Attempt to repair printer.

2100p CIS repair

It has been almost six years since I wrote my latest update on 2100P CIS.
The printer has not been working for four years.

After all this time I'm not optimistic but it was an expensive printer and it's worth one last try before junking it.

The primary problem is clogged jets in the black print head but there are also problems with the ink chips not seating properly. I also suspect I may have a vacuum pump problem.

Because of these !@#$%ing chips it is difficult to fix the other problems.

So part one of the attempted resurrection is fixing the chip problem.

Most inkjets are not worth fixing but this is a high-end printer with extra mods.
The 2100P (or 2200P in the US) is a pigment printer this makes them more prone to clogging and much harder to unclog.
The printer uses a home made CIS (continuous inking system) and a third party inkset made by media-street.

In 2002 I cracked the epson ink-chip protocol and fitted the printer with a circuit to fool it into thinking the chips were reading full.
I also came up with a chip reader/writer/resetter using a PC printer port and a little program I wrote.


Part one.

The cartridges and carriage were in quite good shape after all this time.
I was expecting the carts to be stuck onto the ink spikes but they were all fine.
The CIS lines are still connected. All the carts except yellow still had ink in them.

Image Looking into the carriage you can see the ink spike and the chip contacts. The spike was drilled out a few years ago to clear a suspected clog - I can't remember exactly what I did

One of the problems is the loose fit of the modified cartridges.
The cartridges don't fit properly after my CIS conversion because one side cover is removed.
The original covers are long gone so I glued some plastic sheet in place at the chip end of the cartridge. This improves the fit a little.

Image Gluing on a plastic spacer

I used a CD jewel case as the source of plastic - both the clear and black pieces worked but the clear is a little easier because you can see the glue.

I just used superglue - this doesn't stick well to the cart (nothing much does) but it will do for now.

The fit isn't perfect but a big improvement over what it was.
I had to wiggle and jiggle a little but managed to get the printer to read all the chips and attempt a cleaning cycle for the first time in a couple of years.
The cleaning can't possibly work because the vacuum pump is disconnected.
That's the next job.

Part two.

It isn't going well. Re-attaching the vacuum line is keyhole surgery. I've done it a few times before and it is never easy. I bought clamping forceps for this job before but I don't know where I put them when did this two years ago.

In the process I've noticed another serious problem. These printers use a long strip of film with a fine grating on it for optical motion feedback. This has been splashed with ink. In trying to clean it some of the emulation has lifted off. I'm pretty sure I will need to replace it but hopefully the old one will work well enough to for me to know if the printer is fixable.

31st May,
I went on a failed shopping expedition to buy new forceps.

I've been able to plug the vacuum line in with much difficulty but it always falls out again. This time I dried inside the pump connection (F) as best I could then put a little silicone around the plug before inserting it. Lacking my forceps I use a pickle grabber and a hooked bit of wire to maneuver the plug into place. The tube was much longer than the original.
Image Glued in Vacuum line. The vacuum pad has been removed and the mechanism pulled right as far as possible. The white near the center of the photo in the silcone

Image Attached to vacuum pad

I trimmed the tube so it was still long enough to attach it to the vacuum pad.
Image Reassembled

It is all assembled again all looks good but will it work?

It doesn't work. It appears the pump is not working for some reason.
Removing the pump will be a major problem and the printer will need to be largely disassembled.
I'm not sure when I'll get time to attempt it.
On the positive side this explains why I was unable to get the printer working before so the head may be in reasonable shape.

I don't really have a clue what is wrong with the pump. They are peristaltic pumps and quite simple in operation but because it is buried where I can't see it I don't have much to go on.
It could be that it isn't being driven. It could be a split tube or a blockage.

I don't have a repair manual and it sure would be handy to have one.

30 minutes later.
I do now! I downloaded a manual for £2.50 from http://www.serviceemanualepson.com
Dave from serviceemanualepson has been helpful and can also supply service programs at reasonable prices. For a few bucks it is worth getting a manual so you don't break things (as much) because it shows you where the various hooks are that hold things together.

Serviceemanualepson.com is now dead and Dave is moving to - http://store.payloadz.com/results/results.asp?m=84808



Image A look at the pump unit though the side of the printer mechanism after the panels are removed.

Image The pump and cap assembly after removal from the printer.

The pump doesn't appear to work when rotated by hand. The tube appears to be intact. The rollers are meant to engage when rotated in one direction but not the other. It seem likely one or more rollers is stuck.
A pump I have from another epson works fine when rotated by hand.
I will need to disassemble the pump to see what is going on.

The pump is the only one roller peristaltic pump I've ever seen.

I think the design is crap. If it worked I might think it novel but it doesn't.


In the photos you can see the pressure roller sits in an inclined grove. When the shaft rotates one way the roller is pulled away from the tube and pumping stops with the tube open to flow.
When the direction is reversed the roller is supposed to be force out against the tube and begin pumping.

The problem is the roller stopped half way (or so) along the grove and never engages fully. The design relies on the roller/tube friction being enough the move the roller along and it no longer does.
I roughened the roller with abrasive and it now kinda works.
I don't know how reliable it will be but it is a lot better than it was.

My other espon pump uses two metal rollers and tube in a channel (it works).
I can see the new style could be cheaper to make but this is a crappy pump to have in an expensive high end printer.

The pump is reinstalled and appears to be working.

More chips problems.
After reassembling the printer I had more chips problems with both black and cyan leds on.
Eventually I solved the cyan problem be dumping an old copy of the chips data back into the chip using my chip reader mentioned earlier.
The on-board chip defeating circuit only changes the ink level and I suspect some of the other counts may also need to be reset. Usually to first 9 bytes of new chips are zeroed and I may need to modify my circuit to clear more bytes. Mind you this only needs to happen every few years.


Black also seemed to have this problem but the main issue was dirty pads on the carriage circuit board.
Even with the service manual it wasn't obvious the little blocks were removable and had contacts on both sides.
I bent a small screwdriver to make a removal tools.
Then promptly mangled all the rear contacts when attempting to push the block back into place.
The contacts catch the edge of the PCB. Once you know this is it easy to prevent this by placing a trimmed paddle-pop stick abutting the PCB while sliding the block into place.

I also managed to slip when removing the second block and mangle the front contacts.
In both cases I managed to bend the spring contacts back into shape enough that they worked.
The blocks can be bought as spare parts but are a little expensive.
If I get the printer working I'll buy some but at this point it would be a waste of money.


Note in this photo I have the frame of a yellow cart plugged into black and the error light is off.
I use this cart to check the chip contacts. The mechanical key has been removed so I can plug it in to any slot.
The chip has been reprogrammed with "black" data. It is not even an epson chip - it is a chip from a calidad cartridge.
The chips are on two circuits- three chips on one and four on the other. The printer will detect the chip no matter which slot it is in but the ID will be wrong if it is on the wrong circuit.

More pump problems.
After getting the chips to read it was apparent the pump was broken again.
This time I looked to buy a one as a spare part. I found them to be quite cheap in the US but expensive in the UK.
The US agent won't export and the Aussie agent didn't reply for a while.

I repaired the pump by covering the pressure roller with heatshrink tubing to build it up a little. I think it is now a little too big but it has been working reliably so far.

After the pump was repaired the Aussie agent replied with a good price and I was also able to confirm they can export.

compassmicro.com (won't export).
1303605 pump assembly : Stylus Photo 2200 (us)$13.11
2032470 scale, cr (encoder strip) : Stylus Photo 2200 $3.64

www.partsalliance.com - pump assy for Epson 2100 is (au)$16.12 ex gst (p/n 9303605)
They say they can ship anywhere :-)

Still no joy unblcoking the head so it had to come out. It isn't all that difficult and having a manual helps by showing where the clips are.
The head is stuffed. I obviously applied to much pressure trying to clear the blockage.


The head was leaking at the seam where the front metal plate was bonded. There was also leakage between the heads at the "black" end of the print head. The heads at the other end (ie yellow) seemed to be fine.


Image This view shows the plumbing which connects the ink spike to the actually print head

{img src=show_image.php?id=230&thumb=1"|link=show_image.php?id=230}

Image a scan of the head after the front plate is removed.

I managed to lever the front plate off with minimal damage. While epson's marketing practices are vile there technology is awesome.
The heads are a modern marvel.
Even the $50 printers have a scaled down version of the technology and clear are sold below cost.

The replacement cost of the 2100 head is au$305+gst - quite a gamble to get an old printer going but I'll give it a go. You can't simply replace the head you also need a service program to change the head ID number.

I've also ordered a new inkset from https://www.inksupply.com so I'm somewhat committed to getting an epson printer working again.

I've said I'd never get another epson but I'm tempted to try the newer printers with primary colors in the inksets.

July 23'rd.
I finally have the new print head - over a month waiting.
In the mean time I forgot what screws went where and ended up using the wrong screws to secure the head fastener. I broke one of the plastic lugs that the screw goes into.
It still sits fairly securely - the broken lug works now like a cam. I may glue it later.

On power up the head/carriage would slowly go to the other end of the slide then seek home. The process would repeat several time then the printer would stop with all leds flashing.

I had the service program running and it would say status was "printing" while the seeking was going on then "fatal error" when it stopped.
The problem was the screw I'd use to secure the fastener was too high and jammed before the carriage could home properly.

Once this was fixed it demanded the "out of ink" error is fixed before it would allow the head-id to be changed.

Later - Installed ink, set head ID and now the pump is broken again. I ordered a new one.
Enough for now.
Image Left - the head ID sticker on the new head. Are those zeros or ohs? Right - using the service program to enter the new ID into the printer. Image

August 20.
After a four week wait I was minutes away from sending an email to parts alliance to see what happened to my pump order. Then there was a knock on the door and a courier with a parcel.
I'm too busy now but maybe later today I'll install it.

August 25.
The new pump is fitted and pumping. However there is still a problem. The cap unit isn't sealing against the head so the head still isn't being primed. I've cut a hole in the case now to access the pump area without repeatedly disassembling the printer.
I've manually got some ink into the heads and the head in firing on some of the nozzles so I know the new head more or less works.

The solution may be to buy a new cap unit but I'd like to verify that is the problem first.
If I get a good nozzle test I'm prepare to spend more money on parts.

It is quite possible this is the problem that prevented me from getting the old head working but I'm pretty sure I did have the pump working and sealing after the last repair.

I'm also considering another chip hack to replace all the chips with one micro-controller - Done.

Sept 23,
Almost another month passed before I had the parts and the time to continue.

In the mean time I got my ink chip replacement hack working and took great pleasure in removing all the chip from the cartridges. The lack of chips meant I could but a couple of carts in a sonic cleaner to unclogged the button sponge filters. There was also a blocked CIS line (black) I needed to unclog.

The new cap unit did solve the problem with air leaks. However I managed to jam the wiper assembly and strip a small gear in the pump assembly.

Image Left - The small broken gear. Right - a gear from the paper feed assembly of a different type of epson happened to be the right size to improvise a replacement. Image

I wasn't keen to wait another month for a spare so I had a look in the wreck of another cheap epson I was given and found a cluster gear in the paper feed mechanism had the right size to fashion a replacement.
I cut the gear to size and glued in some carbon fibre tube to reduce the hole size.
So far it is working just fine.
The printer is almost working. I still have a few clogged jets but not many. I haven't done all that many cleaning cycles yet and I think they will clear.


After four years of gathering dusk
$400 worth of parts (not counting new inksets)
four months of having the printer in pieces
A major hack to eliminate the #$%^ing chips.
Much cursing and many days with inky fingers..........

It works.

Image The first successful print to feed out of the printer in four years

There is still some tweaking to do but the hardware works again.

13'th October 2008.

The printer worked but was printing too pink so I detoured to update my colour management software. I also have a few jets still clogged as a result of having the printer sitting around too long while being repaired.
I got all the jets clears today and did a dozen fairly good prints.
There is still a minor problem with paper alignment.
I only learned today that the paper sensor is on the carriage and this is probably is bad shape after all I've done to keep the head wet.

I can't print the alignment test pattern because it tells me there is a paper jam. I can worked around this.
I'm not happy with some of the noises the printer make when doing a rapid paper feed - a sort of rattling sound which I haven't tracked down yet.


Created by system. Last Modification: Thursday 18 of February, 2010 15:41:38 AEDT by eddie.

Main Index

Switch Theme


eddie, 11:29 AEST, Sat 22 of Jun, 2024: Offline wiki repair is mostly done. It is running on tiki-27-alpha
eddie, 18:14 AEST, Sun 12 of May, 2024: Phase one of wiki repair is to upgrade to V22 (the last version with image galleries). All 150 images in the wiki-up directory are broken and need manual wiki page repairs.
admin2, 19:45 AEST, Sun 28 of Apr, 2024: Tested tiki 26.2 on a raspberry pi. Tiki 26 does not support image galleries so I have to try again after migrating images to file galleries.
eddie, 13:04 AEST, Thu 10 of Aug, 2023: Offline tiki 26 upgrade went badly. Waiting for 26.1. Will limp on.
System Administrator, 18:45 AEST, Wed 26 of Jul, 2023: Recovered from lockout but unable to upgrade to V24

Last-Visited Pages

Online Users

3 online users