Clogged Canon print head

[websnail]

Exactly the benefit of this forum, where participants are willing to show their failures as well as successes.

Aye.. where some of us do failure with flair and panache, and not a little crying

 

[Trigger 37]

The sad news continues on this damaged printhead. The leaking Cyan ruined by Magenta ink cart so I knew that any ink cart I put in there would be contaminated. But to print it had to have an ink cart so I put in an old dead yellow cart that had no ink in the reservoir but show some signs of ink in the sponge. The leaking before was limited to the Cyan into the Magenta but as I check today, their was no leading ink to the dummy ink cart I put into the Magenat slot. This points out to me that the wicking was a combination of Cyan ink flow and Magenta ink flow meating at the ceramic wafter. Once the two streams were mixed there was a path back up to the Magenta ink cart. With the dummy ink cart in there, there was no path for the ink to flow back to the cart. This helps to understand how the Cyan got back up to the magenta ink cart.

The bad news is that as this all started the nozzle check showed excellent black, satisfactory Cyan, Cyan/Magneta, and excellent yellow. Today I discovered that the Cyan has now leaked over to the Yellow and totally contaminated that inkcart as well. I've removed all ink carts and the printhead and replace everything with working hardware. I flushed the printhead and set it aside with the capped ink carts for future cleaning.

fotofreek, Websnail,... I must admit I was so anxious about the results that Dralsop had with his boiling water, I jumped in head first to try it. This is not the first hastly thing I've done and probably won't be the last. I'm just greatful that I was able to stop others from going ahead before we had some confirmation that this was a solid solution to EXTREMELY CLOGGED HEADS.

Neil Slade has commented in one of his posts the he believe individual nozzles and sections of nozzles are damaged by bad ink carts that either run dry or can't supply the ink to keep the head cool. When we continue to print when the ink is very low, the nozzles are going to get hot and this makes things even worse as it tends to dry and then burn the tiny amont of ink that is there. The question is, once the ink is burned on to a section of a nozzle, can it ever be cleaned off. Would deep cleaning have any effect at all. It is sometthing like the beef stew my wife made the other day. She left it unattended for too long and burnt the soup to the bottom of the pan. It was soaked for 1 full day in soapy water then scrubbed with an abrasive pad and soap. Finally to get it clean I had to use SOS pads which is basically steel wool. Because the pan was stainless steel there was no permanent damage. I wonder what dried burnt ink looks like on a 2 picoliter nozzle.

Do you suppose this is why Canon is sp paranoid about warning if the ink in a cart is running low. Is this the reason they designed the ink cart to have a reservoir on one side and a sponge holding area on the other side, such that when the reservoir went dry there was still sufficient ink left to protect the printhead from burning up?

 

[mikling]

Trigger there is also another seal which is more likely to be damaged than the rubber seal. I suspect this is where the leak is. That seal in between the nozzle plate and ceramic body. On older HP cartridges, the nozzle plate coming off was a common occurrence on refilled cartridges. Basically you have grooves that correspond to each color ( 6 color head shown)

The nozzle plate shown broken and inverted simply sits on the white ceramic and looks to be held and sealed by adhesive. If this adhesive was not perfectly applied or the plate is slightly distorted, you are going to get a weak bond. If you get a paper jam stressing this weak bond, you might initiate leaking.

Severe rapid temperature increases where the plate expands faster than the ceramic material will create shear stresses which will weaken the adhesive bond. My suspicion is that this was the mode of failure using the boiling water treatment Each material by itself could possibly withstand the temperature of the water but the differential expansion might be the culprit. Higher temperature water will aid in dissolving deposits but this must be introduced gradually where everything slowly heats up. Warm water probably would have been safer.

 

[fotofreek]

Does the elasticity or lack of elasticity of the sealing material come to play?

 

[canonfodder]

................. Is this the reason they designed the ink cart to have a reservoir on one side and a sponge holding area on the other side, such that when the reservoir went dry there was still sufficient ink left to protect the printhead from burning up?

Trigger 37, Only a minor detail in this discussion, but - No, we know that the sponges have a distinct function in controlling the ink as fed to the printhead. The capillary action of the sponge material pulls up on the ink, and this is what produces the negative pressure that the printhead needs to control ink properly. I think we can assume that protection of the printhead is only a side benefit.

 

[mikling]

Elasticity will come into play however, many adhesives are strong on tension but weak in shear and torsion. That is why you pull to the side to remove glued objects but never directly pull. Superglue is a classic representation of this. When there is a difference in expansion rates a gasket of some type may be preferred. judging by the few failures, I would think this is a non-issue in general.

 

[Trigger 37]

mikling,.. Thanks for the detailed photos. The wells or holding tubs above each bank of nozzles provides a lot of information about how printheads can/do clog. From the size of them the could hold something less the 1ml of ink. I'm enclined to believe you are correct that the bond between the nozzle plate and the ceramic plate was lost. In my years working in development of hard disk drives for IBM, the most difficult problem to always solve was the difference in the coieffenct of expansion of dissimilar materials due to heating or cooling. Basically you can get a lot of things to mate and stay together at one given temperature, but change it by 5 degrees and they will separate because of expansion or contraction. We are talking about microinches or nanos.

We used to say that the most difficult job was trying to fly a 747 airplane at 1 foot off the ground at mach-2, and keeping it from crashing. This is the problem we had to solve in order to keep a disk head flying over the surface of a disk. Change the temperature 5 degrees and the airplane wing takes a different shape,... crash.

Back to reality,.. To totally contaminate the good yellow ink cart, I agree that the nozzle plate and the ceramic tub must have come separated and the inks are running everywhere. The proof will be when I take it apart and check to see if there was some kind of leak at the rubber gasket.

This has also opened up many other possibilities on how inks can appear contaminated in prints. Think about a printhead full of ink sitting at the parking station on top of a saturated suction pad. This is why the purge unit goes through so many cycles before parking. One of these is to move the head off and cycle the purge unit to suck any extra ink off of the pads and dry them off. Then the blades wipe the bottom of the head before they finally bring the pads back up to mate with the printhead. If the color suction tube is blocked the pad will stay saturated from just the wicking ink flow out of the printhead. Then again, if it is blocked, it can't even suck ink into the printhead.

 

[Grandad35]

I ran across this post and thought that it is worth a try if you encounter the dreaded "missing half" nozzle check pattern. I have been assuming that this problem is caused by an electrical failure, but this post indicates that it may be a (repairable) programming glitch in the print head's EEPROM.

http://www.stevesforums.com/forums/view_topic.php?id=586754&forum_id=40

 

[fotofreek]

Grandad - is it possible that Canon programs in an eeprom failure? More probably the usual software glitch that we experience in so many areas of the computer world. Many of them require a reboot or re-installation of the software to cure the problem. Eqivalent to "kicking" a mechanical device that is malfunctioning and seeing it come to life again!

 

[Nifty]

I love this part:

For some reason I decided to do an EEPROM initialization. This clears all the counters back to zero. Magically the head started working perfectly again. Not a single nozzle clogged!

If only every stubbor "clog" was this easy to fix!