Anatomy of a Canon print head

[Defcon2k]

I'm looking forward to your new findings!

I can tell you that the ordering of the colors in your drawings goes cyan, magenta, yellow, black, magenta, cyan and then pigment black.

It seems that the position of the dye black is not important then, because it seems to differ from my iP5200:



Btw here are the nozzle setups for two other Canon printers, according to this (german) site:

iP4300 (=iP5200):
512 PGBK
512 photoBK (5pl)
512 yellow (5pl)
1024 cyan (1pl/5pl)
1024 magenta (1pl/5pl)

iP5300:
512 PGBK
512 photoBK (5pl)
512 yellow (5pl)
1024 cyan (1pl/2pl/5pl)
1024 magenta (1pl/2pl/5pl)

The first one I find understandable, but the second one (1024 nozzles with 1/2/5pl) sounds really odd. I think it's either not build symetrically, or it has 6 rows of nozzles per color (like this: Oo. .oO ) instead of the four rows per color discussed here with the iP4000 ( Oo oO ).


btw: I contemplated to make a test with my iP5200 by putting some tape over one side of the magenta and cyan nozzles and see how the nozzle check print would look like. But then I did not dare to do that, because it might damage the head.

 

[ghwellsjr]

btw: I contemplated to make a test with my iP5200 by putting some tape over one side of the magenta and cyan nozzles and see how the nozzle check print would look like. But then I did not dare to do that, because it might damage the head.

I guess I'm more daring because that's what I did. I put a piece of cellophane tape over the first two sets of nozzles (for cyan and magenta) and showed that my last schematic is the correct one. The cyan and magenta sets of nozzles have two different sizes. Thanks for prompting me to investigate this.

With regard to the print heads with three different sizes of nozzles, see this article:

http://www.usa.canon.com/consumer/c...y=plqindex&keycode=developers&fcategoryid=103

If I understand this correctly, they have doubled the number of nozzles over the IP4000. For the larger nozzles they just doubled the number, making the printing region twice as big. For the smaller nozzles, they interchanged 1 and 2 pl nozzles so there are 256 of each but they are twice as far apart as the large nozzles.

 

[Defcon2k]

I guess I'm more daring because that's what I did. I put a piece of cellophane tape over the first two sets of nozzles (for cyan and magenta) and showed that my last schematic is the correct one. The cyan and magenta sets of nozzles have two different sizes.

Wow, very nice! The canon article is also very informative.

May I ask how exactly the nozzle check print was looking when you did the cellophane test? Did you also try an extended nozzle check?
I guess the normal nozzle check looked quite normal, with cyan and magenta somewhat brighter*, and the extended nozzle check (in case you made one) was missing half of the patterns**, right?



*like this nozzle check where one row of the small cyan nozzles seems to be gone completely

** like this

 

[ghwellsjr]

I didn't even think to do a regular nozzle check because I knew it would be difficult to determine anything but now I wish I had. Anyway, here is a scan showing the extended nozzle check after I removed the tape and below that with the tape covering the one set of cyan and magenta nozzles:

 

[Defcon2k]

I didn't even think to do a regular nozzle check because I knew it would be difficult to determine anything but now I wish I had.

That's a pity for the sake of science, but perhaps a good thing for your print head

I think if you had one, we would see one of the following two possibilities:
(a) The nozzle check of magenta and cyan would look normal, but somewhat brighter, because every dot of magenta and cyan consists of only one ink drop instead of two.
(b) The nozzle check would not look brighter, but every other row would be missing. This would mean every visible dot would still consist of two drops of ink, but we can only see every second dot.
(Personally I guess that it is (a))

Anyway, here is a scan showing the extended nozzle check after I removed the tape and below that with the tape covering the one set of cyan and magenta nozzles:

Really interesting! May I ask which two sets of nozzles did you tape? I guess if your nozzle plate looks like this [cyan1, magenta1, yellow, black, magenta2, cyan2 and then pigment black] you taped cyan1 and magenta1, right?

 

[ghwellsjr]

I'm sure the regular nozzle check would be like (a) because the pattern presented earlier on this thread shows two sequential dots and then two missing dots for the first patch of the pattern. At the time, I thought each pair of these dots came from nozzles that were fairly close together, on opposite sides of the same ink feed but now I realize they were coming from about as far away as they could possibly be, almost on opposite sides of the print head. That would explain why there was so much irregularity in the placement of the dots as opposed to the more regular patterns put down by the black and yellow nozzles.

I did tape as you surmised, cyan1 and magenta1.

I think I'm going to repeat the experiment tonight and put the regular nozzle check on photo paper so I can get some good images of it later this week.

 

[ghwellsjr]

Here is the regular nozzle check. The top one is after the tape was removed with everything normal. The lower one is with the tape applied covering up the first two sets of nozzles as described earlier.

 

[Defcon2k]

Very interesting.
I wonder why the letters are completely yellow. I guess this is because they consist of yellow and the cyan and magenta nozzles which you taped. The other cyan and magenta nozzles seem not to be used for the letters.

 

[ghwellsjr]

The same kind of thing happened with the extended nozzle check posted earlier where only the cyan ink from one set of nozzles was used for some of the text.

And here are the magnified portions of the regular nozzle check focusing in on the magenta.

First the normal untaped image:

And the taped image:

 

[Defcon2k]

The same kind of thing happened with the extended nozzle check posted earlier where only the cyan ink from one set of nozzles was used for some of the text.

Oh yes, I overlooked that.

And here are the magnified portions of the regular nozzle check focusing in on the magenta.

Great, thanks for testing this!
I wonder why there are nearly no tiny dots scattered around on the "taped" print. On the first pic you can see mainly two large dots and a small one... on the second pic you can mostly see one large dot.