Canon CLI Cartridge

[mikling]

Has anyone ever tested to see if the prism still works on the chipped carts? For example. emptying a new cartridge and trying to run it out before the chip expires? I can't recall anyone doing that..... gutsy kind of move. Reason why? I have a suspicion that the prism system is actually not there.

B.C.I. Before Chip Installed.
C.L.I. Chip Later Installed.

Maybe Canon has been planning this for quite a while.

 

[jackson]

Has anyone ever tested to see if the prism still works on the chipped carts? For example. emptying a new cartridge and trying to run it out before the chip expires? I can't recall anyone doing that..... gutsy kind of move. Reason why? I have a suspicion that the prism system is actually not there.

B.C.I. Before Chip Installed.
C.L.I. Chip Later Installed.

Maybe Canon has been planning this for quite a while.

But if your bran new cart silently leaked it's heart out, then without the prism, how would you know the cart was empty?
Surely Canon would not tell fibs.

 

[ghwellsjr]

The printer will not report a bunch of empty carts. In fact, I have a set of empty carts which I put in my printer before putting in full carts to reset the ink level indicators, if necessary. I like to think of the prism as reflecting light when the cart is empty.

Seems contradictory.
Are you saying that if you put an empty cart in the printer it will not be reported as empty, but the reflected light from the prism indicates an empty cart?
I know that if you cover the base of a cart with black tape, the cart will not change status to 'empty'.
That seems to say that the light (when the tank is full) passes through the prism and only when the ink level changes does the light reflect through the prism surfaces back to a receiver.
And the Canon site refers to 'refracted' light not 'reflected'.
Whats the diff and when does this light turn on?

The Canon description is faulty, as noted by Grandad35, but the schematic is correct. When the tank has ink in it, the light is refracted. That means that it "bends" as it passes through the interface between two different media that propagate light at different speeds. When the tank is empty, the light is reflected and the sensor knows the tank is empty.

Another way to observe this is to simply look at the prism on a yellow tank. It will look dark if ink is covering the prism and light if it is not.

 

[Grandad35]

I like to think of the prism as reflecting light when the cart is empty.
.snip...
How do you like their animation of the cartridge being removed from the print head the wrong way?

This agrees with their schematic - light reflects back into the sensor when the prism is uncovered. I didn't even notice their animation - obviously done by someone who never changed a cart. These two inaccuracies show the lack of attention to detail by the people design and approve these pages.

I know that if you cover the base of a cart with black tape, the cart will not change status to 'empty'.
.when does this light turn on?

This agrees with the statement by ghwellsjr, and tells us that black tape can be used to disable the prism.

Let me give my guess on how this system works. If you pull a cart and look in the print head directly under where the prism would be, you will see a 3/8" square hole - this hole allows light to pass into each prism from a single fixed sensor. On my i9900, the sensor appears to be mounted about 3" from the left end of travel, and has two small "dots" about 1/8" apart on its top surface. My guess is that one of the "dots" is an LED and that the other is a sensor to sense the presence of reflected light. Before a sheet is printed my printer usually makes a slow traverse to the left, and I suspect that it is looking at each prism during this time. I have never had the printer declare a "low" cart at any time other than between prints.

When there is no cart, there is nothing to reflect the light, so the sensor thinks that there is a non-empty cart in place. It is only when light is reflected back to the sensor that it thinks that the cart is empty. BTW, this is another reason for keeping the printer's cover closed - in a brightly lit room, stray light could trigger the sensor during this sequence, giving a false empty signal.

Now, all we need is for someone who refills chipped Canon CLI carts to run some experiments.

edit: - Others posted while I was composing my message - sorry for any overlap.

 

[pebe]

The printer will not report a bunch of empty carts. In fact, I have a set of empty carts which I put in my printer before putting in full carts to reset the ink level indicators, if necessary. I like to think of the prism as reflecting light when the cart is empty.

Seems contradictory.
Are you saying that if you put an empty cart in the printer it will not be reported as empty, but the reflected light from the prism indicates an empty cart?
I know that if you cover the base of a cart with black tape, the cart will not change status to 'empty'.
That seems to say that the light (when the tank is full) passes through the prism and only when the ink level changes does the light reflect through the prism surfaces back to a receiver.
And the Canon site refers to 'refracted' light not 'reflected'.
Whats the diff and when does this light turn on?

Even Steve's site has got it wrong.

http://www.steves-digicams.com/2004_reviews/canon_ip4000_pg2.html

In fact, IR light shines up through the cart base into the prism. When there is ink present, the light beam hits the ink/plastic-case interface and because there is little difference between the refractive indices of ink and plastic, the beam is marginally refracted through the interface but still enters the ink and is dissipated in it.

When there is no ink preset, the beam cannot be refracted by the air/plasic interface because the difference of their indices is too large and the beam is reflected to the horizontal because the prism face set at 45deg. From there the beam hits the second prism face and is reflected downward into the IR detector.

So, ink present = no IR; no ink = IR detected

 

[websnail]

BTW, didn't websnail report that he was able to fit a CIS system to a Canon printer with chipped carts and that the printer never complained about the carts being low or empty? Was/is this true? If so, it supplies another interesting piece of information.

Bit late but just to correct this bit..

I actually found that the cartridge/printer could not be fooled into thinking it was full.. To recap.. the cartridges have 2 detection routines... the prism is still active and it has a "failsafe" in that the chip and printer will also track your usage via the chip.

Rational for this has been explained before but basically if you get an empty cartridge it knows.. If you don't refill it and attempt to continue when the "empty" message first pops up it will refuse to let you print no matter how many times you click "continue".. Refilling the cart with ink and lo it gives you the whole "accept your fate, mortal!" message and you do the 3 clicks and accept life without a warranty and off you go.

Anyway, the PG5Bk thing was an interesting development that I haven't been able to check out myself but I've heard of others trying it and finding it didn't work... so conflicting info on that score...


Other info on the actual chips though.. The encryption is one factor that makes the whole "saving a few bucks" theory a little moot.. A source of mine in the refill, CIS and aftermarket cart industry has been talking to a technician with a LOT of experience of various chips and they went hunting with the Canon chips and found they are multi layered with 32, 64, 128 bit levels of encryption and you only find out the next layer exists when you've cracked the first bit... So I think you'll have to excuse me a little on this if you figure that Canon are going to go to all that trouble and then leave out some tracking or writable functionality into their chips.

Course I would like to throw in the usual "I could be completely and utterly wrong" as a disclaimer...

Bottom line though I've found that creating a CIS for these units seems to be a hell of a lot better in terms of effort to benefit.. Not easy but so far it's worked very well for me...

 

[Grandad35]

I actually found that the cartridge/printer could not be fooled into thinking it was full.. To recap.. the cartridges have 2 detection routines... the prism is still active and it has a "failsafe" in that the chip and printer will also track your usage via the chip.

Just to be clear on this - if you took a new printer and new carts and immediately installed a CIS, the carts would still report as empty at some point. You would then get the nag screens, accept your fate and print as long as you kept the CIS topped off. Is it also true that you must have at least one cart in the printer that hasn't been reported as empty (the "PG5Bk thing")?

When you talked to your source, did they say what type of hardware was used in the chip? Does it contain some form of PLD and/or rewritable memory?

 

[websnail]

Just to be clear on this - if you took a new printer and new carts and immediately installed a CIS, the carts would still report as empty at some point.

Correct... This has been confirmed on 2 x IP4200's and an MP500 which were all setup the same and all the cartridges still held ink in more than half of the non-sponge area.

You would then get the nag screens, accept your fate and print as long as you kept the CIS topped off.

So long as you have ink in the system it works.. I have not experienced (and I'm not about to risk it deliberately - science is all well and good but ) a situation where one of the tanks has run out (bad seal, etc..) to see what happens when the tank runs out.. but based on guess work I think Canon have allowed the printer to keep on printing in this scenario.. a burnt out printhead being a neat way to punish refillers... but admittedly that's just a guess

Is it also true that you must have at least one cart in the printer that hasn't been reported as empty (the "PG5Bk thing")?

So far I've been unable to confirm this... I have one IP4200 which has been empty across all cartridges for over 3 months now (need to check that date but it's about right) and it still prints quite happily.. I've heard of no more than just one report of the printer refusing to print if all are showing empty so right now the jury is out on that one.

When you talked to your source, did they say what type of hardware was used in the chip? Does it contain some form of PLD and/or rewritable memory?

They didn't because quite frankly that sort of info would have gone clean over my head but I'll see if I can find out more as I need to chase something else with them anyway.

What was said in no uncertain terms though is that any kind of 3rd party chip was virtually impossible for economic and also litigatory reasons. Canon have let it be known that any chip cracking will bring a swift lawsuit. On the up side a formal complaint has been lodged in the EU on the basis that printer/consumer protection is all well and good but why does it require more encryption support than nuclear missiles?!

 

[mikling]

"all well and good but why does it require more encryption support than nuclear missiles?"
It's so darn cheap to do it for Canon, so why not? Probably cost them about $.05 - $0.10 per cartridge.

 

[pebe]

....... Canon have let it be known that any chip cracking will bring a swift lawsuit. On the up side a formal complaint has been lodged in the EU on the basis that printer/consumer protection is all well and good but why does it require more encryption support than nuclear missiles?!

I wouldn't be surprised if Canon have left the back door open for a modification to their printers if legislation in Europe or the US forces them to backtrack on their chipped carts. This might take the form of a device retrofitted in the cable feed to the heads, and it might be quite a cheap and simple modification.

I wonder if the threat of legal action is a bluff by Canon to prevent people from probing too deeply.