I wasn't satisfied with the finicky results from the standard bottle-and-tube bulk ink system on several Epsons, so I did a lot of testing and thinking:
http://www.polyphoto.com/tutorials/bulkinksystem/ExtensiveThoughtsOnCIS.html


I should add, I have built and sold about a dozen CIS. I do not have an online store selling them, but am considering it.

AlienSteve wrote:

I wasn't satisfied with the finicky results from the standard bottle-and-tube bulk ink system on several Epsons, so I did a lot of testing and thinking:
http://www.polyphoto.com/tutorials/bulkinksystem/ExtensiveThoughtsOnCIS.html


I should add, I have built and sold about a dozen CIS. I do not have an online store selling them, but am considering it.

Hi Steve,

I have actually come across your page before but thanks for the link... I figured it was worthy of discussion on its own because the issue of ink pressure is one that I'm sure crops up from time to time. Especially the "why?"

For my own personal experience I've found that the problem with the constant pressure systems (that are available commercially) is that they are poorly made and suffer from a number of reported leaks, cross-contamination between tanks and poorly written instructions for refilling.

As much else I'm ready to test some solutions but as most CIS kits are sold as complete kits with no "options" on reservoirs, etc.. it's rather difficult to take them for their individual merits.. That and a complete lack of branding... No name manufacturing seems to be the norm'

Care to take on some testing? and compare results? and/or share any information you've developed if you'd looked at the commercial offerings..


_________________
Printers: (Canon) MP500/830, MX700, iP4000/4200/4300/4500/5200, iX4000(A3) (Epson) C84/86, D88, CX6600, R285/800/1900 (HP) K550, K850, K5400, L7680

I too have been pondering the "problem". Basically a bottle with a feed tube taken out through the top with a small hole to let air in (or out) is a poor solution because the pressure at the cartridge outlet port varies considerably as ink is used up in the reservoir bottle. Additionally, the top with its feed tube has to be removed to refill the reservoir, risking getting air into the feed tube. This is no problem as long as the amounts are small, but could cause difficulties if the same happens every time.

There appear to be two simple requirements that would solve these disadvantages.

Firstly, the ink feed should come from the bottom of the reservoir - or at least not be disturbed when the top is removed for refilling, but still taken from near the bottom of the reservoir (this is just common sense!).

Secondly, and this is important for maintaining a constant pressure at the print nozzles, the air inlet should be taken by a tube to just above the bottom of the reservoir (the top is sealed so that the only air inlet is via this small tube) but higher than the ink feed tube. Initially the hydrostatic pressure on the ink nozzles will be equal (well proportional to) to the vertical distance between the ink surface and the ink nozzles. As the printer is used, and ink is drawn from the reservoir into the cartridge, the level in the tube will fall until air (to replace the used ink) bubbles from the tube into the space above the ink. Now the hydrostatic pressure remains constant more or less regardless of the ink level until this falls to the level of the bottom of the air inlet tube. The volume of the tube should be sufficient to accomodate likely changes in barometric pressure without ink spilling out of the inlet hole when a depression occurs. I had thought of using a small inverted syringe (without the plunger), with the luer fitting fixed into the removeable top of the reservoir.

So, if the bottom of the reservoir air inlet tubes are arranged to be more or less level with the nozzles - bingo. more or less constant zero pressure, and no tendency for ink to flow out of, or air into, the cartridge when it is parked and idle. Moreover this will be maintained whilst most of the ink in the reservoir is used.

This scheme is not my invention, but appears on my Everest CIS system, where it works very well. However, if this is built into a semi flexible reservoir (e.g. a "Nalgene" bottle), then squeezing the bottle will result in a nice fountain of ink. Enough said! The Everest system has rigid reservoirs, so this is not an issue.

Need to find a neat and leakproof way of fixing a barb fitting into the bottom of a "Nalgene" bottle, and to provide a simple stand which protects the ink tibes and fittings from damage

Already done.

http://www.polyphoto.com/tutorials/bulkinksystem/HydraulicConstantPressureExample02Web.jpg

Where the ink feed comes from has no affect at all on hydrostatic pressure. It could come straight out of the top and go 3 feet in the air and back down again, or come right out the bottom and drop 3 feet below the printer. All that matters is the ink feed level is set to the same level as between the bottom of the head and the base of the ink spike.

http://www.polyphoto.com/tutorials/bulkinksystem/1280/WhiteTanks3QuarterView4096w.jpg

Refilling the system:
http://www.polyphoto.com/tutorials/bulkinksystem/1280/HCPCISInstallationPg09.jpg
http://www.polyphoto.com/tutorials/bulkinksystem/1280/HCPCISInstallationPg10.jpg

I've redesigned to be castable and more easily fillable. Just need to make the molds.

I really, really wish I'd gotten into all this many years ago. Someone just gave me an old Encad Novajet II that had a 3rd party bulk ink system added. I looked it up, that bulk ink system sold for around $2000 or more! And it's just the same old set of bottles and hoses.

I was also just given a Novajet III, looks the same exept Encad built a bulk ink system into it. You guessed it, same old, same old bottle and hose, just more integrated into the case. Well known to be finicky and require constant suctioning to prime the heads, and constant attendance to amount of ink in the bottles.

Steve

Like it - how about a sample to test!!

But seriously, you appear to have done a grand job, but I bet the molds will cost a pretty penny (sorry cent) or two. I assume you are using the standard 1/8"od / 1/16" id tubing that everyone else seems to favour.

Have you any idea of the likely cost of the final production units yet

Peter

Wow Steve! That's really ingenius thinking. I really like the clear tube indicator and leak gauge. Must've cost a few hours doing it up.

Pauline

Yes, takes way too long to build that version. As it does any prototype.

Interested to know what adhesive(s) you use to assemble your system. MIS promote a "Loctite" Super Glue (cyanoacrylate) which uses what they call an "activator". Haven't tried this out yet, but would like to find additional types which work well on difficult plastics.

Peter.

ABS to ABS and PVC to PVC parts are glued with PVC or ABS cement. They are similar enough to each other, both are just a plastic solvent with either ABS or PVC dissolved into them as a thickener. Welds them together as if they started life as one piece.

Although I do use clear PVC cement to glue the gauge tubes in.

The 90 degree threaded hose spikes in the cartridges are glued in with superglue gel. I found other glues shrank as they dried and tended to pull away from the very slick plastic.