HP Designjet XL 3600DR PS Service Manual

[blackcat4866]

More information ... and I know even less than before.

The roll2 media sensor made no difference, and detects paper present/paper absent in diagnostic mode.
The roll2 rewinder is now passing the diagnostic. I have no idea why.

I figured out that I can key the top door and watch the paper passage during the "Unloading". My abbreviated test is to:
Load roll2
Load roll1 (roll2 now parked)
Print an internal print (like Usage Page) from roll2.

So what usually happens is:
Roll1 parks.
Roll2 feeds paper up the vertical path.
Slight hesitation when paper trips the clean-out media sensor.
Pinch rollers raise ~2mm.
Paper advances to ~13mm from the star wheel guide, and pinch rollers lower.
Simultaneously paper advances into the star wheel guide and starts measuring the left edge location x3
Paper advances ~200mm then repeats the left edge measurement
If it doesn't like the measurement, the pinch rollers raise/lower to release any buckle.
Paper backs up ~200mm then repeats the left edge measurement
Carriage moves in ~100mm and the paper advances x3 edging past the line sensor
I think this is where it goes wrong:

When it fails, the paper backs down into the roll2 tray and displays "Unloading Roll".

When it's successful, printing starts.

I am suspecting the Line Sensor-Tetris, or Carriage PWB, or Engine PWB ... except for the fact that roll1 uses exactly the same paper, line sensor, PWB's and always detects left edge and leading edge correctly. We'll see what tech support has to say. =^..^=

[blackcat4866]

Another cross-off:
The Pinches Assembly (formed aluminum frame with ~150 star wheels) is not warped, and raises and lowers appropriately in service mode. I suspect that the pinches assembly is too far down the path to be a cause. Sometimes the paper roll does not get that far. There's about 10 seconds from the time the paper edge clears the grit roller, until it fails some part of the queueing up process.

Tech support provided me with the latest version of the manual. The older version showed photos of the Pagewide XL paper deck with 2 rolls per tray. Somewhere along the way it got simplified to one roll per tray.

I got bumped up a level in tech support. I wonder if that means the fixes will be more difficult? =^..^=

[blackcat4866]

If anybody is interested I got bumped up to Engineering in Spain.

Firmware version 4_20_21.1 seems to solve this issue for the three standard paper roll sizes 24", 30", & 36", in whatever configuration you like: upper or lower. I believe the most current firmware is version 7_21_20.1
There's a version 5, but but I don't know if it presents symptoms or not.

And I don't know how custom sizes, ISO sizes, and less standard sizes like 15" and ANSI widths are affected.
I'll be interested if you can add any detail. =^..^=

[blackcat4866]

Just to confirm, this issue exists in firmware versions:

CYCLOPSMARS_7_21_20.1.fmw
CYCLOPSMARS_8_21_39.11.rfu.fmw
CYCLOPSMARS_8_21_39.14.rfu.fmw

Some machines work with versions:
CYCLOPSMARS_4_20_21.1.rfu.fmw
CYCLOPSMARS_1_19_18.6.rfu.fmw

It's fixed in version:
CYCLOPSMARS_8_21_43.13.rfu.fmw

=^..^=