Possible explanation for strange print patterns and other things

[Grandad35]

I would like to propose some theories to try and explain some of the strange printing problems that have been reported. It is also the purpose of this discussion to develop better operating procedures to minimize the dreaded "clogged nozzle" problem, especially with refilled carts. These are only theories - please feel free to post your experience or knowledge on any of the subjects covered below.

First, some background. When I do a head cleaning on my i9900, I can clean all colors at once, the cyan, magenta and yellow as a group, or the black, PM, PC, Red and Green as a group. I also seem to remember reading somewhere that the head cleaning operations use air to clean the heads. What makes the most sense is that the printer pulls a slight vacuum on the discharge of one or both of the groups of nozzles in the print head. This also explains how a cleaning cycle gets ink to start flowing - it is difficult to understand how wiping the heads or activating the thermal elements alone could do this. Does anyone know if this is the case? Has anyone taken a Canon apart to see how it works? In any case, let's start with the theories (remember that these are just conjecture at this time).

My print head is divided into two sets of nozzles. The cyan, magenta and yellow are ejected from one set of nozzles, and the other 5 colors are ejected from the other set of nozzles. This is why these colors must be cleaned as a group.
When the print head is off-line, some sort of soft rubber-like material comes up against the print head and seals the exits of the nozzles so that they don't drip or dry out.
Suppose that one of the carts in a color group has a pressure mismatch such that it is pulling a slight vacuum at the print head's ink pickup. This vacuum will tend to pull ink for that color back through the nozzles. If the seal covering the nozzles provides a good seal to the outside, but a not-so-good seal between adjacent colors, ink will be pulled from the other colors back into nozzles with the vacuum. When you print a nozzle check pattern, you will get the wrong color (or maybe even a mixture of colors) in that nozzle. This would also explain the post (that I can't find) where someone reported seeing the wrong color ink on the print head's pickup when the cart was removed.
Suppose that one of the carts in a color group has a pressure mismatch such that it is has a slight positive pressure at the print head's ink pickup. This pressure will tend to push ink for that color out through the nozzles. If the seal covering the nozzles provides a good seal to the outside, but a not-so-good seal between adjacent colors, ink will be pushed from the high pressure color into the nozzles for the other colors. When you print a nozzle check pattern, you will get the wrong color (or maybe a mixture of two colors) in the other nozzles. Ref - Nifty-Stuff's post (http://www.nifty-stuff.com/forum/viewtopic.php?id=103) shows a scan of a nozzle check pattern where it looks like cyan got pushed into the other nozzles.
I am coming to the conclusion that you should never remove a cart unless it is to replace it, and that the new cart should be ready to drop in ASAP to keep the top of the print head pickup wet with the proper ink. If the ink film between the print head and the filter on the cart is lost, at least one cleaning cycle must be run to pull ink from the cart all the way through to the nozzle exit, as any air that gets into the pickup on the print head will break the capillary action that pulls ink from the filter. When my printer was new, I never touched the carts and never had a "clogged nozzle" problem (I began to wonder what everyone was talking about). After successfully replacing three OEM carts when the printer gave me a "low ink" warning, I replaced all 8 carts with a complete set of new (old formulation) InkGrabber carts, did a single cleaning cycle and never had a problem. It was only when I started to change carts a lot to test various refill techniques that I started to have "clogged nozzle" problems, often with colors that I didn't touch.
I am currently running a test where I won't touch the carts for any reason to see how often I have to run a cleaning cycle. I recently refilled a complete set (using the second vacuum technique) to get every cart completely full at the start of the test. I have printed 12 8.5x11 sheets of photos so far with no problems.
When I was testing various refill techniques, I noticed that the Canon carts seem to take more pressure to blow ink through the exit (to make sure that there is a continuous ink path to the bottom of the filter and that the bottom of the filter is "wet" when the cart is installed). I attributed this difference to a difference in the sponge or filter used in the two carts. This implies that there could be a pressure differential between carts from different suppliers, and that perhaps carts should always be kept in matched sets to keep the pressures better matched, even when refilling.

There is a recent post (http://www.nifty-stuff.com/forum/viewtopic.php?id=140) that discusses a common approach of periodically pulling the carts to visually inspect the ink levels so that the carts can be refilled before the ink chamber becomes empty. My OEM and InkGrabber carts have always tripped the alarm when they are empty - and I mean EMPTY. Up to that point, the indicated levels on the status monitor display can not be relied on to give an accurate ink level (how does the printer know how much ink is in the upper sponge when the cart is installed?). The sponge needs to hold enough ink after the ink chamber becomes empty to print a few extra pages if you are going to rely on this sensor. I think that I remember reading that the printer will continue printing until it calculates that the sponge is almost empty before refusing to print any more (please correct me if this is wrong). This is fine if you will be replacing the empty cart with a new cart, as you will get more usable ink out of the old cart and you just throw away the empty cart (the OEM's plan).

The problem arises if you refill your carts and wait for the sensor to tell you when the cart is empty. If you happen to be running a large print job when the ink chamber empties and you aren't constantly watching your computer for a low ink level alarm, you can end up with a lot of air in the sponge before the printer stops printing. If the sponge wasn't almost completely filled during the last refill, you may even run out of ink. Refilling carts with nearly empty sponges requires unusual techniques to get the air out so that they will print properly when refilled (http://www.nifty-stuff.com/forum/viewtopic.php?id=124).

The inspect and refill technique seems to be a good solution to this problem, but pulling carts for inspection seems to bring an increased occurrence of ink feeding problems. Which is the lesser of the two evils? What are your experiences? If you refill, please post which technique you use and the problems that you encounter (if any).

 

[Nifty]

Ahh.... a rock and a hard place is where I'm finding myself! If your theory is true (and we're not yet sure that it is) then it is a hard call between frequently pulling the cartridges out and air in the sponge.

For me, if I had to choose between the two, I'd choose pulling the carts out more often, BUT based on some information I've been gathering I'll:
Try to pull them out less often and also be quicker about putting them back in (or swap with cartridges I've already refilled and are ready to go).

Grandad, one of the most interesting things you said was, "This implies that there could be a pressure differential between carts from different suppliers, and that perhaps carts should always be kept in matched sets to keep the pressures better matched, even when refilling." I never even considered this and is a very excellent theory / idea.

BTW, FYI: On my i860 I have the choice between cleaning all or just black (pigment) or just the colors (CYM and dye black).

 

[Grandad35]

This is an update on the "don't touch the carts" test. When the carts were first installed, several cleaning and 1 deep cleaning cycle were required to get a perfect nozzle test pattern. There were a lot of cart changes before the start of this test, and it took a while to get everything back to perfect condition.

A total of (35) 8.5x11 sheets of photos were printed when the low ink level warning came on for both the PM and PC at the same time. During the printing of these 35 sheets, there were two instances where the magenta was missing on a nozzle check (all other colors were perfect and there was no magenta at all) - a single cleaning cycle fixed the problem in both cases. Otherwise, there were no problems while printing the 35 pages. Note that the power to the printer was never turned off during this time to prevent any extra head cleaning cycles. If the printer had been set to go into standby when it was idle and did an automatic cleaning cycle every time that it was reactivated, no additional cleaning cycles should have been necessary.

The total of 35 sheets before emptying a cart was at least 15 sheets less than expected. It is assumed that the main reason for this was the extra cleaning and 1 deep cleaning cycles performed after the carts were first installed.

After the low ink warning, all 8 carts were pulled to inspect the ink levels. Only the M, C, PM and PC were replaced, as the other 4 carts were still more than 1/2 full. Actually, the magenta was also more than 1/2 full, but I wanted to see if there was a reason why it lost its feed twice. A single cleaning cycle was run and a nozzle check pattern printed. There was ink in all of the colors, but not always the correct color (as has been reported by others). A single heavy color pattern was printed to clean the ink channels, and everything was back to normal. It is interesting that pulling the carts when everything was working perfectly just a few minutes earlier and replacing them with either the same or full carts within 60 seconds of the first cart being pulled was still long enough to cross-contaminate the colors.

These two images show the PC and PM carts before and after using the "Pressure Refill" technique to force ink back into the sponge chamber. The (natural) lighting changed slightly between the shots, so the appearance of the bottom sponge may be slightly different - it was actually the same.

It can be seen that the bottom sponge is still almost completely full (there is a little air in the top corner of the sponge next to the ink chamber). The appearance of the top sponge was similar to the bottom sponge when the carts were first vacuum refilled, so it appears that the upper sponge has been at least partially drained. From the appearance of the partially used carts, it can be seen that ink from the upper sponge is used before the ink chamber starts to drain.

The "Pressure Refill" technique was easy to use and allowed about 9 CCs of ink to be injected instead of the normal 6CCs required to fill just the ink chamber.

The only thing that was strange about the magenta cart is that it took more pressure to blow ink out of the exit than the other carts. Two drops of isopropyl rubbing alcohol were dropped onto the exit filter, left sit for 1 minute and the pressure test repeated. It did seem to take less pressure, but this could be because we often see what we expect to happen, the alcohol could have diluted the first ink to come out and the pressure will increase over time, or the alcohol could have broken down an old ink residue.

Based on this single "don't touch the carts" test and the success of refilling using the "pressure refill" technique, I plan to continue waiting for the "low ink" alarm before pulling the carts. The "vacuum refill" technique will be used when the bottom sponge looks like it is getting too much air or if the cart doesn't feed properly. If periodic extra cleaning cycles continue to be required, the printer will be set to go into "standby" mode when it is not in use so that it will do a cleaning cycle when it is reactivated.

 

[fotofreek]

To add to the do not remove theory, I just received my first carts from Alotofthings for my i960 and the caution note on the package says "once you install the cartridge, do not remove it except to replace." These carts are designed for refilling and refilling is suggested by Alotofthings which also sells bulk ink (Formulabs brand). I've seen some good posts about this ink on the printers newsgroup.

After refilling and using my original OEM carts for about six months one of the carts gradually started feeding less and less, leaving a band near the trailing end of the print. Cleaning and printing a purge print didn't improve it so I replaced all the carts with MIS virgin carts filled with MIS inks (Yes, the fill plugs are a nuisance!) Worked fine. I have the old nearly full carts sealed at the air vent and outlet and in a baggie and will go back over the posts to see if I can revive them when I have time. I don't know which cart is the problem, and I am not sure which technique to start with. Maybe the hot water bath first or a bit of Isopropyl alcohol followed by the hot bath. Is the 130 degree suggestion critical? I think that our tap water is around 145 to 150 at it's hottest. Would that distort the cart or cause other problems. No one had mentioned it but I assume that the cart should be in a waterproof baggie.

 

[Grandad35]

Fotofreek,

Thanks for the info on your new carts.

I just ran a few tests on a cart that was previously dissected by placing the individual parts in boiling water for 5 minutes. The cart casing showed no sign of distortion. Likewise, the filter and bottom sponge did not show any change - they still had the same dimensions and water absorption characteristics as they did before the test. The top sponge got about 60% thicker and 10% longer. There was no sign that the top sponge deteriorated in any way, and when it was stuffed back into the cart, I couldn't tell the difference. From this test, it would appear that a 150 degrees F soak wouldn't do any damage to a cart.

I also soaked the sponges and filter in Isopropyl (rubbing) alcohol for 5 minutes. There was no sign of any deterioration from this treatment, so these parts should easily tolerate a little alcohol without any problem.

On your old carts, try blowing into each vent with its exit uncovered to see how much pressure is required to force ink to drip out of the exit. If you find a cart that is difficult to get the ink to drip out of, I'd lay odds that cart is the problem.

 

[Nifty]

Grandad, I noticed in your images that your ink level in the reservoir is pretty high, especially in your cyan cartridge. Have you had any "overfilling" problems when putting this much ink in?

Also, you pull the carts when you get the warning. Is this the warning from the prism or from the internal droplet counter?

 

[Grandad35]

Nifty-stuff,

I haven't had any problems with "overfilling" the ink chamber. The new carts that I have looked at are also filled to the top. Using the "vacuum refill" or "pressure refill" techniques, it is possible to overfill the sponge chamber so that the ink is above the top of the sponge. This will cause the cart to drip until the ink is either absorbed into the sponge or drips out until the ink is all contained in the sponge(s).

The low ink warning was from the prism. As the photos show, it did a great job (at least in this case) of indicating that the ink chamber was completely empty, but that the bottom sponge was still full of ink. I looked at the "predicted ink level" display just a few sheets before the "empty" warning appeared, and it showed the same "30% full" level that it had shown since the new carts were installed (the carts that were replaced at the start of the test weren't empty, so the counters weren't reset at that time).

 

[Nifty]

I looked at the "predicted ink level" display just a few sheets before the "empty" warning appeared, and it showed the same "30% full" level that it had shown since the new carts were installed (the carts that were replaced at the start of the test weren't empty, so the counters weren't reset at that time).

So if I understand correctly, you installed new cartridges and your meter said 30% and then after running that cartridge to "dry" you got the warning that it was empty?

 

[Grandad35]

Nifty-stuff,

In a word, yes.

There is a post at (http://www.nifty-stuff.com/forum/viewtopic.php?id=140) that explains in general how the ink level indication system works. On my i9900, however, the indicated (calculated) ink level counts down to about 30% and holds there until the prism is uncovered. When the prism is uncovered, the display shows an empty ink tank and issues a "Low Ink Level" warning. It will still print at this time, and (according to the manual) stop printing and issue an "Empty Ink Tank" alarm when it somehow determines that the cart is completely empty - probably by a calculation of ink usage and an estimate of how much ink is held in the sponge when the ink tank first becomes empty.

If a cart is swapped before the prism is uncovered, the printer doesn't know that it now has a full cart, so it keeps counting down from the previously calculated ink level. Because of this, the only way that the ink level monitor will be approximately correct is to replace a cart only when it is completely empty. This is a good way to operate when you don't refill, but if you refill it is far easier to replace all of the carts that are running low when the first cart empties. This also minimizes wasted ink and the dreaded "Ink Tank Full" alarm, since the printer always runs at least one full cleaning cycle when it detects that a cart has been replaced.

Right now, my display shows full tanks only on the two carts that were empty when they were replaced. The other level displays remain stuck at 30%.

Does this clear it up, or just muddy the waters some more?

 

[Nifty]

Pretty clear! If I have any more questions I'll post them in that forum... Sorry to get a bit off topic!