to be tricked into interpreting it as always illuminated. It depends on how the hardware interacts with the low-level software; this is very implementation specific and generally not covered in a tech manual.

 

[bernard_L]

Any idea why it would be lighter as opposed to darker?

This suggests some dirt/smudge on the calibration strip. It's a strip tat is read by the sensor at the beginning of each scan, to calibrate out the deviations from uniform response of the pixels. Don't know about the XL10000, but on my V700, that strip is on the underside/top of the top plate, just above the scanning glass. I think (IIRC) it is covered by the glass. Still on my V700, removing the top plate is easy, but separating the glass from its frame to access the calibration strip is not trivial. Hopefully you have a smudge on the bottom (top once you have removed and flipped it) of the glass, above the cal strip.

If you find there is dirt between the glass and the strip, I would suggest to overlay the strip (and glass) with a strip of smooth white paper. Something like satin or semi-matte inkjet paper. Be aware that this strip defines the white point; so strange things might happen if your inkjet paper has optical brighteners...

Good luck