Officejet 8500A printhead cleaning - how does it work? (long)

[Pcarsba]

My printer won't stop trying to clean its print heads. So in an act of desperation I took the whole thing apart to try and understand how it decides on success for the cleaning routine. Here is what I found so far:

1. The waste tank has this enormous sponge inside. It was saturated and ink was sloshing around in the bottom. What a mess but I cleaned it all up and washed the sponge.

2. There is a small peristaltic pump connected to what looks like two rectangular rubber funnels on a movable cart. When the cart gets moved under the print heads they suck on the printheads and force ink through them. At the same time Some of the lines get pinched. Ink first gets sucked into a little canister and then purged into the bigger waste tank with the sponge.

3. There are four (two for each) rubber wipers on that same movable cart. They just wipe off dried ink from the nozzle area.

4. Now comes the mystery part. None of the above has any sorts of smarts or sensors to "know" if the cleaning did anything. However, here is a small PCB in a plastic cover on the left side of the cleaning unit. It also empties into the waste tank. It has a metal antenna (my guess) with a rectangular opening and a second metal shield. The opening has plastic sliding cover that at times covers the opening (when the cleaning cart moves to the back) and then snapps back again pulled by a spring. The circuit board has a single OpAmp. And then this is connected to the main PCB. For lack of better description the metal cover looks like a WiFi antenna. And it is connected to the PCB.

Has anyone any information on what this device does and how it works exactly? Is it some sort of flow meter? Does it measure whether the spray passing through it is dense enough? The printheads seem to spray to a little blast plate angled at 45 degrees in the waste tank. I was not able to see whether they also spray through this mystery device. If not how does the mystery devise measure? And what's with the moving cover?

In my case I am able to make my printheads happy when soaking them in cleaning solution to the point where they show some acceptable blot pattern for all 4 colors. But when I put them back into the printer it literally keeps sucking on the yellow for 10 minutes or more until the tank is empty. In the process it all craps up and if I get lucky to where decides to finally print most colors are missing. When I take the heads out and put them on moist paper the is no blotting pattern. WTF.

I have two running theories?

1. The refill pigment ink (NanoDigital pigment) I am using is so bad it clogs some of the nozzles right away during the test shots in the cleaning cycle.
2. The mystery device is broken. As a result the ink delivery system can not keep up with ink consumption during the sucking on the heads?

I lean towards 1. because when I do get the printer to print it remains in whatever state it then is. In other words once it makes it through one of these super extensive cleaning routines to the point where it actually prints a test pattern one or more colors are impaired to various degrees. And that streak pattern remains. Doesn't get better, doesn't get worse. If I decide to re-do the ordeal I might end up with one or two colors looking better and another being completely gone. And it remains in that state. To me this spells nozzle clogs. I have not (yet) found a way to remove and clean the heads without the printer noticing. Ideas?

I am frustrated - too bad I am not living in a 10 story high-rise. Else this printer would have had an accidental fall to its death by now......

 

[Pcarsba]

Let me add some images for clarification of above. The first is when the two suction cups (green) are under the print heads. You can also see the retractable cover (blue) over the rectangular antenna (red). Inside the metal antenna you can make out the circuit board with one SOIC-8 OpAmp and some resistors. I have not seen any ink inside the rectangular opening after my cleaning. Not sure if the printer never got that far or whether this is by design. The retractable cover now open (red) must be significant because there is a pretty intricate mechanism that pulls the cover shut covering the opening and then it snaps open when a spring under tension releases. My assumption is that it baseslines the response from this device. The bottom of that opening opens into the waste tank and there is a small plastic cover over that protects the gear shaft. It is still spotless. Also, the circuit board wasn't badly contaminated when I took everything apart. So I don't think the head squirts into the opening. There is a second metal plate connected to a wire to the right of the opening. It is pretty much sandwiched between the circuit board and the left sidewall of the waste tank. It is not visible in the image but must be significant, too since it has its own singe wire running to it. My guess is this is some sort of large capacitor or tunable antenna sensing ink?

.


Next a shot with the four wipers (white) exposed. They just act like windshield wipers when the cart moves back and forth during the cleaning. Below the wipers you can see half of the 45 degree angled blast shield with some magenta ink.

The next shot shows the blast shield better. The print heads fire onto this shield and there is a nice drip line at the left with a little cut noth where the ink collects and drips into the waste tank bottom. I can hear the printer firing and I can see ink when I shine a flashlight just right into the printer during the cleaning. I am still not clear how the antenna device detects how well the cleaning went.

In the last image I show the peristaltic pump (red) and the collection cannister (green). The pump has simply a couple of silicon tubes stretched over a rotating square. It sucks on the top two lines that are connected to the collection cannister. The bottom goes into a Y (currently disconnected) from the lower bosses and then into the waste tank. When the cart moves to the front it can pinch off the exit so there is an effective vacuum pulled onto the head suction cups. Ink will collect in the collection cannister. Then it can be pumped out of that cannister into the waste tank bottom.

 

[turbguy]

I would suspect the firmare on your HP detects non-OEM or refilled carts installed (they do have chips on them, no?), and causes a LOT of waste ink to be used intentionally. Some of Canon's models have been accused (AND MEASURED) of doing the same thing. Have you upgraded the printer's firmware?

As far as I know, no inkjet printer actually "detects" the need to run a cleaning cycle. It appears to be done on a counter-based software system that tracks print head usage during p[rinting and time between prints (if unused for several days). Perhaps you have a dead CMOS battery (if there is one) and the printer believes it is being started afresh each time, and repeatedly goes through an initial prime cycle (which uses a LOT of ink). And perhaps I'm WRONG!

Others here...feel free to correct me, please.

Wayne

 

[Pcarsba]

Yes, the firmware does detect and track carts and marks them non-OEM after the ink usage estimates render them empty. This is something every newer chipped HP cart will experience at some point in its life. On this model printer the firmware also tells you the non-OEM ink has been detected in the print heads. I assume it simply does the math: Once the volume of ink in OEM cart plus the volume in head is lower than the volume squirted with the head it assume the non-OEM ink has made it to the nozzles.

But it doesn't explain how the printer detects or decides about the head cleaning being complete. I don't mind if a cycle starts every so often. I know one trigger is when you print a quality report and the internal photo detector attached to the print head holder doesnt see the cross-hairs. However, cleaning goes on for more than 10 minutes always doing the same routine (wiping, sucking, squirting) many time over makes me wonder if something is broken.

Remember, using CIS or refilling will lead to a situation where the carts (chips) are flagged as refilled or non-OEM. I can see where HP is a little vindictive and ups the consumption by throwing in cleaning cycles more often. But this should not cause the printer to go into an endless loop of attemting to clean the heads. Agreed?

I would really like to understand what the device in my first picture does and how it works. It somehow is related to the cleaning - just how?

And yes, this model has a CMOS battery. My assumption is this is for the RTC and maybe settings memory. Maybe I should yank it and make the printer forget about the carts. But then I believe the printer writes to the chips and this is how the chip are rendered "empty"


This brings me to my second question: Has anyone tried the ARC chips for HP 940 carts that start poping up on eBay. I wonder how well they work or whether these are just some harvested chips from old carts?

Ingo

 

[ThrillaMozilla]

But then I believe the printer writes to the chips and this is how the chip are rendered "empty"

Not necessarily. It could just remember the last few chips.

There was a method that worked with some HP printers, in which you put in two or three sets of cartridges in a row, so it loses its memory of the first one, which you have now refilled, of course.

 

[lisl]

i noted that my HP8500 entered more cleaning cycles when I installed a CISS. That's on top of regular cleaning everytime I start up the printer. It's very annoying!

Pcarsba, how did you get access to the waste ink sponge? There's a panel on the right side of the printer that seems to be made for sliding off, but I just can't seem to get it out. Any help?

Thanks