Ad
Ad
Ad
Pages: « 1 ... 51 52 [53] 54 55 ... 74 »   Bottom of Page
Print
Author Topic: Epson 7900 from the inside - out  (Read 263233 times)
Alan Goldhammer
Sr. Member
****
Offline Offline

Posts: 1573


WWW
« Reply #1040 on: December 29, 2012, 12:38:26 PM »
ReplyReply

As the resident chemist who has been following this thread here are some observations.  Acetone is an excellent solvent and as we learned when we first took organic chemistry, lots of thing dissolved int it.  Dissolve is the operative word here, plastic doesn't melt as melting is a function of heat.  If there are any binders that are used in the construction of the print head, then acetone is likely a poor choice of cleaning solutions since these compounds can dissolve, weakening the integrity of the print head.  It would be better to try a more moderate solvent such as isopropyl alcohol which used to be used by audiophiles to clean turntable stylus and tape recording heads without fear of damage.  The Epson inks are suspensions of encapsulated colorants (carbon black or appropriate color dyes) or as we commonly define this 'pigments' and my personal believe is that the quality control of the particle size which is critical here.  As Mark and others have noted the 3880 which is highly resistant to clogging has half the number of nozzles per unit area compared to the x900 printers.  Thus, Epson require far better quality control of particle size for inks in the case of the x900 printer.  The only way to know if it is a QC problem in terms of clogging is to document clogs across all printers by the lot number of the ink which is nigh to impossible to do.

From what I've read in the Epson patents, it is likely that there is some heat generated at the print head by the charge put through the piezoelectric head but I doubt that this is significant. 
Logged

Oldfox
Jr. Member
**
Offline Offline

Posts: 76


« Reply #1041 on: December 29, 2012, 12:45:53 PM »
ReplyReply

Alan, what about ammonia?
Logged
iladi
Newbie
*
Offline Offline

Posts: 49


« Reply #1042 on: December 29, 2012, 12:53:38 PM »
ReplyReply

Mark, printheads are consumables. Roland gives a live span of 7 bilion shots per a dx4 head. For a dx7 head, much more sensitive, must be less. The fact is piezo elements fails in time. They will not keep their properties and will start missfire, fuzzyness will became more and more obvious in prints. For a normal epson printer user, yes, printhead last a life, for a heavy user, a copyshop, like mine, it will last only a couple years. I have used over 35 liters of ink in an epson 9700 printer in one and a half year. Yellow is dead now. On my roland solvent i change heads every 2 years, it is a fact that i'm aware. I have to replace heads that are not cloged at alll, just fuzzy prints and lack of power to fire.

Eric, I have a modified epson 4880, turned into DTS printer (direc to substrat). Althou the manufacturer claims it uses waterbased ink, MSDS  talks about ether and other stuff like this. And the ink smells alot like alcohol.lokks to me more like a lite solvent than water.  The pipes are originals, made by epson. No melting. Monday i will try to see what impact the cleaning solution bfor the dts printer has on original HDR inks and let you know.  Then i'll send you the MSDS.


Edit: maybe Alan will help with the MSDS.
« Last Edit: December 29, 2012, 01:02:38 PM by iladi » Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6767


WWW
« Reply #1043 on: December 29, 2012, 12:55:53 PM »
ReplyReply

Thus, Epson require far better quality control of particle size for inks in the case of the x900 printer.  The only way to know if it is a QC problem in terms of clogging is to document clogs across all printers by the lot number of the ink which is nigh to impossible to do.


Alan, I expect the QC on all their inks for the professional series printers would be rigorous - built-in to the production systems. It just doesn't pay for them to do otherwise, because once it is routine it is cheap compared to dealing with damage and customer relations. I think the real fact of the matter is that they are truly on the leading edge of technology and they don't have a win-win solution in terms of avoiding *all* or *almost all* clogging. As you know, most design and manufacturing involves compromises. My sense of it is that they take a model design as far as they can within what they consider to be a technical improvement within an acceptable range of price and performance. If they had all the answers and they could bundle them into a marketable product, I have no doubt they would do so. And if they don't have all answers, having been the developers of this technology and the true specialists in it for the past several decades, one can never say never, but I'd be surprised if the rest of us will be particularly reliable at second-guessing them, as fascinating and courageous as much of this discussion has been. Remember the beginning of what must be this "Guinness Book or Records Forum Thread" - I'd still like to know *what really happened* to Eric's printer during the time it was in transit from the seller to his place. Of course the discussion has evolved a long way from where it started, and that's OK, but that lingering issue is...well, still lingering :-)
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6767


WWW
« Reply #1044 on: December 29, 2012, 01:03:53 PM »
ReplyReply

Yes, printheads are consumables. ..............

In the final analysis just about anything we use is "consumable". But the term has a specific meaning in this industry. When we say a printhead is "consumable", the meaning generally associated with that expression is that it is inexpensive relative to the printer and designed to be user-replaceable quite easily and at low cost - such as for example the printheads in an HP office inkjet printer. When we say it is not a consumable, we mean the opposite - it is a high percentage of the cost of the printer, and it is not designed to be user-replaceable - the case for ALL of these Epson professional series printers. The Canon IPFs are somewhat hybrid in this scheme - they are meant to be user-replaceable once enough of their nozzles get clogged, but they are not cheap.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
iladi
Newbie
*
Offline Offline

Posts: 49


« Reply #1045 on: December 29, 2012, 02:19:07 PM »
ReplyReply

Maybe you are right, althou my Roland costs 15000 euros, 2 heads are about 1500 euros, including labor, since are not user replaceable. But as i mention earlier, they last about 7 billion shots each head. And you miss my point: piezo DO  tears in time, each shot they fire. Maybe your head will not clog but will fail in time, the more ink, the faster. And canons are cheap. 350 euros for a user replaceable head and 4 liters of ink thru, it is cheap. Epson head is 850 euros the head, plus 450 euros labor and about 10 liter of ink by color.
Logged
Alan Goldhammer
Sr. Member
****
Offline Offline

Posts: 1573


WWW
« Reply #1046 on: December 29, 2012, 02:20:00 PM »
ReplyReply

Alan, what about ammonia?
Ammonia won't do anything if the clog is an organic molecule such as an ink particle.  You need to have an organic solvent.  In the old days we made stylus cleaning solution with isopropyl alcohol, water and a drop of dishwasher detergent.  It might be interesting for Eric to mix up a solution of 10-20% isopropyl alcohol (aka rubbing alcohol) and add a drop of Dawn to it (if it's good enough to clean oil soaked birds it should be good enough for an Epson print head) and see how this works.  It's pretty much the recipe that Windex uses but with more alcohol in it.
Logged

Alan Goldhammer
Sr. Member
****
Offline Offline

Posts: 1573


WWW
« Reply #1047 on: December 29, 2012, 02:23:03 PM »
ReplyReply

Eric, I have a modified epson 4880, turned into DTS printer (direc to substrat). Althou the manufacturer claims it uses waterbased ink, MSDS  talks about ether and other stuff like this. And the ink smells alot like alcohol.lokks to me more like a lite solvent than water.  The pipes are originals, made by epson. No melting. Monday i will try to see what impact the cleaning solution bfor the dts printer has on original HDR inks and let you know.  Then i'll send you the MSDS.


Edit: maybe Alan will help with the MSDS.
I'll be happy to look at the MSDS.  If one looks at the Epson MSDS, the inks are waterbased but have organic solvents in them (mostly glycols and glycerol) to help keep the particles in suspension.
Logged

Alan Goldhammer
Sr. Member
****
Offline Offline

Posts: 1573


WWW
« Reply #1048 on: December 29, 2012, 02:26:27 PM »
ReplyReply

Alan, I expect the QC on all their inks for the professional series printers would be rigorous - built-in to the production systems. It just doesn't pay for them to do otherwise, because once it is routine it is cheap compared to dealing with damage and customer relations. I think the real fact of the matter is that they are truly on the leading edge of technology and they don't have a win-win solution in terms of avoiding *all* or *almost all* clogging. As you know, most design and manufacturing involves compromises. My sense of it is that they take a model design as far as they can within what they consider to be a technical improvement within an acceptable range of price and performance.
Mark, I agree 100% here.  I suspect that the only difference between the x880 and x900 inks are the particle size of the 'pigments'.  As you know I'm also in agreement that the cause of clogs might be multi-factorial and that the installed user base is quite high.
Logged

Eric Gulbransen
Sr. Member
****
Offline Offline

Posts: 298


never surrender


WWW
« Reply #1049 on: December 29, 2012, 02:27:27 PM »
ReplyReply

...to really understand it I think requires a high level of very specialized education and experience...

....I'd be surprised if the rest of us will be particularly reliable at second-guessing them...

Don't worry I'm not fragile.


Listen I'm just a lowly 47yr old college drop-out carpenter who loves photography and (still) races motorcycles.  You're not gonna learn much from me here and I'll be the first one to tell you that.  But do keep in mind the genius buddy who I continually make reference to - he actually has a very specialized education which applies to more than just one aspect of these piezoelectric printheads.  While I respectfully appreciate that you might expect a "genius" reference from a carpenter's perspective to simply mean "Wow that guy's smart," what I actually mean is - he's a genius.  Keep in mind where I live.  I used to wave to Steve Jobbs on my way to work in the morning, he literally lived up my street.  The flip-phone was born above the very wood floors I'm writing this from.  This area that I live in right now, which I don't even understand how I live here, is a mecca for the mindful.  And our resident genius - who has access to more people just like him - has earned his his seat among the best of them.  It's a safe bet you hold in your hands today, Mark, something our very genius made yesterday.  Literally

I feel incredibly fortunate to have the very specialized help we do here.  There's not as much "guess work" going on as my goofy nature  conveys.  Would you like me to explain exactly how just forty six microscopic brass leads inside a channel's tiny ribbon wire can deliver firing charges to seven hundred and twenty individual Piezoelectric nozzles - all needing different information?  Because I can do that - all the notes are scribbled on three napkins from Thursday night's dinner at California Pizza Kitchen.  That's the type of specialized education we are fortunate enough to have access to here.  I have to brace myself before I ask questions.  Seriously some of the answers I get leave me feeling pretty dam helpless.  It's common I'm lost at the first sentence, never mind the last.  

As for "that lingering issue is...well, still lingering :-)," consider this fun fact and then tell me YOUR answer:  The 9900 that I bought in Colorado had a reported 75% of it's green channel missing.  I bought that printer in August, where it was dragged from the back of a mini-van packed with four hundred suit cases and bicycles and juggling apparatus's piled on top of it.  For a week it baked in the back of my cargo van in the high (dry) desert heat (110+).  Once I got it home that machine lay inactive, never even turned on, for 4 months more.  So it's been abused - subject to horrid neglect, super high temperatures, super dry environments, banging, smashing, vibrating and the like.  You name it this 9900 has endured it.  For sure way more than my shipped 7900 ever endured.  The 7900 was back up and running in only ten days time from the day it was shipped, which was also the day it reportedly fired it's last print - and it had terminal clogs from (my) day one.  So what would you expect a nozzle pattern to deliver on this 9900, after all this time and abuse?  Maybe not one channel firing one nozzle at all?  Well you'd almost be right, only magenta was 100% clear.  The rest of the channels/nozzles were pretty much gone.  After just one series of pairs cleanings (about three per pair), this 9900 was 100% clear and ready for printing - except for the very same 75% of missing terminally clogged green nozzles that we missing before I got it.

I don't know what this does for you, but it pretty well wraps up my view of the 7900 that I bought used.  



Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6767


WWW
« Reply #1050 on: December 29, 2012, 02:36:53 PM »
ReplyReply

Good to hear you're getting knowledgeable help Eric.

That comparison between the 7900 and the 9900 is truly intriguing and it's encouraging to see how much abuse that 9900 could absorb and still work. Maybe that says something for the intrinsic quality of these machine builds, but very little about specifically what causes some of them to fail so completely.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
iladi
Newbie
*
Offline Offline

Posts: 49


« Reply #1051 on: December 29, 2012, 02:48:45 PM »
ReplyReply

Ammonia won't do anything if the clog is an organic molecule such as an ink particle.  You need to have an organic solvent.  In the old days we made stylus cleaning solution with isopropyl alcohol, water and a drop of dishwasher detergent.  It might be interesting for Eric to mix up a solution of 10-20% isopropyl alcohol (aka rubbing alcohol) and add a drop of Dawn to it (if it's good enough to clean oil soaked birds it should be good enough for an Epson print head) and see how this works.  It's pretty much the recipe that Windex uses but with more alcohol in it.

You are right. Once i had some cleaning solution for my waterbased mutoh printer. It was alcohol, soap, some glycol and water.
Logged
Eric Gulbransen
Sr. Member
****
Offline Offline

Posts: 298


never surrender


WWW
« Reply #1052 on: December 29, 2012, 02:54:52 PM »
ReplyReply

Ammonia won't do anything if the clog is an organic molecule such as an ink particle.  You need to have an organic solvent.  In the old days we made stylus cleaning solution with isopropyl alcohol, water and a drop of dishwasher detergent.  It might be interesting for Eric to mix up a solution of 10-20% isopropyl alcohol (aka rubbing alcohol) and add a drop of Dawn to it (if it's good enough to clean oil soaked birds it should be good enough for an Epson print head) and see how this works.  It's pretty much the recipe that Windex uses but with more alcohol in it.

I will try this tonight Alan.  This is exactly what I had hoped for..
Logged

DougMorgan
Full Member
***
Offline Offline

Posts: 144


WWW
« Reply #1053 on: December 29, 2012, 07:14:15 PM »
ReplyReply

Happy holidays all:

I don't know if this was discussed elsewhere but I'm rather pissed to discover that the 9900 initiated an ink recharge when I replaced the cleaning unit.   It seems to have averaged about 60 mils PER COLOR or 600 mls total down the drain.   Another 200+ dollars in wasted ink on top of two for the cleaning unit itself.   Happy new year from freaking Epson.

There was no warning that this was going to occur and nothing about it in either service manual.   It seemed like the printer was going to do yet another unasked for cleaning but instead drained a large portion of ink from each color when it came up in regular mode.

The service manual specifies the sequence to use to check the AID function but it did not work exactly as shown.   Note that the AID check function produced blanks for the Mode 1 line on the LCD the first time I set it for 2.2 stop and then NG (Not Good I assume) for 2.8.   After setting it back to 2.2 the Mode 1 line still showed blanks rather than "OK".  Firing the printer up the next day (after the ink purge and reloading the most current firmware) the AID check function now shows "OK" and all is well.

I don't know what triggered the ink purge but the printer also seemed to have reset all the settings including reverting to the initial firmware version.   I must have triggered a complete reset while looking for a way to set the cleaning unit counter without servprog.exe.    Beware that there was no indication that any of this was going to occur.

Doug

Edited to reflect new information....




« Last Edit: December 31, 2012, 03:20:29 PM by DougMorgan » Logged
snsandrze
Newbie
*
Offline Offline

Posts: 3


« Reply #1054 on: December 30, 2012, 01:06:39 AM »
ReplyReply

Mark,
I appreciate your advice on 3880 vs. 4900. I have been debating this very issue for a while with the hopes that Epson will offer a rebate on the 4900 that I couldn't pass up. I'm still leaning towards the 4900 armed with the knowledge that it will require regular use and diligent maintenance (thanks to the enlightenment I've received from this epic thread). I have enrolled my cat in a printing 101 course, in my absence she will be able to output print jobs to keep the 4900 nice and lubricated! A little catnip and a furry toy and she'll do just about anything!

Eric,
I'm in the construction industry as well and I like to keep it simple. Things have to make sense before I can move on. I'm anxiously waiting to see if the cleaning solution is the answer without causing catastrophic damage to the head.

Steve
Logged
Eric Gulbransen
Sr. Member
****
Offline Offline

Posts: 298


never surrender


WWW
« Reply #1055 on: December 30, 2012, 02:06:08 PM »
ReplyReply

...mix up a solution of 10-20% isopropyl alcohol (aka rubbing alcohol) and add a drop of Dawn to it (if it's good enough to clean oil soaked birds it should be good enough for an Epson print head) and see how this works...

Alan I am looking to mix/test your solution now.  I have come up with a new system of flushing the printhead which will take far longer, but will be far safer for all parts considered.  Please if you will, give me an idea of percentages when you say "a drop" of Dawn.  Is this for a shot-glass sized portion of solution, or a gallon?  And when you say 10-20% of isoproply alcohol, what's the rest of the solution made of - Distilled water I assume?


I have tested again and no, we cannot use Acetone to clean these X900 heads.  I have confirmed that it indeed does compromise the bond of the glue used to secure the channel units to the printhead framework.  Interestingly enough it's not the bond of the glue to the channel (metal) that seems broken down - it's the bond of the glue to the (plastic) printhead framework that gets broken down.  Before soaking in Acetone it is very difficult to remove a channel unit from the framework.  Once removed, equal amounts of glue residue are left on both the metal channel surface, and the plastic printhead framework surface.  Upon close examination it is plain to see that the glue itself is what breaks when the parts are forced away from one another, before soaking, because equal amounts of glue residue are left on both surfaces.  In fact in no places at all are either bonded surface even visible - glue residue completely covers both surfaces.  But after soaking in Acetone, it's the bond to the printhead that breaks when the parts are forced away from one another - which is plain to see because almost no glue residue is left on the printhead framework surface.   This is obviously bad, but what is worse is the hidden bond between the tin roof and the piezoboard.  This is also compromised, as the tin roof is also easier to remove after soaking.   ...so Acetone is out.

I am happy with this discovery because quite honestly once you explore all the tiny fragile intricacies of these piezoboards you are left with a very high level of respect for their vulnerabilities.  While the Devil on my shoulder was anxious to watch Acetone melt away dried ink into nothingness from the chamber walls of a piezonozzle, the Angle on my other shoulder incessantly clashed pots and pans of despair in my other ear.  I expected these results, so I feel far better about using your milder solution for far longer amounts of time.
Logged

Alan Goldhammer
Sr. Member
****
Offline Offline

Posts: 1573


WWW
« Reply #1056 on: December 30, 2012, 02:18:40 PM »
ReplyReply

Alan I am looking to mix/test your solution now.  I have come up with a new system of flushing the printhead which will take far longer, but will be far safer for all parts considered.  Please if you will, give me an idea of percentages when you say "a drop" of Dawn.  Is this for a shot-glass sized portion of solution, or a gallon?  And when you say 10-20% of isoproply alcohol, what's the rest of the solution made of - Distilled water I assume?
Distilled water is best as you won't have to worry about any residual salts in tap water.  I think if you mix up two cups of the isopropyl alcohol/water mixture and add one drop of Dawn to that you will have a good working solution.

Alan
Logged

Eric Gulbransen
Sr. Member
****
Offline Offline

Posts: 298


never surrender


WWW
« Reply #1057 on: December 30, 2012, 02:21:36 PM »
ReplyReply

on it, thanks Alan
Logged

davidh202
Sr. Member
****
Offline Offline

Posts: 540


« Reply #1058 on: December 30, 2012, 05:42:19 PM »
ReplyReply

Eric,
 I just had a thought, STEAM as in cleaning oil and grease from engines.
   might it be possible that hot water under mild pressure (steam) from under the nozzle plate, could be enough to soften and break up the "dried" ink in the piezo channels, enough to allow it to be forced out of the nozzles by injecting mildly pressurized  distilled water from the damper side of the nozzles.?

BTW,From your pics and description, I have a suspicion the "speed bumps" are ink flow restrictors (governers) controlling the final flow to the nozzles.  
Logged
Eric Gulbransen
Sr. Member
****
Offline Offline

Posts: 298


never surrender


WWW
« Reply #1059 on: December 31, 2012, 12:08:27 AM »
ReplyReply

I like your thinking David.  Heat, pressure, chemicals, even ultrasonic vibrations - there has got to be a way...

Here's the thing though - we face some tough challenges when you consider that clogged up ink is typically a LOT larger than the hole we need it to exit the head through.  Forcing the dried ink free with pressure is an attractive idea, albeit dangerous considering how fragile the piezo chamber walls are.  Backwards would be my direction of choice to force hardened ink through but regardless, we need to break it up before we can force it anywhere.  Right now it's land-locked inbetween the speedbumps and the nozzle openings.  

Here's a few illustrations I made to show a piezo channel unit from a perspective that I can't photograph.  They go in order - no ink, filled with ink, hardened ink stuck inside the chamber walls.  These illustrations are not based on theories.  I've seen all this crap up close and personal, hardened ink included..







« Last Edit: December 31, 2012, 12:33:45 AM by Eric Gulbransen » Logged

Pages: « 1 ... 51 52 [53] 54 55 ... 74 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad