T-Shirt Forums banner

EPSON SP 4000 Build, Direct drive

[DIY DTG] 
7K views 27 replies 6 participants last post by  pep0na 
#1 ·
Have been thinking about a 4000 build, A2 size printer for quite a while. Followed Customtshirt (Ronnie)s 4000 build and was inspired to get off my rear and get something going. His build is a roller feed design with steel under carriage - Im not as handy with a torch and I had an ideal to do mine as a direct drive using the original paper feed motor and shaft driving the drawer directly. (Cork lined underside) I also felt more confident doing my base in wood because Im good working in wood and the Design has been hammered into bedrock functionality by others.

I plan to use the finished machine doing signage and personalized Items for my business . There are a few pics to show what Ive got going - will have more detail next post.
 

Attachments

See less See more
3
#2 ·
Stripping the printer wasn't too bad - I saved as much as I could - had to cut some of the parts out with a dremel tool - Left the paper exit bar in place and have not removed the head or any of the carriage parts or Ink supply items . The PE sensor rests on the left side of the printer and is driven by a rod that goes through the tunnel and has a gear on the right side. The ideal is to remove the sensor and make a short shaft with the sensor flush against its drive gear- it is driven by another gear on the right side of the printer. Its tricky to align but you must make a stub mount that holds the sensor and its gear in place and still allows your Ink waste tank to slide into its position. I used part of the discarded paper feed chassis to make a bracket. I used the white spacers from the paper feed bar for the tunnel side stop.
 

Attachments

#3 ·
For the drive system I needed to relocate the drive bar and encoder to the wooden box . Was concerned about getting the encoder wheel reader aligned properly along with the PF motor . I used the original mounting plate and holes - just had to cut it out and mount it in the wood cabinet - Used the plastic bushings to mount the pf rod - Drilled holes and used epoxy to hold them in the proper place. You have to play with the alignment until the drive rod turns freely. The end or the pf rod ihas a threaded hole that is handy for adding a hand crank (Friend of mine came up with this ) . Comes in handy for running your table in and out. The encoder wires are a bit short - you will need to splice them with solder and a longer wire and run them through a hole back to the printer board, same with the pf motor.
 

Attachments

#5 · (Edited)
Got it all put together and rewired - the main board sits on a platform behind the frame and above the tunnel . There is just enough cable length to reach the board here - all the ribbon cables are the right length - the tiny pf sensor ribbon is tedious to get away from the frame. There are a few wired that need to be spliced all easily done and no ribbons to bother with . I have the Manual pf switch and the pf photo sensor mounted at the rear right of the tunnel near their original positions . Ive changed the pf switch for an industrial roller switch. Its normally open and makes contact when the table is rolled into its start position, stays closed until the table is anear the end of its run . The pf sensor has a sheet of paper taped to the edge of the table and I have a leading edge board on the table to give the Print head photo sensor an edge to look at .

Had a few error codes but all were minor -the printer powers up goes through its start up, cleans itself and then gives a Ready - I set it to do a test print and it runs the table fwd - checks for an edge with the print head multi sensor, then runs the table all the way fwd and gives either a paper out or insert more deeply message or simply says remove paper . Ive been stuck here for several days now, Not exactly sure what Ive got adjusted wrong . Im a bit unsure of how the sensors are supposed to work . Ill try it again tomorrow. Any ideals ?
 

Attachments

#6 ·
Great project.In which mode you print?Roll or sheet?Did you use printer before conversion?L shaped paper in your case is in the middle of platen and MUST be aligned with begining front edge of platen for multisensor in head to reckognize beggining of paper edge timing.Now is very OFF.Also try to disable multisensor in printer menu because if you need to print dark/black t-shirts you will get "no paper" error for sure.And try to play with timing between manual and PE sensor(try to find right distance between sensors,it`s essential).
 
#7 ·
It was fully functional before the build - Ive printed on it - only has a few hundred prints total - very lightly used machine . Ive been trying to get it to do a test print from the control panel - I got it to go through the motions twice in a row before I mounted the pf sensor and the manual feed switch - didnt have the ink primed in the lines to get any output but it seemed to work like it should. I think the multisensor and the pf sensors are my problem now - I have everything I know how to get to on the control panel turned off - the multi sensor still does its little search for the leading paper edge though - not sure how to turn it off . On the manual feed switch Im not sure if it gets a quick bump at the beginning of print or if it gets actuated for the duration of the paper length you wish to print - The same with the pf photo sensor in the tunnel - I was thinking that it needs to see white , soon after the manual pf switch is made and at the end of the sheet, the white goes away - But this is difficult to understand with the multisensor trying to locate an edge as well - I was thinking that I should get everything working from the control panel on a test print then hook up the computer and test it from there . It will help to understand what the tunnel sensors need to see ( manual feed switch and pf photo-sensor - Its so close to working lol It seems to like a quick bump of the pf switch as opposed to a constant on . Still getting the paper messages on the cp Ill play with more after work - Its so easy to get distracted and spend the day on the printer instead of earnings lol
Ed
 
#8 ·
Breakthrough, Best friend was helping with the build and we were playing with sensors - double checked all the control panel settings - found that I had left the paper position check option on and the printer didnt like it at all . Turned it off and tripped the pf switch - went right into a test print but in Roll feed mode - Now to get the sensors in the right place. Print looks ok but a little faded probably need to do a good head cleaning - trying to save ink till we get it settled a bit better .
 
#9 ·
Ok I think we got it - Multi sensor is completely off, pf photosensor at the leading edge and the pf switch makes gives a ready and puts the printer into roll mode - You get a test print from the front panel - The sensor positions still need work . Wanted manual sheet feed, found that if you open the top pressor foot lever dont use the manual feed PF switch at all line up the leading edge of paper under the PF Photo switch, carriage back inside the tunnel at start position , power up the machine, it does its start up routine settles in sheet mode with a Ready indication on the control panel . Select Test print the machine does a beautiful test print without the pf switch ever having been made . Now we have sheet feed mode . Have more work to do on setting up the proper positions of the sensor - Tomorrow we will get this hammered out . Have found that manually running the carriage back into the tunnel to start position with the printer turned on somehow freaks out the encoder and gives you a reinsert paper indication that you cant clear without restarting the printer . hmmmmm.......
 
#11 · (Edited)
Top pressor foot is simply the input paper feed roller release . You must open the lever to load either roll paper or manual feed cut sheets. Still have a bit of work to do before its ready to do a show print . I printed several test pages in sheet mode then tried to print from the computer - evidently it mixes cmyk to get black because all that came out was faint blue lol - I need to sort out the ink lines and fish out the gunk from the lines . We adjusted the photo sensor to get more print area on the table but now have issues with paper feed - have a Remove paper error . I believe the small ribbon cable has a broken lead or the small board has a cold solder joint . we are investigating replacement of the photo sensor with a roller switch - should be more reliable if it can be done - I need to spend an few hours with the wiring diagrams . Ive also got the Ink counters turned off now . Ill post more once we get this sorted out . will do more pics next posting .
Ed
 
#12 · (Edited)
Ive found the source of my paper sensor problems - Loose ribbon cable on the sub b board that the paper feed sensor connects into . The thing was very intermittent before - now I can get a consistant signal from the paper edge transition . I burnt a lot of time figuring out the workings of the Omron photo microsensor so I could use a roller switch to replace it . It has 3 pins - 1, 2 and 3 --- pin 2 is ground . I suspect pin 1 is the output and 3 is the vcc and shorting pin 1 to 2, with a switch will mimic the photo sensor - there is a load resistor in the circuit on the main board CN54 connector pin A =1, pin K=2 and pin 3 = C--- pin A is the PF adjust, pin C is PF sensitivity and K is chasis ground . Havent decided wether to do it or not. If Im wrong I might pop part of the main board . the sensor works good now so I may leave it be . Just need to lock it down and tidy up and then freshen up the inks . and were ready for testing. Didnt get any photos tonight - I want to finish building the main board case so I can show that . Sure do learn a lot about how these things work :)
 
#14 ·
I wont need a Rip until I do the next conversion . I wont be using tshirt inks in this one - its for signs and promo items . The epson drivers have a pretty decent package in the software that will allow you to vary the color Saturation and dry time between head passes - also the feed rate and others . It will serve me well just as it is. ( Wish White ink( Non Garment Ink)was available in standard pigment inks for the epson).

On the next conversion I will do a uv machine - that one Ill need white ink and a Rip to lay it down . I contacted the EK Rip people and they have an older package made for the 4000 . It will do duak CMYK or CMYK and 4x channels of white . The cost is a bit high at 800 us but its available . Strangely there are not a lot of Rips that support the old 4000 . You would probably do better to convert a newer machine with better Rip support. I have 4 of these on hand so its a no brainer for me .
 
#15 · (Edited)
Here's a Video , Very happy with the machine. I still have some cosmetic touches to add, 1st attempt at video but I think it will do. Printed on poster board with Original Epson inks. We built a scissor jack out of pvc pipe but it was too unsteady and caused a bad image - the box is preset to the right height for signage. No lines blurs or distortion of the print . The handcrank works great but you need to put a switch in the power leads for the paper feed motor . This printer is set up for manual sheet feed - It can work in roll mode as well but I like to keep things simple . The drawer slides can support quite a lot of weight without bogging down the paper feed motor . The reqistration is perfect -there is a hard stop at the rear to start the table at the same place each print and there is absolutely no slip with the drive bar and cork lined base. You do need to use MDF for the drive surface - it must be perfectly flat and level on the bar to prevent slip .

http://youtu.be/XWviC5c-Pbw
 
#16 ·
Good job!Thanks for video.In future you should consider to use aluminium with glued cork on shaft side instead of wood for t-shirt platen if you stick with friction mode .Wood is too heavy for Epson`s small DC motor.For few prints is O.K. but if you have series of prints = motor would easy overheat.I see that few guys managed to use Arduino PIC with big steppers for automatic platen transport system.And my question is how do you managed to print from Sheet Mode?Switches order and position etc...?Please explain with photos or new video.Thanks!
 
#17 · (Edited)
Hi Stampac, The weight is not a factor on this build, the drawer slides are heavy industrial grade ball bearings, top and bottom on both sides - the Brackets are mounted to the slides and adjusted by the milled slots to place just enough pressure on the drive bar to remove any slip -The cork is self adhesive mounted to the bottom of a 1" Mdf board - the pressure on the drive is the same . Might have troubles from the torque required, if you ran something heavier than say 30kg (60 pounds) I have not tried to test the limits- the movement is smooth and has no binding at all . It does take patience to adjust the rails and the brackets to insure nothing binds - We had to take the slides apart, remove the back stops and Buff the metal surfaces with a polishing wheel to make it better.
I used the wood because it is readily available, easily worked and if I make mistakes, its low cost. Good heavy Aluminum plate is very dear here . You could do the whole project in metal for better durability and a more professional job - On Servos, I needed the output of the printer to be exact with no dimensional deviation from the print input - I read German13 s experiences with direct pf drive vs Servo,s and belts - Said the output of a servo / belt drive always has a slight distortion because you cant get the stepper and belts to be exactly the same as the original pf motor and shaft .

Sheet feed mode - Photos are in the next post . I will say that when you set up for a print you must open the top cover, open the paper feed lever on the top right, turn off the power to the PF drive motor You see a small on/ off toggle in my Video ., roll the platform back into the machine, turn the pf drive motor power back on , close the paper feed lever and then close the top cover . the printer will do its thing and move the table forward into print position - it will then set itself on roll mode and you will need to select sheet mode at the control panel . Its ready for a print from the computer now .

You dont use the rear upper manual sheet feed / roll paper switch, only the photo sensor, and you must trick that by giving it a strip of white paper with a couple inches of black electrical tape at the start about 11.5 inches of white paper and another black strip at the end of the white . The photo sensor starts out with nothing under it, this makes it prompt you to load a sheet of paper ( you open the cover raise the paper feed lever then close it and all the way look for paper by moving the pf motor fwd till it runs onto paper strip - it then finds the start of the paper by moving back inside the tunnel a little. - it needs
to see white at the start position and black when it runs out of print table . The other thing you need is the paper exit bar - it has a sensor on both ends so that something can trip the sensors and stop the machine if its too high for the heads - if you cut this out of the tunnel you will have problems with sheet mode - Just remove the little spring serrated wheels and leave the bar in place so the gears can turn it and the sensors will be happy . There is also a set of height sensors attached to the wheels that originally held the paper against the drive pf bar - I just left this in place as well - Ill post pic's later hope that is helpful -

My next step is to do another of these and convert the Ink system for LED UV - Ive seen so many melting issues with the standard ink systems - Even some of the the commercial printers require the dampers and capping stations and lines to be replaced frequently - I need true solvent grade parts for the 4000 to do this one .
 
#18 · (Edited)
The first pic is of the pf photo sensor mounted on a magnet so you can experiment with position . When your table is at the start position inside the tunnel the pf sensor should be over black and it will move onto white paper strip - Anything shiny white will not workand will give a false reading .

Pic 2 is the actual photosensor - The output is on pin 1 and the sensor shorts pins 1 and 2 together to register as paper under the sensor - you could use a simple switch across these points and have it make or drop out as the table moves past . Pin 1 is 3.3v and pin 2 is ground in the circuit -a pull up resistor on the main board keeps the voltage from short circuiting when the sensor operates .

Pic 3 shows the rollers that press against the original drive roller - these are attached to the thickness photosensors and I left them in place .

Pic 4 shows the rear upper manual feed paper switch or the Roll feed switch - This should be left out of the tunnel if you want to use sheet feed mode . Mine is hooked up but its just not used . If you wish to operate in roll feed mode this is the right switch to use .

And last, is Pic 5 . This is one of four adjustable brackets that allow adjustment of your table with the right pressure on your Pf motor drive - You want the drawer slides to carry all the weight and keep the load off your pf motor . I milled a proper slot in a piece of extruded steel angle . This build is not overly difficult. You will need to fabricate your base and some simple sheet metal components . The table has plenty of room to handle large items or Garments - The platents would need to be fabricated for you own needs . Be very careful with the small ribbon cables they get loosened easily. I need to mention that the ink flush tank on the capping station needs to have a longer tube inserted from the white stub to the waste tank else the ink will run down the inside of your machine and make a mess . Note Pic 4 is of the roll feed switch on an unmodified 4000 - wanted you to see what it looked like before its abused .
 

Attachments

#19 ·
Pic 1 and 2 are a rotating bar that has 2 photo sensors attached inside - I'm reasonably sure that this is a pain in the rear if you remove it . The bar stays hooked up in place and it rotates while the machine readys itself . IThis is the bar that has the serrated edge wheels I mentioned earlier - I removed these . Dont confuse them with the gray wheels that hold the paper to the feed bar driven by the encoder .
 

Attachments

#20 ·
A couple of notes on this build . You need to make sure the drawer slides are smooth , I took mine apart and polished them with a buffer . Use Silicone lube to keep them working smoothly. It takes a while to get the table brackets adjusted and level on the drive bar but once done The printer has exceptional precision and will register a print in exactly the same position over multiple prints . The Slides will support a great deal of weight as long as you dont drop anything on the table and cause the bracket bolts to slip and bind the table against the drive rod . I am printing in manual sheet mode, not in roll mode . Next project will be an LED UV conversion using the same platform Epson 4000 .
 
#21 ·
Printer is working great , A few Mods were made to improve operation . The original brackets on the carriage had no slots to allow adjustment between the slides . We used 1.5 " aluminum angle and milled slots for the base as well as the single slot for the rail bolts . Also increased the size of the bolts on the slide rails to 1/4 inch, The old bolts were very small and I stripped them out . I had to mill the bolt heads flat so they would have clearance inside the slide rails - also spot welded them with a wire feed welder to hold them in place . You can use a 24" slide but you will need to cut away the rear slide stops so the slides have greater range of movement I get 48" .

The cork lining on the bottom of the board will become wavy as it is abraided . You will need to sand the waves flat and re adjust the tension of the rails on the drive bar after a while . Too much tension will give you a servo motor error, Too little and your print output will be off. The rough coating on the PF drive bar is very important and the soft cork ensures this coat remains intact .
 
#22 ·
Its been quite a while since Ive posted here and I have learned a tremendous amount about this style of build and my printer in particular . This build functioned perfectly for several months and I started having slight errors in registration a couple thousandths of an inch. I use a fairly heavy table on the drive bar / paper feed bar and over time this has worn out the heavy drawer slides . when the table feeds out the front of the printer they bend slightly and cause binding . This was fixed by ordering a heavy set of truck slide rails rated at 500 lbs . These rails extend out the front of the printer box and are full extension vice the 3 part lite rails we first used . Works great now .
 
#24 ·
Michele , Looking at the picture you sent - You have removed the important bits where the sensors are mounted and you need to decide how you will move the carriage under the print head . My system is usint the original paper feed drive shaft and a cork lined table . that rides on the drive bar . You will also need to trick the paper feed sensor. It is driven by the gear train with the sensor opposite the gears . You will also need to relocate the Control board above the tunnel of the printer .

Others have adapted the 4000 and 4800 using other systems like Arduino control drive motor but I like high accuracy movement so I went with the original paper feed motor and drive bar . I have asked the gentleman at UFO about adapting the 4000 for his base and he does not know the printer . I wouldnt recommend spending that much to adapt something that might not work well . You need a copy of the printers repair manual to understand how it works before you do the build . Some of the parts you removed will need to be re installed if your going to do the same build I did .
 

Attachments

#27 ·
Ive had several months running this conversion, Have rebuilt the lower end with Truck slides rated at 300 lbs , also have to adjust the tension of the cork against the drive bar occasionally, too loose and the platent slips ...too tight and you get a servo maint error . Another issue affects the accuracy of registration and that is the wood base changing dimensions as the weather changes, humidity etc . The ink system is rock solid as long as you run the printer daily. let it sit over the weekend and you will need to clean the head a couple times . Overall its a good build but for better registration the base needs to be metal .
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top