On ink sedimentation, agitation, clogs, and Hamlet

[cybis]

Not sure if this applies to printers other than the Epson 7/9900, and not sure if ink sedimentation is a real or imaginary problem, but I had this thought while thinking about John’s missing channels problem:

Epson recommends agitating cartridges before installing them; therefore we can safely assume that ink sedimentation can occur. But what to do once the cartridges are installed: to shake or not to shake, that is the question. Whether ‘tis Nobler the carts to agitate.

One idea being floated around is that there is no need to regularly remove and agitate ink cartridges: in the normal process of operating the printer, the ink inside the cartridges will be sufficiently mixed by the pressurization cycles squeezing and releasing the ink bags inside the cartridges.

But how could this work? There shouldn’t be any air mixed with the ink, and the ink, being liquid, doesn’t change volume with pressure changes. None of the components exposed to pressurized ink on the inside and ambient air pressure on the outside are flexible enough to expand significantly under operating conditions.  Therefore the ink bags don’t change shape during pressurization cycles, and no mixing of the inks should occur. The ink bags do very slowly change shape over time as ink is being consumed, but this won't be sufficient to mix the inks.

Maybe some of us have been misled  (or at least I have) by the sound coming from the carts during pressurization cycles, which seems to be the sound of ink bags changing shape, but is actually from the membrane sealing the air chamber inside the cartridges. The membrane is being pushed against the inside of the cartridge and makes a sound as it expands and contracts.

Relying on pressurization cycles as a way to agitate the ink and prevent ink sedimentation without removing the cartridges and manually agitating them, is probably not a working solution.

If ink sedimentation does actually occur, and if the denser part of that mixture contains a greater fraction of pigments, and if higher pigment concentration tends to cause more nozzle clogging, then such clogging would tend to appear when ink from old, unagitated, almost empty cartridges reaches the printhead.

It takes a long time for ink to travel from a cartridge to the head. So clogs would tend to show up very many prints after such ‘bad’ ink first enters the ink lines and potentially very many prints after replacing old cartridges with fresh ones. The long time delay between the cause and the symptoms would mask the causality link to the user.

More clogs, means more cleanings, and if cleanings shorten the lifespan of x900 printheads, not agitating cartridges in low volume environments could be another factor in x900 printhead failures, as it has been suggested by others in the past.

Is there any evidence of that happening?

[Ernst Dinkla]

Yes, I never imagined agitation happening by the air pressure system Epson uses. Carts also are inserted horizontally with the seal halfway in height at a vertical side so if there is pigment settling it happens in the lower half of the bag and that ink will be used at the end of the cartridge's life if not shaken periodically. Very passive way of applying pressure on the system.

On the other hand modern pigment inks stay in suspension very well. The pigment particle coatings have special characteristics for that purpose. One way to test pigment settling is using a (DIY) centrifuge like Paul Roark has done to test his quad ink mixes.

Actual ink agitation in the carts happens in my opinion in the HP system used on many of its printers, not just the Z models. I took the bag out of an empty HP cart and cut the top of that bag off. Cleaned most ink out. Filled with water. Syringe needle in the seal, pushed the rubber ball = membrane pump with a pencil and water went through the syringe needle. Nothing unexpected so far. Then I removed the syringe needle. Did the activation of the pump. Looking inside the bag I saw remaining ink from the pump section flowed in and out of the pump section and the bag. This and the more or less vertical position of the carts and seal + pump beneath suggests to me that ink agitation happens in the HP carts. I was already wondering about the fact that all pumps or at least 6 cart pumps at one side are activated when there is an ink request, the actuators of the pumps are driven by a kind of camshaft driven by one electric motor so all rubber balls are pushed in, one after the other. Actual ink flow is controlled by the butterfly valves in the head ink buffers, with buffer full there is no ink flow = ink agitation. The periodic cleaning also comes with the sound of pumps being activated, kind of high pitch sound like a leather or rubber belt slipping on a drive wheel.

There are more ways to keep inks agitated, white and metal color inks are known to have fast pigment settling. By pumping inks through a tubes loop, ink ending in the cart again and the heads halfway that loop, the inks remain in agitation over the entire channel. Heavier tube assembly of course. For normal inks the sideways movement of the tubes should already agitate the ink in the tubes. In all inkjet heads there are special compensations in the ink channel structure to overcome influences of the head carriage movement, sideways acceleration-deceleration, on nozzle ink supply pressure, indicating agitation there.

Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots

[JayWPage]

I have a 3880, so my observations may or may not be applicable to the larger Epson printers.

I recently had a PK black cartridge run out 1/2 way through a print of a night time image with a jet black sky. The ink cartridge had never been taken out and agitated other than when it was installed and other than the normal slight jiggling that the printer does while printing. I replaced the PK black cartridge with a new one which was gently shaken for about 20 seconds. When I examined the print, there was no difference what so ever in the blackness of the sky, but there was a difference in the amount of gloss. If one looked carefully at the print surface reflection you could identify exactly where the cartridges were swapped.

So I don't know about the pigment settling in the cartridge, but there may be other things that may vary between the beginning and the end of the cartridge.

[cybis]

Jay, that's an interesting observation. I have no experience with that printer, but I guess on a 3880, you were able to observe the effect of a cartridge swap instantaneously. How soon after the print came out of the printer did you observe the difference?

[cybis]

Ernst, it's very cool to have your insight into the various technologies out there. Epson definitely seems to be lagging behind HP in that area. I wonder how the Canon LFP deal with the issue.

[Ferp]

When I examined the print, there was no difference what so ever in the blackness of the sky, but there was a difference in the amount of gloss. If one looked carefully at the print surface reflection you could identify exactly where the cartridges were swapped.

I haven't had this happen to me, so I have no experience to draw on.  But surely what you're seeing isn't the new ink, is it?  The 3880 has a fair bit of ink in the ink lines and dampers. and it's going to take some time for the ink in the new cart to make its way to the print head and onto the paper.  Isn't there something like 15ml in each ink line?  I'd be interested to know what caused what you're seeing, but if I'm right in what I say then it can't be different levels of ink agitation and pigment suspension.  Not so quickly.

[Ernst Dinkla]

I haven't had this happen to me, so I have no experience to draw on.  But surely what you're seeing isn't the new ink, is it?  The 3880 has a fair bit of ink in the ink lines and dampers. and it's going to take some time for the ink in the new cart to make its way to the print head and onto the paper.  Isn't there something like 15ml in each ink line?  I'd be interested to know what caused what you're seeing, but if I'm right in what I say then it can't be different levels of ink agitation and pigment suspension.  Not so quickly.

Right, I wouldn't expect that either. I wonder more whether the end of the old cartridge influenced the ink amount applied versus that of the full new cartridge on the PK channel.


Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots

[JayWPage]

That's a good point. In retrospect, I couldn't tell exactly where the ink ran out because that part of the image was still inside the printer. It was only some time later when I picked up the print that I had set aside to dry that I saw an abrupt boundary in the amount of gloss.

[cybis]

Jay, is there any chance at all that what you've observed was simply the ink on part of the print having had a chance to dry at lot longer that the other part of the print? Was the difference still visible a day later?

If the difference wasn't due to differential drying, then Ernst is on to something when he says the difference may be due to a change in the amount of ink. Maybe an almost empty ink bag behave like an almost empty toothpaste tube: it's hard to get to the last bits. 

[cybis]

Here is another thought (not an explanation for Jay's observation): maybe when ink sits too long, it's not really the pigments that settles but the chemicals. I wonder if the water - glycol - glycerol solution can separate over time (any chemist out there?). If this is the case, those chemical have different viscosity and surface tension properties that may affect how ink behaves in the printhead.

[JayWPage]

I have located the print, (it was printed last winter and discarded), there is still a band about 2 cm wide that looks a bit different. It was printed on Han Fine Art pearl so the gloss differential is not so apparent when I look at it now. I don't have an explanation for it, but that band hasn't appeared on other prints of the same image.

The lesson to take away from it is don't let your ink run out!

[Ernst Dinkla]

I have located the print, (it was printed last winter and discarded), there is still a band about 2 cm wide that looks a bit different. It was printed on Han Fine Art pearl so the gloss differential is not so apparent when I look at it now. I don't have an explanation for it, but that band hasn't appeared on other prints of the same image.

The lesson to take away from it is don't let your ink run out!

Cybis' explanation that the ink had time to dry at that band before the printing was continued is more plausible. Just part of the ink droplets weaving happened at the band and was continued after the cart exchange, the absorption by the inkjet coating in two steps and each step with half the normal ink layer (more or less) makes a difference to the laying down of the total amount of ink within some seconds. Gloss difference is easily created if I see the difference between Economy and total sheet Gloss Enhancer applied on my Z printers, the total sheet mode lays down more GE and some papers (for example Epson Proofing White Semi Matte) show a dramatic gloss difference between the two modes.

Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots

[Some Guy]

Jon Cone mentions agitating the carts every two weeks here: http://www.inkjetmall.com/tech/content.php?133-Printer-Cleaning-and-Preventative-Maintenance&s=949279a4dffe258e01ee28a53a016f8d

"Agitate ink cartridges every 1-2 weeks to maintain in-suspension pigment, and use printers at least once a week to avoid settled ink in the pro model internal ink lines and to keep the print head moist."

I have seen "Ring around the cart" where some ink has left a tell-tale bit of pigment on the walls of the refill carts and some refill bottles.  Often with the bottles I can dissolve it back into solution with a lot of shaking.

As for the gloss differential of the OP's black, that can be due to the thinner in the ink, and also how much of it is being sprayed out of the head.  Spray to much and it can reverse in the density value easily in the black region.  I use QTR (A RIP software) sometimes and one can control the ink densities with it, but if I go too much to make it blacker, the outcome may actually be a lighter black which seems wrong, but too much and it begins to dull in appearance - and the spectrometer sees it too.

Over time printing with the same carts, I have seen a profile change between a new cart and once they get towards being empty; i.e. Reading a new cart vs. old cart test print.  Dunno.  Sometimes I've had nozzles clog midway through printing too which is more maddening and maybe an air bubble or something.

When I delved into that expensive Piezography B&W stuff, the entire first run of that ink was a costly "flush ink."  The former Yellow cart had a very light shade of black ink and on occasion it would show a yellow tint in the light black areas of an image.  Same for one of the Magenta inks.  Really became a crap shoot in trying to match something.  Sometimes some small crevice was holding some of the old colored stuff and it just got loose.  The second round of ink helped a lot, but I wasn't too crazy about spending $450 for an initial "flush ink."  Still not that crazy about that system either as some papers have their own tint and there were times when I'd like my tint to be different than a fixed tint off the Pizeography ink set.  Also, too much gloss differential on glossy papers that was not addressed by their "Gloss Optimizer" second-pass printing.  So back to Adobe RGB printing and away from the fixed-tint of the Gamma 2.2 piezo stuff.

SG

[cybis]

Regarding the issue of water - glycol - glycerol - pigment separation, I did a quick search:

• Water and glycol are very hard to separate. Doesn't seem likely that gravity would do the work.
   
• Glycol and polymer nanoparticles won't separate in centrifuges. I don't know what pigment particle coating Epson uses but it seems very likely they've developed a method to prevent glycol - pigment separation. Epson may have solved that problem. Third party ink manufacturers, and DIYers, probably not.
   
• Glycerol and water does separate easily in centrifuges...
   
  

Glycerol is the denser of the three and would settle to the bottom. It also significantly affects the viscosity of the mixture. I wouldn't jump to the conclusion that the higher the viscosity the more nozzles will be missing. In fact, it could be the opposite. I'm currently experimenting with water with a bit of jet dry (in an attempt to match the ink's surface tension) and color dye in my 9900 cartridges. I'm surprised to find that the printer is more prone to exhibit missing nozzles after a period of inactivity than with Epson ink. This seems to confirm missing nozzle are mostly not clogs. Is viscosity the key? This is a bit far fetched but is viscosity too low in high temperature environment somehow leading to more missing nozzles? Ink solution contains about 15% of glycerol. If glycerol does separate to the bottom in low volume - no agitation - high temp environments, for most of the time the printer would be fed with a low glycerol, low viscosity solution.

I see a centrifuge in my near future 

[Wayne Fox]

Therefore the ink bags don’t change shape during pressurization cycles, and no mixing of the inks should occur.

The ink bags I do believe change shape, when the printer is off, the ink will settle to the bottom and expand the bottom of the bag out.  When the printer is turned on, the bag is squeezed so the ink in the bottom is pushed upwards.

As to whether sedimentation is a problem, I've never seen any evidence of it and no longer remove my cartridges to shake them as I believe that  can cause more problems.  Shaking new cartridges is recommended because they possibly have been sitting for an extended length of time, but I think the slight vibration of the printer as it moves back and forth plus the pressure changes when turned off and back on are sufficient.  It's not like things settle out quickly and easily.

[cybis]

The ink bags I do believe change shape, when the printer is off, the ink will settle to the bottom and expand the bottom of the bag out.  When the printer is turned on, the bag is squeezed so the ink in the bottom is pushed upwards.

As to whether sedimentation is a problem, I've never seen any evidence of it and no longer remove my cartridges to shake them as I believe that  can cause more problems.  Shaking new cartridges is recommended because they possibly have been sitting for an extended length of time, but I think the slight vibration of the printer as it moves back and forth plus the pressure changes when turned off and back on are sufficient.  It's not like things settle out quickly and easily.

Wayne, I think physics makes it unlikely that the ink bag could change shape the way you described it. After your post I tested this by pressurizing a cartridge with the cover removed and observed that indeed the ink bag inside did not change shape whatsoever.
Here is the link to the video of my test: https://youtu.be/_irPncMwIbc

[John Nollendorfs]

Cybis:
Third party ink settling was a real phenomenon 8-10 years ago. When I quit using third-party bulk inks in an old Epson 9000, I observed yellow pigments obviously settling toward the bottom of a bottle, that had sat undisturbed for a year. Ink formulation has steadily improved, and the new co-solvents and pigment encapsulation have greatly improved. But it is still wise to not buy ink carts larger than what you might use up within a year. As much of a problem with pigments, is the coagulation of pigments into larger particles. In any event, just shaking the cart on a periodic basis will not really do much to help prevent either of these conditions. The best solution is to use fresh ink. FYI, Pigments are more prone to settling and coagulation than dyes.

As Ernst has observed, HP probably has the best system for keeping their inks thoroughly mixed. Also, HP grey inks are created with a consistent color balance. When you print in grey and black only mode, no color pigments are used to neutralize the warmth of carbon black. HP uses a lot of grey component replacement--evidenced by the much greater usage of especially light grey and grey inks. This probably explains why fade tests consistently show a linear fade and 100-200 print life unlike Epson or Canon.

[Ferp]

Jon Cone mentions agitating the carts every two weeks here: http://www.inkjetmall.com/tech/content.php?133-Printer-Cleaning-and-Preventative-Maintenance&s=949279a4dffe258e01ee28a53a016f8d

"Agitate ink cartridges every 1-2 weeks to maintain in-suspension pigment, and use printers at least once a week to avoid settled ink in the pro model internal ink lines and to keep the print head moist."

I have seen "Ring around the cart" where some ink has left a tell-tale bit of pigment on the walls of the refill carts and some refill bottles.  Often with the bottles I can dissolve it back into solution with a lot of shaking.

One of the challenges in using non-OEM inks, color or B&W, is sedimentation.  So that's one of the cost-benefit trade-offs you make when you elect to go that route in a Pro printer.  Hence Cone's reported instructions for his inks.  The general consensus here on Lula is that it's not much of an issue for OEM, and that's my experience as well.

As for the gloss differential of the OP's black, that can be due to the thinner in the ink, and also how much of it is being sprayed out of the head.  Spray to much and it can reverse in the density value easily in the black region.  I use QTR (A RIP software) sometimes and one can control the ink densities with it, but if I go too much to make it blacker, the outcome may actually be a lighter black which seems wrong, but too much and it begins to dull in appearance - and the spectrometer sees it too.

Over time printing with the same carts, I have seen a profile change between a new cart and once they get towards being empty; i.e. Reading a new cart vs. old cart test print.  Dunno.  Sometimes I've had nozzles clog midway through printing too which is more maddening and maybe an air bubble or something.

The OP had a specific problem, which was changing a cart mid-print left a band which could be seen as having a gloss differential but not a density differential.  I can't see how any of this relates to his specific problem.  There's no way that the ink from the new cart could have made it to the print so quickly and as the changeover only took minutes, I can't see that anything could have happened in the ink lines either.  You'd see that as a density differential.

I think Ernst's explanation in his last post is on the money.  The brief pause in printing meant that the two step absorption produced a differential gloss to the rest of the print which has a one step absorption.  Which is unlucky.  I thought you were supposed to be able to change carts mid-print in a Pro printer.  I wonder if this only visible in such a large dark area like a night sky.

[cybis]

The OP had a specific problem, which was changing a cart mid-print left a band which could be seen as having a gloss differential but not a density differential.  I can't see how any of this relates to his specific problem. 

There might be a little bit of confusion here. The OP (me) is interested in discussing ink settling and how it might relate to missing nozzles and/or impact the printing process in a very general and broad way, but with an eye on the Epson x900 printhead issues.

[Ferp]

There might be a little bit of confusion here.

You're right, there certainly was.  I forgot the original question and thought it was Jay's post.  My apologies to you and SG.

My experience is that sedimentation is an issue with 3rd party inks, but I've never found it to be a problem with OEM.  But then I've never had an x900 printer.