What's causing this?

PeterBJ said:

The freeware program Irfan View can do a RGB separation. By inverting the image before the separation this will approximately be a CMY separation. After the separation the separated images are inverted again. This is my result with a rotated crop of your uploaded image, It is seen that the problems are mainly in cyan and yellow. This is consistent with the magenta cast in the stripes:

Following the ink lines in your photo of the printer it is seen that cyan and yellow are at the edges of the ink tube ribbon, I think this makes mikling's pinching theory very plausible.

With a Canon printer you can open the lid to centre the print head for cartridge change, and then pull the power cord without using the printers on/off switch. This allows you to move the print head from side to side to check for friction problems. Could this be done with the Epson printer? You could then move the print head from side to side and observe the ink ribbon.

Can you trick an interlocking switch to operate the printer with the lid open? This could show how the ribbon flexes when printing, and if striping also occurs with the lid open

Click to expand...

The R2000 has Red and Orange inks in addition to CMYK, so it's a bit more complicated to figure which causes the problem. I tried printing solid blocks of CMY colours, the problem is always there, although it's more pronounced when printing dark grey tones.

Here are 4 pictures of the cartridge moving around (we can do what you describe for Epson printers by simply unplugging the printer while the cartridges move). Can you spot something wrong?

I have no experience with CIS systems, so I cannot spot any errors, but I have a suspicion that the ribbon might bend too sharply, causing problems at certain print head positions. I edited post #10 to include this, but had not noticed your new post while editing.

Ok I edited my thread #8 where I said right instead of the left side.

Now I noticed in you nice action photos that you have secured the tubing onto the bottom of the extension bar and it looks to me that when the print head travels back under the bar that there’s not much room.

Could you move the securing point to the top of the bar or even better still bend the bar upwards to give more room to the tubing where it’s doubles back on the print head.

Nine time out of ten times where the tubing comes in contact with the moving print head it causes a slight judder and knocks the print head out of alignment for a fraction of a second, I had similar problem while printing photos with a CISS on one of my printers.

Try a setup and print a couple of plain paper sheets on the highest setting this should make the print head move the slowest and then look and listen carefully to hear where the knocking sound is at the loudest and then observe the tubing at that point to see if you can make some changes to its securing position.

CISS unit are great when they work perfectly but that can sometime never be achieved so compromises need to be made and in my case it was not printing any more photos with any of my CISS printers.

Take the print you made and put it back into the rear feed.
Now move the printhead carriage across and observe the flexure of the ribbon. You might have to do this with some speed to recreate what is happening when it is actually printing. Observe carefully and see what the ribbon is doing where the bands occur. Or simply place another sheet in front of the printed one and observe while the front one is printing.

dear costadinos your problem is simple when you use draft or fast speed printing its allow more fatal to air balancing in tubes and tanks -bottles - so when you back to normal it seem be well
cis system so difficult to work with printers in high speed cause its attached with tubes and its allow to get inaccuracy air flow(pressure) or ink from the CISS system.

The CISS setup is out of my league, but I do have the R1900 and noticed when I used high speed printing, it left strange lines in the print. I did the alignment check and print head clean, all to no avail. It is when I used the RPM profiles to print slowly did my problems go away. So, that is all I use now. I've read Red Rivers comments that the RPM setting is only marginally better in print quality than Best Photo, but it slows down print time considerably, but for me, I don't care about print time, but the highest quality print I can get. I mean really what is another, say 30 seconds.

Try the RPM print profile setting if using Epson profiles

Just One Man's Opinion

http://www.redrivercatalog.com/infocenter/setting-recommendations/66-polar-pearl-metallic-recs.pdf

Ideally you want to minimize the variance in the internal volume of the tube as the printhead carriage moves across back and forth, So the tubes needs to be set up as carefully as possible. Typically, when the machine is at rest the ribbon needs to have very little slack. Similarly when it is at the extreme left it will also have little slack. Adjust the looseness and centering point accordingly by trial and error.

Theory behind this. Take a straight piece of tube and calculate its internal volume. Now take same tube, calculate its internal volume when curved. Knowing that the tube is likely to stretch to accomodate the curvature you'll see that the volume increases when curved.

The CISS must accommodate these volume and pressure changes and it needs to buffer it out before it reaches the printhead nozzles.

PeterBJ I know has some electronics background. So what is needed is some resistance and capacitance to smooth these bumps out. The analog to resistance is resistance to flow and capacitance will be some volume arrangment that can absorb changes in volume. So ideally, the ribbon line irregularities need to have some resistance to flow as well as a chamber that can isolate the changes from the printhead. i.e RC smoothing circuit.

If you observe the ink lines with a bubble in it, you'll see it go back and forth as the printhead prints. The system must absorb and isolate that effect from the nozzle as best as possible.

On the cartridge, there is supposed to be a damper system......( damper = resistance + volume) . On refillables there is a long maze....that serves a purpose....resistance. The interal volume and some air in the cartridge is the capacitance.

So unless there is something electronically wrong inside the printer.We come to this .... something along the ink delivery path and how it is delivered is causing this issue. Trial and error with some background knowledge will find the problem using careful setup and observation.

Ink properties can have a part as well but is not likely in this case. Different viscosity and flow through a maze will vary resistance.

If you want the ultimate quality in printing, You need to engineer the tube system carefully as there are quite often small artifacts of these issues that goes unnoticed by most people and sometimes it is magnified by a particular situation and becomes noticeable. You'll notice that even on desktop printers like the Artisan 730 etc... particular engineering attention has been paid to the flexure of the tubes with varying stiffness braces to control the flexure carefully...now you know why. Take a look at a wide format and this becomes even more evident.

While the volume changes discussed as supply tubing changes from straight to curve may indeed occur, they are REALLY small. An even larger force for the print head to "contend" with is the pressure pulses resulting from the acceleration of the mass of ink in a tube that is at times stationary, then periodically forced to move, then stopped, then moved, etc. Anything that can reduce the acceleration (such as slower printing speed) reduces pressure pulses in the ink supply.

I finally managed to resolve the problem, turns out it has nothing to do with the CISS setup (I did try what mikling and The Hat suggested, many thanks to them for their help).

I did a google search for "vertical banding", and got a hit from an entry in the R2000 service manual. It's mentioned in the troubleshooting section, and it can be resolved by running the "Bi-d adjustment" routine found in the service program. So it seems it was indeed a hardware issue afterall (probably something to do with the timing when printing in high-speed mode).
Anyway, after running the adjustment program the banding was gone, hope this helps if somebody else faces the same problem in the future.

So, speed printing was the issue. Can you provide the link? That would be really helpful.

Thanks