Ink usage on a Canon i9900

[Grandad35]

I have been keeping a detailed record of the number of pages printed and when carts are changed for the last 10 weeks. The number of prints is recorded as the "Number of 8x10s" by counting three 4x6 prints (printed three to an 8.5x11 sheet) as one 8x10 (72 sq in vs. 80 sq in) and one 12x18 print as two 8x10s (216/2=108 sq in vs. 80 sq in). The 4x6s are 10% under the print coverage of an 8x10, but the 12x18s are 35% over. Given my average usage of these sizes, the unders balance out the overs fairly well, so the "8x10 equivalent" is fairly close.

Even though I refill, the printer is always run until it reports a cart as "empty" and refuses to print any more. The point at which any cart is reported as "Low" is also noted, but only to log the point when this occurs. With a large supply of spare (already refilled) carts, an empty cart just requires a simple cart swap. The carts are never inspected between changes, relying on the printer to indicate when one of the ink chambers is empty (a "Low" warning). However, when any cart is replaced, every other cart is inspected and replaced if there is less than about 3 mm of ink remaining in the ink chamber. The "Low" carts are replaced at this time because an extended cleaning cycle is run when any cart is replaced (the printer sees that the ink chamber of a "Low" cart now has ink). This cleaning cycle uses a lot of ink and can cause previously non-empty carts to become empty before anything is even printed.

During the last refill cycle a total of 23 carts were refilled and the carts were weighed before and after refilling to get an accurate idea of the ink usage.

A summary:
1. The carts are always filled as full as possible in both the sponge and ink chambers. Some people say this can cause ink delivery problems, but I haven't seen any problems from this practice.
2. The refilled carts took (on average) 8.7 grams/cart to refill. Since the ink's density is about 1 gr/CC, this is about 8.7 CCs/cart.
3. Two carts were completely purged and dried before being filled, and they averaged 16.5 CCs of ink/cart. This indicates that up to 8 CCs of ink can remain in an "Empty" cart, but this value would be lower if the sponge wasn't completely full on a prefilled cart (OEM and some 3rd party prefilled carts).
4. The "average" ink usage is very close to one cart for every twelve 8x10 prints. At the current $12 list price for OEM BCI-6 carts, this give an ink cost of about $1.00 for each 8x10 print with OEM inks. The ink cost is about $0.06 for the same quantity of 3rd party bulk ink and $0.20 for 3rd party prefilled carts.
5. Using 9 CCs/cart, this gives an ink usage of 0.75 CCs/(8x10), or about 0.6 CCs/(8x10) if you deduct a 20% ink loss for cleaning cycles (startup./shutdown/cart changes/etc.). The 20% value wasn't measured and is only an estimate based on what I have seen elsewhere for a printer with a high usage and no extra cleaning cycles.
6. The red, green and black usage was so low that it is impossible to know whether any ink was actually used for anything other than cleaning cycles.
7. The usage reported here is based on my prints, and will vary if your prints tend to be lighter (I tend to edit my photos a little darker than most) or of lighter subjects (prints of children playing against a background of grass/trees tend to use a lot of color). It is also probable that other inks will require either more or less ink, depending on their coverage.
8. The following are the approximate ink usage on an i9900 8 color printer in CCs per 8x10 high quality photo print for each ink (these values include the losses during cleaning cycles):
PM 0.23
M 0.064
PC 0.17
C 0.056
Y 0.13
K 0.033
G 0.033
R 0.033
As a point of interest, a 2 pl printer using 0.23 CCs of PM ink will generate 115,000,000 individual PM dots of ink on a typial 8x10 print - enough to print a dot every 30 microns in both direcions.
9. The following are the average number of 8x10 sheets that were printed after a "Low" signal and before the "Empty" signal was given. There are no good values for the inks with a low usage rate - they were always replaced before they ran empty because one of the other colors (usually PM or PC) ran "Empty" first. The values for the PM, PC and Y aren't consistent with the rate of ink usage.
PM 3
M >15
PC 8
C >15
Y 6
K >15
G >15
R >15
10. There was one time period when the printer was unused for 2 weeks (it was turned off). There was a long cleaning cycle when it was turned back on, but there was no sign of any printing problem even after 2 weeks of idle time. In fact, there has not yet been the need to run any extra cleaning cycles for any reason since switching inks, Over 40 carts have been changed during this time and a total of almost 500 8x10 sheets printed.

 

[Nifty]

Grandad, again, a VERY excellent post. Personally, these are are some of the experiments that I enjoy the most and really appreciate the time and effort you put into this.

This cleaning cycle uses a lot of ink and can cause previously non-empty carts to become empty before anything is even printed.

Very true! The first time I used my new MP750 I put the carts from my i860 into it. A few of the carts had about 1/8 inch of ink in the reservoirs and after a VERY lengthy first cleaning / priming cycle two of my carts went from ok to empty. That would be a pretty expensive experience had it been OEM ink!

This indicates that up to 8 CCs of ink can remain in an "Empty" cart,

This is an excellent point to why (1) Using non OEM cartridges & refilling is a great idea (due to the fact that one would be leaving $$ in the OEM cart in the form of unused ink) and (2) why refilling can be provide even more savings when compared to 3rd party cartridges without refilling.

Thanks for all the great data! I'll have to revisit my "Inkjet Printer Total Cost of Ownership" worksheet to incorporate some of this information. http://www.nifty-stuff.com/inkjet-printer-tco.php

 

[JV]

Grandad,

I calculated the ink usage and cost for my iP8500 printer with BIJC#2, to compare with your values in Post#1. The comparison is reasonably close considering the many plain sheets (text and some graphics) I printed and allowing for the different colors we might use. I assumed that ten plain sheets use as much ink as one photo print.

See the BulkInkJetCarts Topic, Post #34 for my results.

JV

 

[Grandad35]

JV,

I agree that both sets of results are amazingly close. I averaged 8.7 CC/refill and you averaged 8.9 CC/new cart! (http://www.nifty-stuff.com/forum/viewtopic.php?pid=3139#p3139). I thought that my prints were darker than most, and your numbers confirm that I use about 15% more ink than you (0.75 vs. 0.66 CC/print). The biggest difference was in PM usage, where I got 0.23 CC/print and you got 0.156. Even this difference could easily be due to the types of photos that we print - my prints have a lot of flesh-tones, which use a lot of PM. I wasn't nearly as meticulous as you and I didnt weigh each individual cart, but my PM usage varied from 33 to 46 prints/cart, or 0.19 to 0.26 CC/print (assuming that every cart supplied the same amount of ink), so it is clear that there is a lot of variation in these numbers due to what is being printed, even on the same printer.

I particularly liked your graph of the ink in the sponge chamber vs. the ink chamber. Your data clearly shows that the ink in the sponge chamber supplies all of the initial ink, which initially lead me to think that the printer was using very little ink when it was new. I am of the opinion that the ink level in the sponge chamber drops until the top of the vertical grooves in the back wall of the sponge chamber are uncovered, allowing air to enter the ink chamber, allowing ink from the ink chamber to flow into the sponge chamber.

Your value of 0.6 CC between the "Low" and "Empty" signals fits my value of 3 prints on PM almost perfectly. Your data also confirms that there is a lot of ink left in the carts when they are declared to be "empty".

Since we now have a good handle on the ink consumption on an 8 color Canon printer, it would be nice to get similar data on a 4 color printer to see if the PM colors use as much extra ink as was predicted.

 

[Kenyada]

Even though I refill, the printer is always run until it reports a cart as "empty" and refuses to print any more. The point at which any cart is reported as "Low" is also noted, but only to log the point when this occurs.

Another excellent post, Grandad, and a great reference for all i9900 owners, as well as everyone else. Have you ever noticed any problems running your printer until "it refuses to print anymore"?

 

[Grandad35]

Richard,

It is conventional wisdom that allowing the ink chamber to run empty will cause a buildup of "glop" in the sponge and cause ink starvation problems. I saw evidence of this glop buildup and had ink starvation problems with my previous bulk ink (http://www.nifty-stuff.com/forum/viewtopic.php?pid=1676#p1676). When I changed ink suppliers I also performed a purge procedure on every cart at the same time (http://www.nifty-stuff.com/forum/viewtopic.php?id=311). Part of this procedure is to make sure that the sponge isn't wedged against the back wall of the sponge chamber, and I think that this additional step has (at least so far) helped to eliminate the glop blockage problem. The change in inks may also be at least partially responsible.

In any case, I have refilled many of these carts that have been "run to empty" 3 times, and (at least so far) always get good ink delivery. In fact, the exit port is always wet when a prefilled cart is removed from its storage clip, and there are generally 5-10 drops of ink that drip out without even blowing into the exit port. There was a risk that all of my carts might be blocked up by running to empty, but having a simple purge procedure capable of rescuing them made it worth the risk to test running them to empty.

If the sponges in your carts tend to block up with glop, I would refill them when the ink chamber is still 1/3 full (conventional wisdom) to minimize any glop buildup. If your sponges don't tend to block up, you might want to try running to empty, knowing that your carts can be rescued if they block up. Printing is much simpler, as you never have to open the cover to look at the ink levels - just print until the printer won't print any more. As long as your "low ink" sensor works, there is no danger of harming your print head since there are still several CCs of ink left in the sponge when the printer refuses to print.

If you buy prefilled carts (OEM or 3rd party) there is no reason not to run to empty, and it can save money by getting more prints from your carts.

 

[billrogers]

Excellent work. I had independently measured my cost to be about 1.3 cents per square inch for photo quality printing with my Canon i9900. I use Canon cartridges and change them when the printer tells me that they are empty.

For an 8x10 print, 80 square inches, the cost works out to be $1.04 per sheet.

I'm a photographer, and I sell prints, so it's important that I am able to assure my customers that I use genuine Canon ink.

But Canon should give a coupon with each ink tank they sell - after you save a certain number of coupons, they should just give you a new printer. They have to be making a fortune on the ink.