Canon InkTanks and the Notorious Air in tube issue

[websnail]

How safe is the using a medical syringe method to suck back ink through the delivery tubes?

I've be a bit wary of using a syringe directly via the port as you might stretch it a bit too much. Use of a tapered plastic needle/tip as an adapter would provide a good seal. You would probably need to cut the length down to get a good fit but it wouldn't stretch the port too much and potentially damage the seal.

Beyond that it seems to work if that video is anything to go by.

 

[brwinters]

I have a G7020 and G620 (basically a G3262 with a 6 color process), and started out with a G3202 before. This is a universal problem with MegaTank printers from what I can tell. In my experience, air in the lines not from moving the printer around always comes from disuse.

My working theory is that "aspiration," "ink ingestion," etc. tends to happen when there's negative pressure in the ink lines and there is a path for air to seep in. Normally this is to a very negligible degree. However, when you don't put any pressure back in the lines for however arbitrarily long a time period, it seems to get to be a real pain. My theory is that the air gets up through the print head as it seems to be, more often than not, a spontaneous change. I.e. I open the lid 4 days, no air. Open it on the fifth day and suddenly there's 4 inches of backflow in my K feed. Though I have noticed more drawn out issues on occasion. As of now, I had 4-5 inches of backflow in my K and did a deep clean the other day to fix it. A couple days later, my K is good but I have, cumulatively, about a pea-sized amount of air bubble in two or three of my color lines.

Canon seemingly didn't put any real effort into integrating their CISS with the printer. Great printer and a pretty solid CISS design, but then they are using print head modules which are, if I understand correctly, in essence a disposable sponge style cart/head combo with some holes on top. Admittedly I have yet to break one open on account of taking care not to run them dry and kill them. And their relative scarcity in Canon bothering to stock them online. But it seems like just a TS3522 style sponge cart/head with a CISS port for whatever color. I don't see any outward evidence that it would be a proper damper inside. Lacking a damper in favor of a basic sponge would explain why my earlier, pre-automated-maintenance-print experience with any Canon tank kit, has been riddled with aspirated ink lines like clockwork every few days of downtime. No valve, and a sponge which air can make a predictable path up through with time. Not sure if the purge plate should be letting that much air through though...

This problem is mostly, but not entirely, mitigated by so-called "maintenance prints," wherein I have Windows task scheduled to print a page with all primary color blocks each day. Large enough blocks that it should in theory get full coverage of the print heads and use a slightly measurable bit of ink to get the CISS to add more pressure to the lines. It also stops it running repeated cleaning cycles every few days, which is what I imagine is Canon's half-half-solution to this issue. (Really they do that on any printer to hide head clogs, so this would probably be a side-effect more than not...)

Though I may need to make the color blocks larger on account that I still get occasional issues. Or bite the bullet and buy Qimage instead of relying on hopefully-accurate CMYK colors in a PNG file.

 

[The Hat]

Canon seemingly didn't put any real effort into integrating their CISS with the printer.

From day one, all home installed CISS units on Canon printers were plagued with issues, even the best setups were still troublesome, and now just because Canon makes and install them themselves, doesn’t mean diddly.

In a nut shell, CISS don’t work very well on the Canon printer, I had 3 CISS units on some of my Canon printers and they were printing all day every day, but problems still crept in, in the end I dumped the CISS in favour of refilling the carts as the best option.. You can’t beat a proper refilling procedure for reliability..

 

[brwinters]

You can’t beat a proper refilling procedure for reliability..

Canon ought to overhaul their design to be CISS to begin with. I suspect it's the overall theme of "retrofitted CISS" that's the cause of the issues. Canon having also done essentially an OEM retrofit falls in line with that.

I agree that refilling is, in a way, more reliable/less fiddly. Though it is inconvenient and, in its own way, also fiddly. Convenience is the whole reason I got into CISS instead of buying another cart set after my starter 275/276 set died. Refilling was messy and a pain compared to turning a bottle over in a socket.

Especially from an average end-user perspective where cart refilling isn't really in the question, CISS is the most complete and promising direction forward. It just needs to be properly done. I.e. when I explain economics of ink systems to customers, "check the ink lines and do deep cleans until it's filled*" is still less complicated than the concept of refilling a cart properly.

*Also that Canon lacks user-serviceable waste tanks on consumer models of CISS, being a major oversight and a design flaw to be aware of... I have redirected some people to Epson offerings on that front alone.

 

[websnail]

One nugget to add to this as I've spent a few months working on this... If one or more of the printhead labels are left in place they can create a tiny air gap that allows air in between the ink outlet in the lid and the printhead receiver. That showed up here with a G6050 that had a very slow backflow in the yellow channel. As soon as I removed the label it was resolved.

Still playing with other Megatanks to see if I can identify the other issues on this but at the moment I strongly suspect the design of the ink outlets in the lid likely results in the silicon rubber degrading over time.


As for the Epson/Canon attempts to "invent" the CIS system, they really did make a whole bunch of design decisions that make maintenance harder not easier and reinforce the idea that obsolescence is built in. Incredibly frustrating to see them keep making these basic mistakes.

 

[The Hat]

As for the Epson/Canon attempts to "invent" the CIS system, they really did make a whole bunch of design decisions that make maintenance harder not easier and reinforce the idea that obsolescence is built in. Incredibly frustrating to see them keep making these basic mistakes.

I’m not, give the customer something they want, then walk away..

 

[brwinters]

If one or more of the printhead labels are left in place they can create a tiny air gap that allows air in between the ink outlet in the lid and the printhead receiver.

Now that you mention it, that rang a bell... I remember seeing labels on the print heads for all of my tank printers and being under the impression that they were permanent-type ones not meant to be removed.

I just peeled them off both my G620 and G7020 printheads, reinstalled and did a couple deep cleans each to get the ink lines completely recharged. (One legitimate improvement I can appreciate is the manual service cutoff valve that Canon seems to have started including across the board... My old G3202 didn't have one and it always took a lot more effort to get that recharged versus these ones...)

I'll pause the daily maintenance print jobs on my print server and see if there's any noticeable uptick in reliability over the next few weeks. I might also tell the friend I gave that G3202 to, to also do this if I notice any improvement. The G3202 was worse with losing charge than either of these. He also has my daily waste print job automated to counter it (and he got the printer basically free).

Still playing with other Megatanks to see if I can identify the other issues on this but at the moment I strongly suspect the design of the ink outlets in the lid likely results in the silicon rubber degrading over time.

What's leading you to this hunch? Also, is the whole socket silicone or is it some type of internal piece to it? I only briefly poked around on the receiver on mine but it doesn't feel rubbery, more like a hard rigid plastic. (The white cylindrical bits.)

 

[brwinters]

I'll pause the daily maintenance print jobs on my print server and see if there's any noticeable uptick in reliability over the next few weeks. I might also tell the friend I gave that G3202 to, to also do this if I notice any improvement. The G3202 was worse with losing charge than either of these. He also has my daily waste print job automated to counter it (and he got the printer basically free).

No change -- Only a couple days later and my black line was full of air again on my G7020.

 

[websnail]

No change -- Only a couple days later and my black line was full of air again on my G7020.

Only other things I can suggest are:

1. Clean the top of the printhead, particularly around the ink receiver in the top, to remove any crud, hair, lint, etc...
2. Clean the carrier lid carefully as per the printhead and look for any signs of damage to the silicon rubber that should lock down over the printer receiver/nipple.
   
   
3. Purchase a new printhead (last option of desperation?)

It might be possible to create a silicon grommet/gasket that seals up any gap between the receiver and the outlet but I wouldn't know where to begin. Only other thing I can think to check is for any physical damage or disconnect between the tubing and the lid but that's just educated guesswork.

 

[brwinters]

Only other things I can suggest are:

1. Clean the top of the printhead, particularly around the ink receiver in the top, to remove any crud, hair, lint, etc...
2. Clean the carrier lid carefully as per the printhead and look for any signs of damage to the silicon rubber that should lock down over the printer receiver/nipple.
   
   
3. Purchase a new printhead (last option of desperation?)

It might be possible to create a silicon grommet/gasket that seals up any gap between the receiver and the outlet but I wouldn't know where to begin. Only other thing I can think to check is for any physical damage or disconnect between the tubing and the lid but that's just educated guesswork.

I'm skeptical that it's some kind of damage that's causing this issue. All three of my Canon tanks have done it. Different models, different heads, different ink processes, etc. And they all do it. I'm inclined to believe it's a design flaw.

Canon's bad about stocking replacement heads in their store and every other option is massively overpriced. I suppose if I get into production work soon I'll have new heads ordered to have on-hand spares. But even then it's not worth replacing a working head for a problem that seems universal and can be fixed with maintenance prints. (Which I have to do anyway to stop wasteful auto-cleaning.)