Canon ip4200 CIS - returning to the fray!

[websnail]

Hey Mikling, yeah that consideration has been at the forefront of my mind but in this particular instance the black banding appears to be a printhead issue for this particular printhead as it did it immediately with the OEM cartridges installed from new.

The pressure fluctuation from bending/flexing is something I did incorporate into my posts from earlier in the last 24 hours hence the air buffer which should at least attenuate the issue somewhat.

For all that the evidence I've seen in my print-outs indicates the banding issues return is not related directly to the flexing as it's not localised to one or other side of the page, instead it seems to be a starvation or over-pressure issue as I've demonstrated and thought I'd cured in the past.


Further to that it appears that there's levels of this "gradient" banding that are indistinguishable visually.

From what I can tell, the initial gradient banding can come about either due to over-pressure (from which the printhead can recover) or print-head starvation (tubing ID too small, blockage in lines, etc..). IIRC the problems with the printhead I started with a couple of days ago, were actual starvation issues so I'm pretty sure I stressed out the printhead enough to make it feed badly. Time will tell if I can revive it or not but we'll see..

 

[websnail]

Right... I may well have finally learned most of the lessons I need to on this having just completed two kits and had them work right from the get go.


The main change I've instigated is one that most refillers will recognise as being fairly sensible... ie: allowing the cartridge to reach an equilibrium. My problem is that I've been vacuum filling the cartridges and in the process removing virtually all air, most importantly including the area in the sponged segment of the cartridge.

If you look in a virgin Canon OEM cartridge you'll note that the cartridge has no ink in the area above the sponge and the secondary/top layer of sponge is not exactly soaked in ink either. I've been ignoring this rather obvious "feature" in the OEM cartridges and essentially creating greater weight of ink thus over-pressuring the printhead. The word you're looking for here is "Doh!".


This is by no means 100% tried and tested as I've only managed to try one of each kit type but it seems that placing the fitting over the sponge or the spongeless tank segment of the cartridge makes no difference to the actual flow of ink, nor creates problems. If I was a betting man I'd suggest that the fitting above the sponge will produce slightly more stress but when you offset that against the need to physically adapt the printer casing to accommodate fittings above the tank I think it's worth the risk. It's also worth bearing in mind that I've had kits setup with fittings in this position and (on reflection) the air buffer in the sponged segment... and they are still working a year on so it can't be a major issue.

Overall the key here is that the ink in the cartridge has levels that are equivalent to a virgin OEM cartridge (or possibly slightly lower) before you incorporate them into a CIS kit.

Additionally it's important that the tubing is of sufficient inner diameter to be able to feed ink to the printer quickly enough. I've used some tubing with an ID of around .5mm which simply isn't enough and it stressed the printhead beyond use even with OEM cartridges so be warned.


Anyways... I've got a few new things to play with now as well so I'll let this thread lie and I'll write up a knowledge base article based on what I've worked out sometime in the next week or so over on continuousink.info .

T'is nice to finally get somewhere.. but then, how many times have I said this before

 

[websnail]

... you know having been and said soooo many times

That's it, I understand it all now...

... and then being promptly bitten on the ass I've spent the last 2 days waiting for the gremlins to appear and start beating the cr*p out of my new found confidence.

Well, after the fourth refurb'd printer where I've sat and re-primed the cartridges and kits, I'm now starting to feel very confident that my paranoia is at an end (well almost ).

It seems the only other thing that can creep up on me and cause problems is ink foam, which is something I've learned about through all my Epson experiences.

Essentially you cannot force air back into the cartridge and you had better not open the ink exit port to air before all vacuum has disappated through the top of the cartridge.

What I've found is that if you remove the tubing from the cartridge and let the cartridge drip to get its equilibrium it'll work just fine once you reconnect everything. Oh and in the process of re-priming I found a few cartridges where they'd taken in a reasonable amount of air as foam and they were really in trouble. A quick re-prime with hard vacuum followed by the dripping routine and wham the cartridge was good to go. Only in one instance did I have to remove and replace the cartridge because it was cloggated. A quick purge should fix that one up and it'll be useable again.


All in all though, I think the troubleshooting guide is actually pretty complete now and just needs writing up with pictures..

 

[websnail]

Ha! Spoke too soon...

It seems the gremlins were patient this week allowing a 2 day confidence boost But anyway, a little more digging has indicated that whilst the equalisation approach has merit if you're using your fittings over the sponge there's still an issue which I hadn't considered.

In a nutshell I'd always assumed (yep, that word!) that the sponge was the key regulator in terms of ink flow but as it turns out there's something else too... Specifically, the channel leading from the spongeless ink tank to the sponge itself. The sponge will help but the channel stops the ink from flowing into the sponge too quickly so it would seem that there is a very good reason to install your tube fittings over the spongeless segment after all.


So, it's back to the drawing board again but in fairness I think I'm running out of variables to screw up so it can't all be bad

 

[websnail]

*sigh*.

This is like snakes and ladders... Printhead, was new, started leaking ink because the seal around the rubber grommet wasn't enough... I pull the kit to put in silicon grommets (thin ones) and pop it all back in... Bingo, the gradient banding starts all over again and this is with a stabilised kit, that has fittings over the spongeless tank part of the cartridge, etc...

WTF am I missing here?... *sigh x2*

 

[Grandad35]

Websnail.

A few questions from someone who has no experience with a CIS:

*Did you "reinitialize" the cart before it was reinstalled? Is it possible that the "ink link" in the cart was broken when it was removed, and that the sponge no longer pulled enough suction on the ink tank to pull ink from the CIS bottle?
*Would the printer resume normal operation if a conventional cart was installed in the problem position?
*If the ink tube from the CIS is removed to vent the ink chamber, would the cart drip from its exit port as a normal cart would? If so, fill the ink chamber back to full, let the exit port drip until the ink level is down to your desired operating point to allow the sponge to become "normalized" with ink, then reconnect the supply tube.

 

[websnail]

Websnail.

A few questions from someone who has no experience with a CIS:

If there's one thing that I know, it's that you've amassed a lot of parrallel and complimentary information so believe me I'm listening when you speak, besides a fresh pair of eyes n'all that

*Did you "reinitialize" the cart before it was reinstalled? Is it possible that the "ink link" in the cart was broken when it was removed, and that the sponge no longer pulled enough suction on the ink tank to pull ink from the CIS bottle?

In truth, no I didn't re-initialise the cartridge although I do check that there was ink flow pulling through from the tubing by lowering the cartridges way below the printer to check for normal flow. There was.

*Would the printer resume normal operation if a conventional cart was installed in the problem position?

Well here's the thing.. This printhead had originally had gone through a lot of change-overs with at least 2 different CIS kits. When the 1st kit had shown banding I immediately killed the print job and removed the CIS, putting in some virgin OEMs and did a cleaning routing, high quality 4 colour purge and then a standard purge print. That worked fine. Where the second kit came unglued was when it had been working just perfectly for 3 days and then I noticed that there were small amounts of magenta appearing on pages... I looked carefully at the ink flow in the tube (I let a small air bubble in as an indicator) and noticed immediately that the ink was flowing towards the cartridge slowly but definitely and obviously there was an air leak.

Put in some silicon grommets (not new ones which may have been significant) to seal the obvious air leak around the rubber grommets and cartridge exit nozzle and put the CIS back in the printer.

Now, additionally to this I noticed something else... I deliberately left the tubing long as I intend to install a bag system underneath the printer rather than bottles (greater capacity, no air, little if any mess, etc...) but it does mean that there's a lot of movement in the tubes when moved/flexed... I may have inadvertently forced ink into the printer with the tubes clipped close to the reservoirs and the ink only able to go into the cartridges and subsequently the printhead. It's another variable but I'm leaving nothing out nowadays.

Once I'd run another clean, nozzle check and a high quality purge it seemed ok but on "normal" it started banding horribly so I cancelled immediately and swapped out to the OEM's again... This time it didn't work out... So the only changes were:
- tube flexing
- silicon grommets added
- an additional removal and insert of cartridges (wear on the rubber grommets)

*If the ink tube from the CIS is removed to vent the ink chamber, would the cart drip from its exit port as a normal cart would? If so, fill the ink chamber back to full, let the exit port drip until the ink level is down to your desired operating point to allow the sponge to become "normalized" with ink, then reconnect the supply tube.

To a large extent that's what I've been doing with since I learned about the pressure issue... but it's definitely worth noting and bearing in mind to normalise everything before attaching it all to a CIS.




Right, now to bring this all bang up to date I've reached the conclusion that the latest design approach and priming approach are pretty much the best I can hope for now. So starting from scratch I've done the following:

1. Removed the printhead from the printer that's "gone"... I figure I need to be giving each kit a proper try out and I feel a lot more confident about it now.

2. Punch out a new set of silicon grommets to be used to resolve the seal at the rubber grommet -> cartridge port.

3. Before inserting the grommets, wipe them to remove grease, etc.. and use a little stick back plastic to remove any hairs, dust, etc.. from them so there's no potential ink wicking effects or crap getting into the printhead to worry about.

4. The CIS kit itself has low profile fittings glued into place with an flexible epoxy that does a great job with the fittings over the spongeless "tank" part. The printer cases have to be "adapted" to remove obstructions but it works fine. The tubing is a little wider than I'd prefer but it's better than the small bore stuff of some of the PVC tubes which get brittle. The down side is that the tubing does indeed have a "flex" force to worry about but allowing an air buffer in the tank end seems to have a positive effect here too.

5. Put a "used" printhead in the printer and check the spacing, tube pathway, etc.. for the kit, removing, replacing, etc.. as required until it's "set" in terms of layout, etc.. thus reducing any wear on the grommets of the printhead that's going to be installed for actual use.

6. Install the printhead, silicon grommets (as per 3) and then install the kit as planned out.

7. Run a nozzle check, printhead alignment, high quality purge print and then a standard print...

8. For a long tube run, split the bonded tubing to where the tubing is mounted on the printer side and move any tubing clips into position there. This way when ANY movement, refilling, or manipulation of the reservoirs/tubing is required the pressure from the flexing involved can ONLY go in the direction of the reservoirs and thus not affect the printhead or cartridges.


Having just gone through this routine with only one hiccup (yellow cart chip wasn't seated properly requiring removal and fixing) I've found it to work a HELL of a lot better. Granted a new printhead makes life easier and ironically the problem isn't the magenta now but a small clog in the cyan which I should be able to resolve pretty quickly. The key thing however seems to be that removing the cartridges should be kept to an absolute minimum.




One thing that is evident throughout though is that the Magenta is definitely a weak spot in the whole process. I know the ink itself is not at fault as I've tested with OCP and Image Specialists solutions and found the same behaviour... it does seem that something in there, perhaps the nozzle position, design or similar is causing the heads to suffer. Cyan tends to go next with yellow having it's own peculiarities.

Perhaps the most telling thing in all of this was a comment Sam Cahir from RIHAC (who does CIS's in Australia) made about the printheads lifespan, intended or otherwise... It's actually not really that high and I suspect what I've been doing is un-naturally stressing the hell out of them with some poor designs.

I'll let you know how this approach works out in the long run anyway but for now this seems to be a little more successful... Be nice if I could understand why a little better I guess..

 

[websnail]

Dagnabbit... I was right in the middle of writing my opus piece on what works when I go and short out the PC by catching a USB cable end by mistake. Doh! and Grrrr!!

So, version #2 and here's hoping I haven't forgotten anything.

Silicon Grommets
- I've reached the conclusion that these are vital. The rubber grommets have a habit of conforming to the cartridge exit ports and after a bit will start to let in air, especially if you have a negative pressure in your CIS system.
- Following a small tip from Sam Cahir (Rihac CIS) it's best to clean your silicon grommets before installing them, using a lint free cloth to remove any grease, or ink followed by a little double sided sticky tape to remove any dust, hairs, etc...


Negative pressure is king
Obvious really but as HP's work differently I'll say it anyway, you need a certain amount of negative pressure in the CIS to ensure the printhead isn't flooded and to ensure you don't get overpressure


Kit setup
This has been mentioned a lot lately but it's now obvious that you need to prime your cartridges as if they were just being used refilled rather than as part of a CIS. To to this end I've found this works quite well for DIY kits.
1. Purge your empty cartridge and let it dry
2. Temporarily seal the air vent over the sponged area (you'll need to open it later so use electricians tape, etc.. to close it for now)
3. Drill out/remove the filler ball over the spongeless cartridge segment and clean up around it.
4. Glue/fix your tube fitting into the new hole and allow to dry as appropriate
5. Seal the exit port with electricians tape or use the orange covers that usually cover this port but make sure you get an air tight seal
6. Fill a suitable syringe* with approx' 20ml (40ml for PGI-5BK) of the appropriate ink
7. Attach the syringe to the tube fitting on the cartridge and then hold the syringe upright (tube end at bottom)
8. Pull gently on the syringe to draw some of the excess air into the syringe and then allow ink to replace it.
9. Repeat (8) until the spongeless area is around 4/5ths full
10. Allow the pressure to equalise in the syringe and then remove from the fitting on the cartridge
11. Attach a clamped bit of tubing to the cartridge tube fitting to stop air entering this section
12. Hold the cartridge over some tissue or a container and then remove the temporary seals on the air vent and ink exit port. This will allow the ink in the sponged section to equalise properly (avoiding the overpressure issue again!).
13. When the ink has stopped dripping seal the cartridge air vent permanently using a glue gun or something similar.
14. Re-seal the cartridge exit port using electricians tape again for storage

15. Repeat for the rest of the cartridges and then with the exit ports sealed (as per 14) remove the clamped bit of tubing


From here it's academic with the need to fill the tubes, attach them to the cartridges, etc... but the key stuff about equalising the pressure that seems to work best with virgin empties are the steps above.

*This assumes you have a syringe with a fitting that can connect to a small piece of tube and thus to the tube fitting on the port.
I use a luer lock 60ml syringe with a barb<->luer attachment.


Printhead health
One big mistake I made in the beginning was in thinking that printheads would "work themselves in"... They don't! If you get horizontal banding or anything that indicates ink starvation (or overpressure) you need to stop printing immediately and solve the problem. If you don't the printhead will never recover and the stress will kill the head pretty darned quickly.

Despite a lot of attempts to resolve printhead problems I've realised the best approach is to get the kit right and then install a new printhead. Use an old printhead to test it out first if you're not sure and work out the kinks with that so the new head doesn't get stressed out.


Printhead grommets
One thing in particular to watch out for is the black rubber grommets that sit in the printhead receiving bays. These grommets can become distorted or conform to specific cartridges quite quickly so removing and reinstalling a CIS cartridge set tends to stress these in very little time at all. As a result it's often best to test things out with an old/dead-ish printhead before you go with a brand new one.


Tubing
Wish I could say I've gotten this one completely licked... It seems the optimum tube inner diameter is around 1.5 - 2mm in a silicon based material. PVC tends to stress relatively quickly and can become brittle causing it to crack.


Fittings
Something to watch out for... Apparently white nylon fittings can and will expand over time when in contact with liquids. This can result in fittings splitting or cracking cartridges, brittle glues holding them in place, etc...


Reservoirs
I've been using bottles with dip tubes in for a long time but it's obvious that Canon printers in particular are very pressure sensitive. Whilst most CIS manufacturers have opted for the constant pressure (Mariotte bottle) type system I've found the dip bottles to work pretty well with the following proviso:
- The ink level must always be below the bottom of the printhead

Recently though I've started trying something new using bags of ink located in a low profile container underneath the printer. It's early days but if the negative pressure isn't too great this should reduce the ink level changes from 10cm to just 3cm complete with a system that doesn't require an air vent at all. News/Images on this will be coming once I've had a chance to get organised and all my bits have arrived.




So, with all of the above in mind I've been refitting a lot of printers with new kits that have been balanced... In a number of cases it's mean't replacing the printheads but in fairness the printheads have been lasting upwards of a year or more doing duty cycles far in excess of their norm' so it's to be expected.