The Internals of the Canon Pigment Ink Cartridge PGI-9

[Grandad35]

The other detail that I am seeing are the two small rectangular cutouts that appear to have no purpose, except that I have seen something similar internally inside the printheads before. It helps establish a ribbon of ink across the pad/mesh layer.

Could the explanation be that these cutouts allow faster cooling of this thick section when the cart is injection molded? Note that there are several other similar cutouts on one end of the cart - probably for a similar purpose. This link explains why having a uniform part thickness is important to prevent warping, twisting and cracking.

This will fully explain why on refilling we can get higher weights than when the cartridge is new.....because the diaphragm/bladder is already partially collapsed a bit when new. When we refill, we can get the diaphragm to fully extend because we do not need the step of air withdrawal that was intially required. Thus we can get higher weights and hold more ink .

Are you sure that there isn't at least a little air left in the folds at the top of the bladder? If the bladder is pushed completely out to the sides and there is any air expansion at all due to barometric pressure/temperature changes, ink might be forced out. In addition, won't you will lose the mixing action of the paddles when there is no lateral movement in the bladder?

Great photos and description.

 

[mikling]

When empty according to the chip there is still some lateral movement, This is where the benefit of a refill kicks in actually. By having the bladder overfilled over the original, it means a larger volume as compared to the original and thus more lateral movement. Hence, more mixing. Thus it is better to have the cartridges filled as much as possible versus underfilling. This shows why it is possibly superior to have the cartridge filled by a visual indicator and the limit to handling determined by handling itself. Superior mixing throughout the usage cycle is attained. If you underfill, then the point of no mixing towards the end is accentuated.

About the barometric pressure issue, this would only be the case if air is trapped inside and if so, the effect would vary depending on the volume trapped. The volume changes would be accommodated by the spring. Remember the sides are pulled back in a little during the evacuation of the air at the time of assembly. So there remains some spring travel to accommodate the possibility of increased volume during expansion of the bladder volume. My instructions to squeeze the sides of the cartridge at the end performs a similar action. We regain some spring travel to accommodate temp changes and we also remove the ink that would drip out if the bladder was fully filled and then squeezed during handling.

About the temperature issues due to injection molding? We'll never know. Fortunately it doesn't matter in the refill process but it is a possibility.

 

[ThrillaMozilla]

Nice pictures. You obviously have this very clearly figured out. Thanks. An elegant design.

My rough measurements indicate around 2-3 inches of water actually.

The vacuum can go to zero, or you can even have positive pressure, if the piston is at the limit of its travel. If you have even a little air and the piston is at the limit, I would think there's some danger. The more air there is, the greater the chance of having positive pressure. That's why it's so important to give it a little squeeze after refilling, as you pointed out.

If it were mine, though, I would also stop a little before filling to the tippy-top -- maybe 31 g instead of 32 g -- to ensure that you always have a vacuum. Would that be a good idea? I don't know if that's really necessary, but for what it's worth, it does apparently correspond to Canon's fill level. Would that be a good idea? I guess one possible disadvantage is that you couldn't get any air out with a squeeze.

 

[mikling]

I'd take the extra capacity, for safety margin of empty and better mixing as I've pointed out. If you're worried about not having enough vaccuum. you could squeeze harder and get a little more out.

What might have been missed is that the onset of vacuum does not set up linearly. It is a non linear function when it goes from being against the wall to having no contact with the wall. It goes from zero or positive to a negative value immediately when there is a gap between the wall and the diaphragm because the spring is already compressed quite a bit.

Additonally, if you squeezed it gently, when it is placed into the printer, the first action of the printer is to prime the printhead and on the 9500, this takes some time. This priming will withdraw ink and definitely set up the negative pressure for you. So there really is no need to place to much concern about setting the pressure up. If it doesn't drip during handling, the printer will handle the balance for you. Couldn't be simpler.

 

[ThrillaMozilla]

...the first action of the printer is to prime the printhead.... This priming will withdraw ink and definitely set up the negative pressure for you.

There you go. This doesn't work for sponge cartridges, but it should work for this one, if the printer always does that.* You've obviously thought this out carefully.

==================
* I have an HP that does not purge if I change only one cartridge. It's a good trick. I don't know if it will work with the 9500. On second thought, it's probably not a good idea on that printer.

 

[rodbam]

Thanks for doing this Mike it takes a lot of the mystery away from these carts. There's two things I'm not understanding, the inlet into the head is a lot smaller than the ink outlet pad on the cart as though it only takes ink from the middle of the pad & the other is why the 9500 has a function for conditioning (mixing) the ink. I would have thought that the carts whizzing back & forth on the head while printing would have mixed the ink.
I'm surprised how well the pigment inks performed compared to my dye inks, I've just done my first test print with the 9500 to compare with the 9000 output & the pigment seemed to pop just as well as the 9000.

 

[The Hat]

I am glad to see your enjoying your wifes new printer and I do hope it works well for you and therefore dont get nagged about it.

The quality output on the 9500 will mostly be down to the paper you use it on,
while on some papers the image as you say will pop but on others it can be an absolute disaster,
its something to watch out for.

Nice test prints by the way..

 

[mikling]

Rod, the pics are not the same so a valid comparison given the conditions is very rough but it is quite evident that shadow detail is superior on the 9500.

Now why do we need the paddles. You will answer that question yourself if you do this at home. Take a bottle. Fill it completely full with water with absolutely no air inside. Close the lid. Now shake the bottle. Did the water inside move and slosh around? Well, that is exactly why there are inertia paddles in the cartridge. The question posed is indicative that you still think that there is air inside when the cartridge is working. No there isn't and there shouldn't be and whatever you do, minimize the potential of doing so. The printhead literally sucks out the ink from the cartridge and the spring is providing opposition to this.

Obviously the worst way to refill these is by using a plug and drilled holes. Why? because it simply keeps plopping more sediment onto the mesh screen during use and there is no action to clear the screen and channels which is going to build up some sediment over time and eventually stop ink from flowing through. The only chance you get to do something about this settlement is during refill via a backflush over the entire screen. Use that opportunity.

About the printhead. Honestly, I don't have time to delve into that one as I committed to getting progress on my website. Others can venture into that. It might even be more complicated.

 

[rodbam]

Thanks Hat the prints should be OK on the 9500 for a while because of the posh papers they gave away with the printer. I will do another test print on my usual cheap paper & see how that goes.
Mike that makes sense about the paddles, what doesn't make sense is such a well engineered cart being for one use only although we make use of it a I was going to say your web site needed attention so get to it mate.

 

[websnail]

Not had the time to delve into this myself but glad I read through this thread and caught up today.

Fascinating images and design that certainly show a well engineered solution from Canon and make you wonder just how well, if at all, the third party sponge alternatives work. Frankly, I shudder just thinking about it.

Not entirely sure I agree with the third hole acting as a vent hole for the filling process as noted in the flushing thread, the air would need to be in exactly the right place for it to exit but in truth the filling process will be a mystery unless someone goes all Mission Impossible at some point . Either way it's a moot point.

The only thing I'd add is how the bellows compress when air/ink has been evacuated to show how the design effectively blocks off the fill holes. I'm prepared to be proven wrong but I still suspect there's an element of design here that is less reliant on the sponge moisture content than has been postulated. I don't have the engineering vocabulary to explain it but take a look at this image:
http://i2.photobucket.com/albums/y20/mikord/IMG_2697.jpg
... and note the "shelter" style design around those outlet holes.

If you look at the way the bag collapses onto this shelter it seems to form a sort of closed gate using the "shelter" as a analogue of the gate posts/frame. When you couple this with the way the bag folds inwards to rest against the inside walls, thus creating a resistance against the spring and re-inflation as well it looks as if the design was deliberately intended to stop an empty cartridge appearing full.

Stepping back from the mechanics a second, there's a good reason for wanting to do this (and not because it hands refillers a golden ticket of non-air ingestion). A cartridge that re-inflates/expands would potentially fool a user into thinking the cartridges was still full/useable so retaining the plate in the "empty" (far-wall) position acts as a visual indicator of being empty for the user. Granted you have the chip but as with the chipped sponge carts (eg: CLI-526's) you still have the prism backing up the chip so redundancy appears to be courted in Canon designs.

As I said, willing to be proven wrong but it does seem more than a little odd to have that shelter design over the outlet holes otherwise.