I am officially an IDIOT!

[websnail]

At some point over the past couple of days I've been trying to get my head back around the whole "constant pressure" side of things and trying to figure out why my internal "mariotte bottle" model wasn't quite jelling with the information I'd been reading here, there and everywhere.

Well, today the penny dropped and I feel like a complete bl**dy moron of the first order of knights of the idiot table. In short... Doh!

I had understood the whole thing to be a sort of closed loop which somehow stopped gravity having any kind of effect on the liquid contained within. So in effect my head somehow had itself convinced that gravity was not involved and hey we have a special anti-gravity system here.


In the cold hard light of 5 million watt realisation I now understand what a complete buffoon I am..



What makes it all the more unforgiveable is that I have this ruddy title of inkjet master in my profile.. :/
Oh dear!

Anyhoo... the realistion has helped me come to terms with the further understanding that all the issues I've had with Canon printers is not going to be solved by buying in constant pressure reservoirs... Nope... Instead I can make my own systems very simply...

Instead of a simple dip system with a vent hold you have a dip system for supplying the ink and a sealed top that has another dip tube to a specific length which provides the same function as a vent but with the necessary pressure. So long as the top is sealed and the two dip tubes are also sealed in so that no air can enter through anything other than the "vent" tube then you're away.

Anyway... inspiration has some odd face smacking properties to it... but feel free to laugh... *cue: red face* :/

 

[Grandad35]

Websnail,

I'm not going to pretend that I understand how your new design works, but have you analyzed what happens when the atmospheric pressure changes by several percent? A severe storm can cause as much as a 5% change in barometric pressure, and even "normal" weather patterns can cause a 2% change in a few hours. Depending on the design details, this can cause a Canon cart to flood or starve.

 

[websnail]

Websnail,

I'm not going to pretend that I understand how your new design works, but have you analyzed what happens when the atmospheric pressure changes by several percent? A severe storm can cause as much as a 5% change in barometric pressure, and even "normal" weather patterns can cause a 2% change in a few hours. Depending on the design details, this can cause a Canon cart to flood or starve.

I guess the thing to point out at the moment is that until recently the amount of pressure in the system was directly related to the ink level in the reservoir and whilst it worked ok, it bore very little relationship to the pressure normally available in an OEM cartridge.

Until now I've eschewed the idea of using a constant pressure system completely but on reflection even with the problems of barometric pressure it's a lot more stable than allowing the actual ink level in the reservoir to determine the pressure reaching the cartridges and printhead.

I haven't actually carried out this change yet and I'm still getting my head around the implications...

 

[Grandad35]

On a standard Canon cart, the vent over the sponge chamber guarantees that changes in the atmospheric pressure do not affect the pressure differential across the print head. With a CIS, the cart's vent is sealed, and it is the responsibility of the CIS to guarantee that this balance is maintained.

Given that just a 2% change in barometric pressure gives the same pressure change as a 200 mm change in the height of the ink supply in a sealed system, it is important that the CIS be designed to cancel out barometric pressure changes in addition to ink level changes.

I know that Canonfodder did a lot of work on this when designing his CIS, and that it is not a trivial problem. I believe that his CIS works well for him after he figured out how to compensate for both the ink level and barometric pressure changes.

 

[websnail]

Given that just a 2% change in barometric pressure gives the same pressure change as a 200 mm change in the height of the ink supply in a sealed system, it is important that the CIS be designed to cancel out barometric pressure changes in addition to ink level changes.

I need to review my physics obviously but given that there is still a vent would it necessarily follow that this is a "sealed" system.

If the pressure changes then the air in the top of the kit will expand or contract certainly but because the vent is still there then surely it would allow the excess pressure to force ink up the vent tube or pull air into the reservoir as it compensates. At most I'd imagine considerably less than a 200mm change in pressure reaching the actual cartridges as a result.

Is my logic on this accurate?



EDIT: Sorry, realised I used the words "sealed top"...

What I was thinking of was pretty much a mariotte bottle design with a normal vent (ie: a hole in the lid) being replaced with a vent tube which provided the air input at a much lower height/depth in the bottle. Thus air would be allowed in but the lower vent hole would reduce the amount of pressure passed to the cartridges.

Does that make a little more sense now?

 

[canonfodder]

I do worry about atmospheric pressure changes. Perhaps too much. I am still running my "Spring Float" system which keeps constant ink level and is not affected by atmospheric pressure. See: http://www.nifty-stuff.com/forum/viewtopic.php?id=2498

WARNING ! The following is way too wordy for those who don't really care about CIS tanks and the atmosphere.

How these function:

The Chicken Waterer keeps a constant level of water in the dish. Each time the level drops enough, air bubbles are allowed into the bottle, reducing the negative air pressure at the inside of the bottle top, and allowing water to come out of the bottle neck until the water in the dish rises enough to block any more air from entering.

The Constant Level CIS Tank is similar to the Chicken Waterer, with the "BOTTLE" functioning like the bottle of the waterer, and the "DISH" functioning like a covered dish of the waterer. Although the "DISH" is covered, it is not sealed, but is vented to the atmosphere. When ink is consumed, the ink level drops enough to allow air to enter the "BOTTLE" and the reduced negative air pressure allows some ink to move into the "DISH" and again seal off the air path into the "BOTTLE". The Ink Output will deliver ink at a pressure that is dependent on the elevation of the point at which the ink is released to atmospheric pressure. If a tube were connected and its open end were raised to the same level as the "Constant Ink Level", the pressure at the open end would be zero, relative to the atmospheric pressure. If the open end is lowered, the pressure would be positive and ink would flow. If the tube end is raised above the "Constant Ink Level", a negative pressure would be present. In Canon printers using cartridges like the CLI-8, the correct elevation for the "Constant Ink Level" is equal to, or below the print head bottom elevation. This results in a zero to negative pressure at the print head, which is achieved with the original OEM cartridges.

Atmospheric Pressure Change Results:

If the atmospheric pressure drops, as often occurs during a local storm, the water level in the Chicken Waterer dish will rise. The reduced atmospheric pressure allows the air in the bottle top to expand, pushing some water out of the bottle and into the dish resulting in the water level in the dish being above the "constant level". Should the atmospheric pressure return to its original value, the air in the bottle top will contract, pulling the water back into the bottle. How much the water in the dish rises during the reduced atmospheric pressure is dependent upon the pressure change, the volume of air in the bottle top, and the relative cross-sectional area of the bottle and the dish.

The more air in the bottle top, (the less water in the bottle), the more a percent change in atmospheric pressure will change the water level in the dish.

The smaller the dish area relative to the bottle area, the more a percent change in atmospheric pressure will change the water level in the dish. In the Constant Level CIS Tank, the "DISH" area is smaller than the "BOTTLE" area, and that will cause an exaggeration of the ink level change during atmospheric pressure changes. Designs which have quite small cross-section area for the "DISH" function are the most sensitive and should probably be avoided.

If we observe what results from the reverse of the above stated pressure cycle, we see that an atmospheric pressure increase will compress the air in the top of the bottle, and water will enter the bottle until an air gap allows air to bubble into the bottle neck, and the water level in the dish is thus maintained. Then, with a subsequent drop in atmospheric pressure, the air in the top of the bottle will expand, releasing water from the bottle and causing the water in the dish to be higher than the constant level.

This Marriotte Bottle Variation can be used as a constant level ink supply. The vent tube serves the same function as the "DISH" of the Constant Level CIS Tank. The possible problem with atmospheric pressure changes is greatly exaggerated as the vent tube cross section is very small compared to that of the Bottle. With a pressure drop, the ink will climb up the vent tube, raising the effective ink level with it. It would help to make the vent tube diameter just as great as can be worked into the design. Even a scheme which enlarges the portion of the vent tube which is inside the bottle will help. A small tube passing through the stopper has no influence on the result.

 

[websnail]

Blimey... Ok... Now I feel educated

I see what you mean now about the air gap at the top of the "bottle" or reservoir and at a guess if you have a small tube instead of an enlarged one what you're saying is that you could end up with a mix of overpressure in the cartridge as well as a really impressive impression of a whale with ink out the vent.

Quite a lot to digest but thanks for the explanation.

 

[websnail]

Well, first off, my thanks to Grandad and Pharmacist for their input on this... I was able to utilise your input in a "tester" design and after a few hiccups I've managed to get my MP830 running again with 5 x 250ml bottles.

The design has taken on the Variation II suggested by Pharmacist with a 10mm diameter tube providing the "dish" functionality.

I further refined the design using some valves I have in rather large supply to provide a simpler vent on/off function but that requires some pictures which I don't have yet. Suffice to say, as a version #1 it seems to be working quite well.


My next adventure into the realms of Canon CIS will be to utilise a design that Epson use in their wide format cartridges and re-invest time in a "bag" design which has no vent at all. I have already had one of these working in a school for the last 9 months and had abandoned the concept due to problems with ink and the valves I'd originally sourced. Since then I've found a new valve which resolves all the issues I'd had and it's going to be time to return to the idea.

The bag idea works from an air pressure point of view because there's no longer a vent. This is no longer needed because the bag is completely sealed and can expand/contract with air pressure changes. The location of the bag (under the printer, in a 10cm high container also ensures that the pressure is negative although I do have some concern that the pressure may be on the overly negative side.

As I said, there's one working iP4300 which hasn't had any issues at all in 9 months (save a jammed valve) so it seems it's viable. The only other issue appears to have been the seal between grommet & cartridges and Sam was kind enough to send out a slew of slim (0.5mm) silicon grommets which seem to do the trick of resolving any potential leaks perfectly.


I still need to play with the physics some more as I have a few more hair brained ideas which my gut says shouldn't work but life's all about learning so I'm enjoying myself most of the time

 

[Froggy]

er...

My cat's breath smells like catfood.

 

[websnail]

????????