Epson PCB Printer Hack/Mod
- March 6th, 2011
- Posted in Uncategorized
- By rucalgary
- Write comment
So I went to my local Futureshop (Canada FTW!) and picked up the only clearance Epson they had… No need to buy the extended warranty here!
Brand new:
After taking off all “Unnecessary” parts:
I removed the paper feed and paper out trays, you can see the scanner bed on the top left, and the power supply left. The paper detector is also seen poking out of the left side of the image. Next is when you remove the head mechanism and raise it:
All I did was use some slightly longer #4 machine screws and pop in a few washers… Did this to both sides where the head rail attaches to the chassis.
*EDIT*
I forgot to add a CRUCIAL part to this build. The printer Head rest mechanism must also be raised by the height of the washers…
This is the infamous mechanism here: 
As you might be able to see, there is a little metal pin sticking out just under that little arm. That was added by me, as after I added height to the tray, I also had to lower that pin so that it would continue to interact with the Jam detection mechanism…
To raise it, I dremmelled off the existing pins and drilled holes just beneath the previously existing pins. I then inserted drill bits (as they happened to be the right diameter) and hot-glued them in place. I had a bit of an issue with friction, but after a little WD40, it was as smooth as butter.
*END EDIT*
I then proceeded to add the aluminum railings for the feeder tray:
I also had to add the paper detection mechanism right after the feeder rail.
This: 
ended up going just after the feed tray on the left.
I also removed most of the feed rollers so that it won’t touch the board and mess with the new printing. Now it only has the most outmost feed rollers.
I was also able to replace the casing once I added a few washers for spacing…
Yes, I kept the protective plastic on…
For operation!
To actually run this, I found that a steel tray was too heavy to be reliable, and cardboard, while light enough, was not reusable enough, and tended to warp under the heat needed to pre and post heat the board. An aluminum sheet was found to work the best!
I also use yellow MISPRO ink in a black cartridge as I found it to resist the etchant the best. It can be found Here. Don’t forget to buy the empty black cartridge!
If you take a look at the bottom left of the tray, you will see a slit in the tray. this is to provide the delay needed for the paper sensor to read “Paper”. I will place the empty tray in the printer, heat it up with a heat gun, run off the pattern, and that will tell me exactly where I need to place the board. I place the copper clad board onto the feed tray, and place small board bits and tape them around the board I am printing on to keep it in place. Once I am done, I etch the board, and wipe up any and all ink up with a touch of acetone and paper towels.
I plan on adding video after my next project.
Let me know if you have any questions so I can update if I missed anything!
*EDIT*
Just finished a new Board, so decided to share the actual creation process!
Result:
And for those who asked for the front tray:
This is very awexome! So that’s an inkjet?? Looks too small to be a laser. I’ve seen this done with Laser printers but not an inkjet. I figured inkjet ink is water soluble and wouldn’t resist very well. I would totally do this if I didn’t have a decent method already.
Standard inkjet ink is water soluble. Pigmented ink is required and in most cases will have to be placed in a cartridge manually. Yellow ink seems to have more pigment in it than other ink colors, which makes sense when you think about it.
It is indeed an inkjet! As it so happens…
Epson printers use their own fancy micro-pezio electronic print heads. This allows them to use Pigment based inks instead of Dye based inks like most printers. What this means, is that the inks can be cured with heat after printing to make them water insoluble! (or you could wait a couple months for them to dry…) However, some inks tend to work better than others, and I found it best to use the yellow ink from the refill company. I just refill the black cartridge though so I don’t have to worry about trying to print the perfect yellow.
Very interesting! Thanks for sharing this build. I have a brand new epson printer cx8400 that could potentially work for this. What do you think? http://img263.imageshack.us/img263/8931/epsoncx8400.jpg
You could probably do it. I will give you a quick warning though about one thing I forgot of when I was writing it up.
The head rest mechanism has to be raised as well once you raise the head rail. This is so that none of the ink dries up, and the jam detection mechanism (the white little square thing near the rest position) operates properly.
This was probably the most difficult task of all. what I did was remove the little tabs that stuck out initially, and then added my own by drilling holes just underneath the last tabs, and hot-gluing drill-bits in place. Then I did something similar with the tab protruding from the base of the head rester and added a shortened drill bit there to lower the pin that sticks into the jam mechanism to alter it’s operation.
I can’t believe I forgot about this, but it seems I forgot to take pictures. I will have to add this in as it is essential for actual use.
That, and if you can’t get your hands on MISPRO ink, the next best is either the Epson yellow, or Epson magenta.
Nano also makes a pigment ink.
http://www.nanodigitalink.com/products/displayitem.asp?pid=749
I want to try and do a silkscreen with this method after I get the PCB thing working. About to start modding a canon printer. I’ve heard most people mention that they don’t use them because of the paper sensors involved. You mentioned the epson print heads being special. Should I expect problems with canon print heads and pigmented ink?
UV silkscreen:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=170505715005
Fake silk screen that works (heat fixed, really a stain glass paint, lots of color options)
http://www.theartshop.com.au/category497_1.htm
Depending on the printer… It needs to use a non-themal method of printing. Most printers use thermal printheads that heat the ink to surface-of-sun temperatures, causing it to expand and squirt onto the page. Pezio-Electric heads use (exacly as the name implies) tiny pezio devices that push ink, allowing them to use pigment and not dye ink. So just do some quick research on that printer to check. I know of some HPs that used a non themal mechanism, and I think Canon may have dabbled in them, but Epson doe almost exclusively pigment ink, so it is almost always a safe bet with them.
What model of printer did you use?
nx215. I think they made one without the scan bed, but this was the one on clearance.
Great project! – could you put a picture up of the exit feeder (looking into the printing bay)? Or do the vertical edges of the front guides just simply continue from the back tray?
Just Did right now. Also Posted a new video of the actual process…
Nice. I might suggest adding some registration/alignment marks to you image prior to printing. Would assist with making sure everything was properly aligned. http://en.wikipedia.org/wiki/Printing_registration
I am not sure that would help… I am only using one color (yellow but it is in the black cartridge) so I have no need to make sure that the colors align. Thanks for the suggestion though!
Hi,
Great project! I have an epson very similar to your’s. The black ran out and it had a chip on the cartridge itself that, I assume, counts pages printed which shuts it down after so many prints (damn consumerism). I suspect your cartridge does the same. How do you get around this?
I’m gonna do the same that you did now that I have a different printer
you should integrate the scanner into the project somehow. Maybe make it scan in unpopulated damaged boards that you do not currently have artwork for? Be a way to import and fix/clean damaged traces and re-print on a fresh pcb…
The Ink cartridge has a chip that resets itself once you start the change cartridge routine and place the cartridge back in apparently. I have yet to “run out” of ink though, so I am unsure of the exact workings. It is warrantied by the guys who sold me the ink refit kit though (The MISPRO guys) and other have used them with success so I guess I will see.
I like the idea, though I am afraid I don’t have any boards that I would need that done for. I have used it for some legal documents though, so who knows. I do know that without the scanner, the printer reports an error, so it was a necessity that I left it in one piece.
Greeting from Argentina! First of all, great project! I’ve been looking to modify a printer myself for some time now and I haven’t found a “normal printer to Awesome PCB printer” modification tutorial that is “idiot proof” yet. It seems that every other tutorials uses some extra sensors and some microcontrollers to function, yours seem simpler, that’s why it looks so promising. I would love if you could post a more extensive step-by-step tutorial on how to modify the printer so the “DIY impared” like me can replicate your results. Thanks and keep up the good work!
I have been playing with the Idea of making one for my engineering department, so if I have some spare time one of these weekends, hopefully I will be able to document it really well.
What model epson printer are you using? Seems like there are a lot of missing details. Might be a little tricky for someone else to replicate your work.
Epson NX215… unfortunately during the build it never came to mind to really document it. I plan on doing another similar mod for my engineering department sometime soon, and will be sure to document everything.
The link you have provided to the ink is for a $99 kit which includes all the cartridges, not just the black cartridge and ink. Does this vendor offer just the single ink color and black cartridges? I did not see that they did.
I see you raised the print head to accommodate the thickness of the circuit board. But you also show the plate being printed and then you tape the circuit board to the plate and print again. There is obviously a difference in thickness between the plate and the plate with a circuit board taped on. So, how are you able to account for the difference in thickness without the print head slamming into the side of the taped on circuit boards?
Also, it looks like the rails do not go all the way through the printer. It appears one set is screwed to the chassis on the front. And then another set of rails are attached on the back somehow (no pic for that). Seems to me there could potentially be a feed alignment problem here.
Looking forward to the updated documentation. FYI for anyone interested, this printer is available new on Amazon for between $78-$120.
The MISPRO ink guys do sell individual cartridges and ink refills, (or at least they did)
The head just has to be raised enough so that the boards+plate thickness can pass underneath. The plate happens to be close enough to the heads for an ok print though.
Correct. The feed rails do NOT go all the way through. They are only really needed for the start of the print as the friction of the feed rollers stop any lateral movement or shifting of angle.
Here’s the link to the 4 ounce mispro yellow
http://www.inksupply.com/product-details.cfm?pn=MISPRO-4-Y
I just looked through my collection of used computer parts and electronics, I have an NX100. Wonder if that would work for this application?
I just looked over the specs for the NX100, and it seems to indicate that it would support this type of mod. It uses the micro-pezio tech, and uses pigment ink, so if you have nothing to loose but an hour or so, there doesn’t seem to be any indication that it wouldn’t work. However, I do recommend looking at how the print head rest mechanism works first, as it is the most troublesome component to “fix”.
I did like you. But the problem right next to ink Although water-based inks (Durablit) was not effective.
Copper pattern need not occur because the acid bite
But the good thing I like most is a small pattern. Is small up to 8 mill.
Tell me what you use ink with the print head does not make tons.
I made a version Epson T10.
you should either use ONLY the Durabright Magenta color, OR you can use (the better option in my opinion) the MISPRO Ink (for me that is http://www.inksupply.com/printer_products.cfm/p/Epson_NX~Series_215.html).
After printing, the ink must be baked on somehow. I use a heat gun to burn the ink on. If I don’t do this step well enough, I will usually get tiny pinholes in my print.
I have also recently started pre-baking my boards, which seems to give even better results for some reason. I might recommend that you try that as well.
Greetings. Thanks for a great tutorial. I modified a NX215 based on your tutorial, set the paper detector int eh back of the unit (where paper goes out). I use a14 inch piece of PCB (60mil) as a holder. I do have problem with acid eating the design even after baking when using single pass print so I am trying to get multiple print pass with heat gun baking between them. I found out that if I do not use the heat gun (cold print), registering of multipass is perfect but as soon as I heat the board, the registering gets off by many mm. It looks like the more heat I apply, the bigger the registering error.
Are you doing double sided board and how do you perform your registering ?. Do you has the same heat issues ?. I have another NX215 for spare parts and replaced both paper detector and Y axis encoder, Keyboard module, main CPU module and power supply. I cannot get rid of the heat induced registering error.Using MISPRO also. Are you using multipass or do you have success with single pass ?. How do you prepare the PCB before printing ?
Thanks again for a great enabler idea d:-)
M.T.
I have noticed issues with multiple passes. I assume it is either due to the heating of the carrier OR the metal rod pulling the tray through. I find that cleaning the board in certain ways can increase the likeliness of a good etch. Are you using Ferric Chloride or HCL etching? I have read on the InkJet PCB mailing list on yahoo groups that HCL can be too rough and eat away at traces. I have always used Ferric Chloride so I don’t have personal experience with this. I have also read that a brake fluid (dot3) pretreatment before printing (without preheating the board) supposedly results in a better etch, but I have yet to try that either. I have attempted multiple sided boards without success, so I have stuck to single sided boards as a result.
Hi,
I am trying to get Epson Stylus T13 (which is easily available in my place through ebay-india). Will this printer work for this mod, can you also let me know where I can get the spec of the printer for figuring out by myself.
unfortunately, I can’t make any guarantees. http://www.epson.com.ph/epson_philippines/printers_and_all_in_ones/inkjet/product.page?product_name=Epson_Stylus_T13 the ink that that printer uses isn’t listed in the replacement cartridges that are available from http://www.inksupply.com. I have heard of people successfully printing boards with the Epson Magenta though, so it may work.
I’m working on a build like this currently (on an Epson C86) after seeing your project featured on Hack A Day a while back and I had a question for you. I’m setting up a rear rail system similar to what you have in the images above, but I’m not sure exactly what size to make it. So my main question is what are the dimensions of your aluminum carrier tray? I read somewhere that the paper sensor needs a gap approximately 3.5 inches long so I was thinking the carrier should be about 14.5 by 8.5 is that about what you have been using? I’m planning on posting my build on my own website once I get it up and running. I’ll make sure to link back here so people have plenty of info and help between the two of us.
my carrier was the size of aluminium that was available to me… 8″x12.5″, so pretty close. 3.5 inches seems about right for the feed gap. I didn’t know if it would be different for other printers or not, and would also depend on where the sensor is relative to the feed location. Sounds Good! I look forward to reading your build! (PS. if you live in Canada or the States, be sure to submit it to Instructables under their epillog contest/challenge.)
So I finished my build and printed off two nice boards that I wanted to test. Here’s where I ran into trouble though. I’m curious what type of etchant you use on your boards and if that makes a difference in your experience? I ask because I tried to etch my test boards in a hydrochloric acid/hydrogen peroxide mixture and ran into some problems. Basically all of my Mis Pro yellow ink flaked off in the etching solution and I stopped etching them because of it. The prints were exactly what I was aiming for and I preheated the board and heated it up to set the print after. The boards actually sat overnight too because I didn’t have time to etch them yesterday. Any advice or thoughts?
It seems I somehow missed this comment… anyways, Ferric chloride is the etchant I use, and I have heard that the HCl mixtures tend to be too rough on the resist and almost never work. I would try FeCl if it is available, otherwise you might want to try diluting your HCl solution.
HI!
What does the 3.5 inches feedgap mean?
I built a PCB printer like yours and I am having trouble with timing of getting the aluminium sheet into the printer…
Printer alwas rejects the sheet I guess because it arrives too early or too late… It is just a guess.. I am not really sure why…
How do you manage the feed?
Thanks
Andy
I was referring to the little slit in the tray that delays the paper sensor detection. depending on where you placed the paper sensor the gap may have to be larger/smaller. You can determine your gap needs easily by taking a piece of cardboard (cereal boxes work great for this) and start with a small slit. Gradually increase the slit, pass it through, until it successfully prints. You can then measure the gap and use that size on your aluminium tray. I would also double check that the sensor is actually being passed through to make sure paper is actually being sensed.
That is a bloody good idea…
The only problem I might have is I noticed on my modded printer later on after I left my previous message.. that I need to activate the paper feed sensor when my printer turns the main paper feed roller to the wrong direction very fast for only half a second or less.. So if I push the tray into the sensor during that very short time interval while the main paper feed roller turns backwards then it will try to print onto the tray…
So I might have a problem with your version because of this…
I was actually wondering why your tray had that piece missing…
I thought that piece of aluminium sheet was hanging around at home (in your home) and it was good enough.. although it wasn’t a complete near A4 size…
Thanks anyway… I will try your way as well because my way doesn’t give me consistent results…
Just two more things…
Is your tray longer than A4 length (297mm) because of the way how you do the tray detection?
My printer has been left without the properly designed rubber seal mechanism inside the printer for a while.. because of the experimenting..
So now it is in dried up condition..
I have tried to clean the head by the softwer driven “Print Head Cleaning” several times.. at leas 15 times after each other.. It didn’t help…
Do you think I will have to throw the printer head away?
Thanks Again,
Andy
The entire tray is a bit longer than a standard sheet of paper, but the distance after the slit to the long end is actually a bit less than A4 length. My printer also does the forward-back-forward bit near the beginning, as the drive mechanism to load paper did a different function when moving in reverse before the printer was modded. I have seen people recover their printer heads by using some of the tricks mentioned here: http://www.inkmagic.com/epson_clogged_printhead.htm, although I would use Kimwipes (http://kimwipes.org/) instead of regular paper towel as they don’t leave fibers where you rub them. You can find those at electronic component stores and laboratory supply stores.
HI!
Your tray cut out for the paper feed sensor has worked great with my printer as well… Thanks,
Does you printer work fine with MissPro inks? Have ever used any of those?
Do you use auto resetting cartridges?
I do use the MISPRO inks, and use their resetting cartridges.