Mimaki UJF-3042 restoration

[kehall]

Hi everyone

I thought it may be helpful to someone if I'd document/share my experience of bringing back to life an old UJF-3042 printer, purchased with the aim of helping out a local business print product samples and experiment getting colour matching right (printing on various plastic, needing white ink) whilst leaving their main production line free, at a low entry cost (couldn't justify cost of new or even refurbished and the affordable low-end Chinese printers look somewhat 'buy cheap buy twice' etc).

I have a background in IT, electronics systems repair and a 'can fix anything' tenacity so felt this was a bit of a project that could possibly have a satisfying, rewarding ending!

So I saw this printer for sale apparently only needing new heads, so did some rough calculations and made a suitable offer based on cost of new heads after some research and possibility of flushing the Ricoh Gen4 heads, picked it up, removed a few covers and gave it the once over before applying any form of power, armed with copy of service/maintenance manual and good information found here amongst other places.

Thinking ahead I bought some Ethyl Acetate to help with the cleanup and head flushing, and was/am considering building a pumped flushing unit as per some very helpful Ricoh youtube videos.

What a mess I found!

Didn't really know what to expect, but it feels a bit like this was a daily workhorse that didn't get much time if any for maintenance or cleaning, only got any attention when a problem occurred. I poured the (full) waste ink container into an old empty IPA container I had laying about, and tested some of the ethyl acetate to clean up the congealed ink mess, which worked well but it evaporates super quickly, meaning you have to be quite quick to mop up!

The head slider board, ffc cables and head connectors at first glance all looked ok, but later turned out to be an issue.

The UV cooling radiator looks like it could do with a blast of air, fair coating of dust here

The LED UV unit itself has a very dirty glass (no photo yet) but that's on the list for cleaning as it will no doubt be reducing the effectiveness of curing.

Wow, the mess under the station wasn't expected, this was the first glimpse that what I bought wasn't just retired for new heads only - look at the pump assembly in pieces! The middle pump cover was laying on the bottom left, the insides of the pump and a split peristaltic tube circled.

This area is still due to be addressed, but first things first, since it wasn't any more just a 'simple' head replacement job I needed to check the printer systems electrically, in case I needed to make any electronic repairs too. Gingerly I applied power... a few fans started up and then beep beep beep, oops, a problem:

I couldn't see any of the suggested voltages on the slider board (first step in diagnosis). Now the information I had suggests a connection issue on one of the FFC cables between the slider relay board and the slider board and I'd read about a modification to early serial numbers (including this machine) to apply protection to stop screws from damaging the cables, so I disconnected both ends and metered out the lines one by one. All tested ok, however one of the blue stiffening back strips was coming away from the cable so ordered 3x 2m long 30 pin 1mm 'reverse' cables from aliexpress, which would take a week or two to arrive. In the meantime I reconnected the cables, and powered up again.

This time I got 11f SLDR PCB halt, so a difference. I watched the slider relay LEDs during boot up and noticed D1 wasn't lit, the information said a fuse was open but I wasn't so sure. I disconnected the FFC cables again and looked to see if any difference, nope. So out comes the slider relay board so I could check fuse etc.

... more in next post as reached limit on attachments.

 

[kehall]

Slider relay PCB with FFC cables to slider board disconnected, showing no D1 light (located between CN3 and CN7):

Inspected the board for fuse problems etc but found nothing obvious, all fuses intact (the fuse for the 'D1' led power indication is on the reverse of the board).

I started tracing out the circuitry to determine the source of power and this is switched on by a P-channel mosfet, driven by a tiny SOT523 NPN transistor from the Altera FPGA. With the board back in the machine, I could see a very small drive to the base of the NPN transistor but not enough to turn on the mosfet.

I started thinking the NPN transistor was faulty, so was preparing to replace that, but equally the fault could possibly have been in the FPGA (a part that wouldn't be available, and BGA soldered, not something insurmountable as have the equipment to do but not something I enjoy doing!). Looking around at replacement boards, the costs were silly prices, exceeding that of the machine itself, but a last resort if needed.

So put the board back in and I decided I'd ground the collector of the NPN transistor to enable power manually, and whoopee! the D1 led lit up and so did the LEDs on the slider board at the top of the machine.

So I rebooted the machine with this power still applied, but alas, still got the 11f fault

Put my thinking cap on, and could tell Mimaki had engineered this to be pretty robust, with various checks etc before applying power, so thought, ok maybe it's using some lines of the FFC to run the power enable signal to and from the slider pcb before enabling power (to avoid the possibility of reversed connections blowing something up), so postponed any more messing with the transistors etc and waited for the replacement FFC cables to arrive.

Temporarily fitted the new FFC cables not particularly expecting miracles, but imagine the delight when seeing D1 glow (and another LED near the FPGA that didn't light up before) ! It's ALIVE!

The other connections were left off in order to just get past this phase.

The rest of the connections were made, and printer powered up again. This time no HALTs, and the whole top section started sliding back to 'home' I presume, which led me to hit the emergency stop button as otherwise it would have trapped the new cables and things.

With the coast clear, powered up once more... this time a bit more whirring, the head slider shifted to the left, a bit of jiggling about and then get another error!

Oh no... This can't be good. After a while I also got a continuous beep and ERROR 707 !HD HEATER BRK so clearly needed some more attention.

... to be continued

 

[netsol]

Regarding the fast evaporation rate, we add butyl carbitol snd or butyl cellosolve to our acetone to make cleaning a bit more managable

 

[kehall]

Two of the head EVEN/ODD connectors were then swapped around and the HD THERMIST error moved with the head cable, so it wasn't a fault with the slider PCB. It was felt to be a bit odd to have two even channels with the same failure and I'd understood these heads to be very reliable/durable, so looked around for any other cause, that's when I found this!

Can see two very clear (but easy to miss if not looking for it) breaks in the cabling, at the same point on both. Can only put down to repetitive fatigue stress, understandable given the speed of the head motion and the inertia in the cable assembly without any additional support.

Now I'd repaired similar cables in the past so this was my next mini project. Only I needed to take the heads home to work on them under the microscope, and they were stuck in a 130Kg heavy printer at my industrial unit. So out they had to come.

Oh if it were only so simple. I removed the (seemingly full) cartridges as to avoid draining ink everywhere (or maybe they should have been left connected to provide a 'vacuum/air lock').

The head removal instructions were clear - disconnect the ink lines from the damper (find something to plug them with or raise them high to stop them dripping everywhere), remove the damper cover (there must be an art to that as I didn't find it that easy trying to tilt back to release from the tab without moving the damper), flush through (ok, my first chance to do this) with air valve open etc. I decided I'll flush them out of the printer so I can also see stream from nozzles.

Then tried removing the middle (Y/K) head securing screws with flat blade driver... 'This is tight and chewy' I thought, so wiggled back and forth a bit but then snap! head screw sheared off didn't it Had a go with the back one, more care this time, snap! that one went too... DAMN! I thought..

Those head screws are made out of putty it seems (and old ink I guess acting as a very effective threadlocker).

So anyway out comes the head, and I decide to flush it, first with ethyl acetate through air bleeder hose, cap removed (two of those were stuck fast too!), then with cap on through nozzles, dribbled at first but a little more pressure and could see a nice stream. Then ran some IPA through as it was less volatile (the service guide states refill with ink within 24 hours) so didn't want anything solidifying in the nozzles really, but it is what it is, still no idea if the heads are usable or not but try to not make it worse if it can be helped!

I then go for the left hand head (the other one with the broken cable), and with utmost care and dribbling ethyl acetate down there to help release, the head screws came out a little easier. Repeated the flushing on that one.

Can see below the remnants of two broken screws in the middle head bracket. Scratched head as to how to best approach that, needed drilling out/extracting, but not something that can easily be done in situ, better on a bench drill in a vice.

A bit of reading suggested I could remove the bracket assembly by taking out the top and bottom hex screws (vertical head alignment locking screws) without upsetting the alignment (as not touching the angle adjustment or the positioning lever locking posts), so did that, and noticed a spring underneath which would need going back in the same place.

As it didn't matter if the screws were a bit smaller, they were trimmed down on the lathe and re-tapped with the required M3 thread.

The head cables were then repaired, first by grinding away the coating, exposing the copper tracks within... (The back side had a glob of hot melt glue applied to stabilise the cable during this process and provide support afterwards)

view through microscope

a bit more delicate grinding needed and then would be ready for the next step... more to follow!

 

[netsol]

VERY impressive!

 

[kehall]

Solder bridged the gap on the first repair... Covered with a UV cured conformal coating and additionally hot melt glue to both sides of the cable for additional mechanical support.

The second cable repair wasn't so easy, as unfortunately the gap was a little large for solder, so had to use some extra fine wire in addition, again followed by UV solder mask/conformal coating and hot melt glue

Decided I would also bolster the 3rd (W/W) head with hot melt glue too, whilst it wasn't yet faulty, felt only would be a matter of time, so that head had to come out... head screws stuck in that one too... spent a good 30 minutes dribbling solvent down and gentle attempts to free up but nope, SNAP! (or rather a squelchy sounding/feeling "snap") those screws sheared off too... ARGHHH! So now had two brackets to drill out and tap, and another set of screws to re-machine threads on.

Unfortunately didn't get any photos of the drilling out/tapping process but suffice to say one went well, the other was slightly offset, but didn't feel was too much of an issue... Foreshadowing?!

A small tool was fashioned out of a bit of thin metal to allow the replacement of the bracket/spring affair without removing the alignment post (thereby retaining alignment - although maybe will have to check/adjust anyway!)

The spring sits with the back in the lower half (in the printer) and the front part of spring in the upper part (the bracket that is going back in). So the tool was needed to compress the spring so the bracket can go in flush (otherwise spring is in the way), and then some manipulation to push the rear of the spring down into the lower half before releasing and hoping. It took several attempts to get right, but easy to know if it's worked or not as the bracket can be pushed/slid back and spring back forward when released. All that to avoid disturbing the alignment post

Some fresh white lithium grease was applied to the sliding surfaces and screw threads and got the white bracket and head replaced first, all went swimmingly. Then the middle bracket was replaced, followed by the left hand head (no problem there after clearing the remaining muck out of the threads with the M3 final tap).

The middle head... I couldn't get it secured The slight offset in the drilling was just a bit too much for it. ARGH! Spent about an hour getting those brackets back in place and it had to come out again for more engineering!

Managed to pluck out what was left of the threads of the original screw (inset) from the bracket, and thankfully, even though the hole was now a tad wider, the head was able to be secured once again into the bracket (checked outside the printer this time!).

Back in the printer it goes, and now it boots up without head errors, yay!, and I now see a !WORK High message, but I can go through various menu options, test sensor etc. Turns out the !WORK High message was down to my forgetting to plug in the CN27 cable on the slider relay PCB, doh.

Now waiting on new dampers before being able to do much further, and a carriage/Y belt on the way as there's a section of missing teeth (plus the repairs and cleaning to the station/peristaltic pump assembly) ! So more to follow, if anyone's interested!

Keith

 

[kehall]

VERY impressive!

Thanks! It's quite a journey, but hopefully will be of help to someone else too. Thanks for the information on the butyl carbitol, seems we can get this as Butyl Diglycol here in the UK. I wonder if that's what is used in the 'flushing/storing' agents with a slight oily consistency.


I guess I should also remove the battery to reset the clock (although I don't really want to lose any data), as I suspect the ink may well have 'expired', and I have no way to read or write the date on the ink chips at the moment, I understand there's a way to do such things but information isn't that easy to come by (understandably!).

In the meantime does anyone know if there's any data written to the ink chips that would permanently disable them post-expiry date, or can I risk putting them back in and booting up without resetting clock (as I don't want to lose any other configuration data)?

 

[Smoke_Jaguar]

Had a MK1 6042 in similar condition, scrapped it. While it uses Gen4 heads, you MUST get the long cable versions (basically the Mimaki version) and those set you back a couple grand each. Getting the head bolts loose is a bloody nightmare and the dampers are tricky to install right. If time isn't an issue, you'll be fine. Can probably restore it for reasonably cheap. However, the thing is going to be a nightmare with ink wicked into every corner inside the printer.

Head holder screws can be found here. https://www.digiprint-supplies.com/en/ujf-3042-r-head-holder-screw-m206792-pscmi18213

If you break a bracket, this guy https://www.digiprint-supplies.com/en/ujf-3042-r-head-bkt-ang-m400437

Order a few extra, they will probably all break, and new ones might snap too if you have my luck. Might be a good idea to source a tap to clean the threads as well. Grease the screws before installing to protect threads from ink buildup.

As for chip data, the 13th hex character is going to be date. Should be good setting it to 19 or 20. In 6 months, +1 it. Or just set date earlier. Just back up all data on chip and restore to reset the chip back to the ink level state it was in when backed up.

 

[netsol]

you and smoke jaguar make me feel like a rank amateur.

 

[Smoke_Jaguar]

you and smoke jaguar make me feel like a rank amateur.

Just wait until you see the crap I'm doing with the JFX500's in the shop!

 

[netsol]

Just wait until you see the crap I'm doing with the JFX500's in the shop!

Can’t wait to see it.
i am the veteran of probably 100 acquisitions & repairs/rebuilds, but, often take the “easy way out”
buy 2 machines & make one good one

 

[SGC]

So ready for more

 

[Smoke_Jaguar]

So ready for more

More what?

 

[SGC]

More what?

The story/journey…

 

[kehall]

This evening was a bit more cleaning ready for the pump repair (waiting on the replacement tube hopefully tomorrow)

My hands have been a bit of a pain recently, I think down to sweating in nitrile gloves for 2-3 hours (not contact with ink), developed an itchy nobbly rash, so trying to limit the time a little. Although that plan didn't last long this evening as I decided to clean out the station/capping area, I put some cotton gloves to help but not sure they did much to be honest! So have taken care to gently wash afterwards and apply some cream/vaseline and presently wearing some cotton gloves so I'm not greasing up the keyboard and mouse! Hopefully will help.

Installed the new FFC cables in the slider and channel adding a bit of additional securing tape (3M 8915 which I have plenty spare from my day to day product manufacturing).

The new FFC cables are about 10" longer than the old so they were gently folded around the cavity at the bottom of the slider relay pcb area

Despite the serial number being an early, affected model, the service action to protect the FFC cable from the screws on the cable clamps was never performed - meaning the protruding screws from the ink line holder (circled below) can damage the FFC cable. I applied some protection (a bit of old PVC cable jacket) to the screw threads (really the screws are just a bit too long) and a bit of strong tape to prevent such an occurrence, to the clamps at both ends.

The station was released (three screws and a few cables disconnected), and brought forward to reveal the separated pump housing sitting at the back.

The top plastic cover that helps retain the head capping setup has broken in a few places around the screw areas - a replacement would be ideal but will attempt gluing first (after a good wash - I have a crude IPA pumped wash bath used in my Formlabs 3D resin printing setup - I may even possibly 3D print a replacement top cover if I can be bothered spending the time on the measurements and design).

Below is the other bit of the pump head attached to pump head plate. I couldn't precisely see how they were held in, it looked like a single clip (turned out to be correct) but didn't like how it was feeling trying to remove them, didn't want to break anything further.

Does anyone know what the absorbent matting material is, and if it can be purchased in sheet form anywhere that can be cut to size? I'd like to replace those if I can and not at Mimaki prices as they appear to be totally saturated. It feels and behaves the same as the wicking material in permanent marker pens etc.

...

 

[kehall]

Getting the pump head plate assembly out wasn't particularly easy, the capping section had to be unscrewed to make it easier (had to come apart anyway for cleaning)

Separated the pump heads, and flushed through the peristaltic tube sections with ethyl acetate just to clean up a bit so don't get ink everywhere when removing the barbed connectors to replace the tubing!

The piece of tubing laying loose at the bottom and extricated earlier for measurement of replacement hose (Platinum Cured Silicone, 3.2mm ID, 6.4mm OD was ordered) has a nice split in it:

The caps (cleaned a little) and capping assembly separated for attempted repair of the top plastic plate... They appear to be in pretty good shape. will clean the springs and apply a little grease on reassembly.

The top plate that's split in several places, going to attempt gluing it first (after decontamination!), as it's not difficult to get out again if a more durable repair is needed.

I wonder if the plastic here has been weakened/attacked/made brittle by cleaning solvents or possibly ink itself along with the spring pressure (interestingly the dark grey paint coating on the printer inside panels peels off a bit in places too where ink has sat presumably for years).

to be continued...

 

[netsol]

we recently had a leak in one of the pumps on our mutoh hybrid.
the cleaner (ink with no pigment bubbled the s**t out of 70 year old concrete floor paint (that's how we noticed the leak)

you should see the things your PRIMER will eat through

 

[netsol]

This evening was a bit more cleaning ready for the pump repair (waiting on the replacement tube hopefully tomorrow)

My hands have been a bit of a pain recently, I think down to sweating in nitrile gloves for 2-3 hours (not contact with ink), developed an itchy nobbly rash, so trying to limit the time a little. Although that plan didn't last long this evening as I decided to clean out the station/capping area, I put some cotton gloves to help but not sure they did much to be honest! So have taken care to gently wash afterwards and apply some cream/vaseline and presently wearing some cotton gloves so I'm not greasing up the keyboard and mouse! Hopefully will help.

Installed the new FFC cables in the slider and channel adding a bit of additional securing tape (3M 8915 which I have plenty spare from my day to day product manufacturing).

View attachment 172385

The new FFC cables are about 10" longer than the old so they were gently folded around the cavity at the bottom of the slider relay pcb area
View attachment 172386

Despite the serial number being an early, affected model, the service action to protect the FFC cable from the screws on the cable clamps was never performed - meaning the protruding screws from the ink line holder (circled below) can damage the FFC cable. I applied some protection (a bit of old PVC cable jacket) to the screw threads (really the screws are just a bit too long) and a bit of strong tape to prevent such an occurrence, to the clamps at both ends.
View attachment 172387

The station was released (three screws and a few cables disconnected), and brought forward to reveal the separated pump housing sitting at the back.

View attachment 172388

The top plastic cover that helps retain the head capping setup has broken in a few places around the screw areas - a replacement would be ideal but will attempt gluing first (after a good wash - I have a crude IPA pumped wash bath used in my Formlabs 3D resin printing setup - I may even possibly 3D print a replacement top cover if I can be bothered spending the time on the measurements and design).

Below is the other bit of the pump head attached to pump head plate. I couldn't precisely see how they were held in, it looked like a single clip (turned out to be correct) but didn't like how it was feeling trying to remove them, didn't want to break anything further.

View attachment 172389

Does anyone know what the absorbent matting material is, and if it can be purchased in sheet form anywhere that can be cut to size? I'd like to replace those if I can and not at Mimaki prices as they appear to be totally saturated. It feels and behaves the same as the wicking material in permanent marker pens etc.

...

i will give dave from solvent printer conversions a call & ask him. he knows the source for almost EVERYTHING

 

[Smoke_Jaguar]

For absorbent pads, I use bulk oil absorbant spun polyester material usually sold to mechanics. Almost identical, good for ink soak pads, especially when doing rebuilds. Just cut it with a nice sharp hobby knife and scissors.

 

[kehall]

Dampers and belt have arrived! However before the exciting and highly anticipatory printing test can be done (Schrodingers cat anyone, the heads are both dead and alive until observed!), the repairs to the peristaltic pumps is needed and a bit more clean-up in that area first.

I think we may get away running the print test on the old belt, so this will be looked at later, assuming the heads don't need to move to a section of the belt with stripped teeth!

The new tubing arrived today, so cut to size (about 175mm long), the pump head insides greased with a good quality silicone grease, and exercised with a powered screwdriver for a minute per service manual. The new greased unit was quieter than the old ones so the same treatment was done to the other 2, so all 3 are the same with fresh hose.

.

I noticed two of the pump heads were a bit loose, it appears a piece of plastic was broken off which helps keep the unit square when running, can see additional wear marks on the lower right corner too. Additional support was provided by inserting a metal pin in the location of the missing plastic (left mid-bottom of picture) to support when under motor torque.