Resurrecting a Dead 3D Printer




Introduction: Resurrecting a Dead 3D Printer

Resurrect a dead 3D Printer? Why?
Sure, these days, if your printer reaches the point where you no longer want to mess with it, or it's just beyond your capacity to repair, you could tear it apart and use its parts elsewhere. In my case, I had a Solidoodle SD3, my first operational printer, which served me faithfully for 2 years before 2 Sangui boards and 5 hot ends failed, eventually had me throw my hands up in frustration and the printer went to the garage floor. Fast forward to 2 months ago and I met Ray Kholodovsky (via Google+) of Cohesion 3D and we got to talking....and talking...and next thing I know I am agreeing to test a Smoothie based remix....and so, the SD3 became the Najia-Doodle-One. Check out the Before and After pictures. Obviously the fluorescent red version is the 'After'.

What This Guide Is
This 'ible is about showing you how it is possible to resurrect a printer that you have no upgrade path for -- either the board is dead and there are no replacements, or your vendor is gone/defunct/not answering your calls, etc.

What This Guide Is Not
This 'ible in no way means to show preference for any particular vendor, printer design, mechanical choice or anything else. I am just putting a dead machine back into operation because I thought I could, rather than junk it like everything else we seem to do in this consumerist world.

First a Teaser Video
When all was said and done, I got the thing working. Here's a video of it running its initial test.

Most of the features are related to the board itself, but I've added a few extras:

  • The carriages in X and Y are unchanged. How is this a feature? Well, I didn't have to take that apart. Maybe someday, if I can manage it, I will replace the remaining plastic parts with metal.
  • The Hot End mount is able to accept another module, currently slated to be a laser
  • Metal heatbed. The SD3 came with a wooden table mounting a metal bed, but you could buy a metal table. Mine is all metal and uses a silicone heater
  • Mounted power supply. The SD3 originally came with a power brick. Ugh.
  • Separately powered LED strip. The SD3 LED strip sucks down so much power the Smoothie Remix hates it. So I ran an old laptop power brick for the LED, works like a charm.
  • Larger print area. Wait, what? Yes, it's larger, because I can specify the limits to use the whole bed. I get a few extra millimeters that way.
  • Better board. 'Nuff said. If you want to get your hands on one of these, contact Ray Kholodovsky on Google+ or wait for the Kickstarter. But really, you can use any board you like, I am just sick of RAMPS and the Arduino IDE. I love editing the configuration in a text editor, loading to the SD card and reboot and voila!
  • E3D V6 hot end. Buh-bye to the original.
  • Bowden Extruder: more speed, no more Jigsaw puzzle, all metal goodness mounted on the outside
  • Totally enclosed case. Keeps the heat in, keeps the smell (mostly) in. Polycarbonate walls mean I can run a laser and not go blind.
  • A door. Well, duh! Keeping all the bad heat in, means a door.
  • Stick PC and LCD Monitor. That's right, the Printer has an integrated Windows 10 PC in addition to its LCD controller. This means I can address the printer as a complete unit. Yes, I can replace parts and all that, I can even store stuff on the SD Card the PC sees on the board, or on the SD Card in the Stick PC. Bottom line, my laptop(s) are free! Free! FREE, I tell you!
  • OctoPrint computer with the switch of a USB cable. Oh yes, that's done, it's sitting right here next to me.
  • Did I already say more print speed? Well, with the extruder no longer on the carriage I can go faster. Doesn't mean more accurate, just faster.
This list of parts is hardly all inclusive. If you are converting an SD3 too, then yes, this should cover it. If not, well, this is article should at least give you some pointers. So let's get started.

Step 1: Gather Your Parts

Parts is Parts

We can break this up into:


As mentioned before I am using a Smoothie Remix from Cohesion3D. Feel free to use whatever you want. My advice, go for more features in a board. The Remix is a 32-bit board, can handle 3 extruders (or spindles or lasers), is configured using a single text file on an SD card and is easier to wire than, well, anything else.

Because the SD3 already has motors, ordinary tap switches for end stops and in my case a very commonly available heat bed and hot end, you are limited mostly by your imagination.


I measured the walls of my SD3 and found that 12" x 12" pieces of polycarbonate work fine. If you have no desire to ever run a laser, feel free to use acrylic instead. I went with 3mm material, mostly because I had it lying around. Use what you want or need. I got most of what I needed from around the web, starting with US Plastics. To cover the top, I used a piece that is 6mm thick, and for the door, a 3mm thick piece like the walls.

Printed Parts:

Here are the printed parts I came up with, again, your needs may differ (you can see some of them in the photos in this step):

  • Some sort of clip to hold your wall material. I modified something I found on Just as a note, I mounted the walls so the clips are on top, bottom and rear. This allows me to slide the walls out should I need to.
  • If you are replacing the hot end mount, something like that. I found one on Thingiverse too. If there is nothing else, you might have to adapt your own.
  • LCD Controller Enclosure (you could buy one, but I'm a nerd, what can I say).
  • LCD Monitor enclosure (of course, and you're surprised?). In my case, the LCD monitor is made up of a dead laptop screen that I added a driver board for (from eBay) and then just designed and printed a box to put it all in

Extruder/Hot End:

You can print one, have one printed, buy one, whatever. I got a cheap all metal one off Amazon. I didn't recycle the SD3's extruder motor because I couldn't get the gear off the shaft. I set that aside for later and just went with the bowden extruder I had in my parts bin. Seriously, you can get these just about anywhere on eBay or Amazon. I chose one that I could mount to the the side of the printer. I am using an E3D V6 hot end. Keep in mind that if you need to print a hot end mount, you need to make sure your hot end can fit or be mounted on it (of course).


You need wire. Remember to use suitable wire for what you are doing. Your motors already have wires, and I trust so does your hot end, bed, thermistors. That just leaves wire for the power supply and any connections.

Power Supply(ies):

I am using a 12V 30A supply left over from another project. Typically, you are going to use 1A per motor (so 4x1A = 4A), your heatbed will likely want 20A, your hot end will want 1A...whoops, here you are at 25A. That doesn't include the board, the LCD controller and whatever else you hook up. Consider running accessories off other supplies or use a 24V supply with more current. The Smoothie allows me to connect separate power for the bed and hot ends and let the board run them. Didn't need to here, but you might.

Step 2: Wall Installation

The picture in this step is fairly self-explanatory, add the walls and mount using clips or whatever you want to use to hold the down. You can also dispense with clips and drill holes to secure the walls directly. I didn't want to poke holes in my material, because I speculated that I would likely mess the material up. Caveat Emptor.

Install your walls so that they overlap any gaps. You are trying to seal in fumes and heat here. Also install your walls on the outside. Makes access way easier. I used a hobby mini-table saw to cut off areas that interfered with rivets or other frame protrusions.

Step 3: Install the Hot End Mount and Hot End

Have a look at how the Hot End is installed here. The mount goes on the carriage but I haven't installed the top lid yet so that I can have access from above. Then the hot end's wiring is routed through the rear of the printer along with the PTFE tube. You can see, in this picture, the metal table for the Z-axis.

Step 4: Connect the Board

The next thing to do is hook up the board and confirm it has a heartbeat. In my case, I didn't really plan on running it yet, I just wanted to confirm it got power and was able too accept commands. I loaded a basic configuration, hooked up a bench power supply (watch your polarities) and turned the whole thing on. You should not connect all the motors and things like I have here. This picture is what it looks like back there when the wiring is a rat's nest. You should, after you have sorted out your wiring, clean up the wire with some wrap or zip-ties.

If you have an LCD controller, configure your board to accept/recognize it, connect it to the board and power back up. Confirm the controller sees the board and you can move around the menus.

Unlike a RAMPS, the Smoothie lets me connect motors and run them without the end stops in place. If you can, run motors without end stops, this is a good time to do so. I connected each motor and confirmed that I can move each with the controller. With a plastic screw driver, I turned the potentiometer on the stepper driver up if the motor seemed to hesitate until it moved normally. Be careful that you don't slam a carriage into a detente. Take your time and take small steps. Don't do anything with heating elements just yet. You need to address the motion first.

You can now add end stops and determine if you are using Xmin or Xmax, Ymin or Ymax, Zmin or Zmax. Determine where your 'home' location is. In my case, I had Xmax, Ymax, Zmin because my home was all the way to the right and back of the printer (as viewed from the door). Adjust your position and origin and adjust your bed size. You will probably have to also adjust your end stop detentes to stop the head where it should stop. In my case, I had to replace the 25mm M3 adjustment screw with a 50mm screw because my hot end's nozzle is higher in the frame than the original.

Step 5: Test the Hot End and HeatBed

This is a good time to connect the thermistors. As always, do this with the power off and the USB cable (if any) unplugged from any PC. In your configuration define the hot end thermistor. Reboot and look at your LCD controller. It should show bed and hot end at the ambient temperature. In the pictures here, you will see my LCD showing this (I ended up replacing the enclosure). You must show a temperature before you hit the heat.

Shut the system down and then connect the hot end and heat bed. In my case, the Cohesion3D has mosfets for the hot ends (3 of them), which makes it easy to test -- I can power everything separately. When you apply power to get the bed and the hot end heated up, you might smell first and then see a wisp of smoke from the hot end. You might also see your temperature shoot up past your setting (as much as 10 degrees C). The smoke is just the preservative and machine oils burning off. The temperature should settle down and go back to where you set it. Your hot end will heat up much faster than your bed. Be ready to pull the plug in case you have a runaway heater on your hands. Set your board to control the overshoot as appropriate. Be sure to set your temperature to a point that an overshoot will not fry the hot end. I used 195 degrees C, giving myself plenty of leeway in case I needed to kill the power abruptly.

Step 6: Finalize the Build

Once I was happy the printer worked the way I wanted to, I set the minimum Z height (I didn't want the table to smash into the nozzle) and did a full test with the LCD. That includes a home test, and a full heat test (including letting the printer stay hot for 30 minutes).

From here, it's mostly cosmetics. I added the lid, set my settings in Repetier to match the print bed and home positions in my configuration, installed the door (I used stock Home Depot hinges) and placed the LCD comfortably on top of the printer. The LED strip is connected to power via an old laptop power supply 'brick' because it draws too much power to get everything running as well as the strip.

This is also a good time to clean up your wiring bundle. In my case, I used a connection block to avoid the use of long wires and to let any moving wires experience the least amount of strain. I bolted the connection block to the back of the printer along with the power supply. You can see this in this picture as well as where I parked the extruder. You can also see the LCD Monitor for the Stick PC.

Step 7: Closing Considerations

I built an LCD monitor that I can plug any HDMI device into using a dead laptop monitor and an eBay sourced driver board. I also 3D printed most of an enclosure (the rest is acrylic). The stick PC does great, runs Windows 10, has 2GB of RAM and a 32GB drive. An SD card is drive D: and gets me more space. Wi Fi is already on board, so it makes sense to make that part of the printer. I use an RF cordless keyboard for input with a mouse on board. The overall lashup works very well.

So you see a dead printer need not stay dead. If it is mechanically sound, you can upgrade it, and bring it back to service. I highly recommend a Smoothie based board based on my experience with RAMPS and Duet. I also highly recommend the Cohesion3D Smoothie Remix, especially since when upgrading you are essentially swapping board connections and changing a text file for configuration. If anyone has any questions, feel free to let me know.

Be the First to Share


    • Pocket-Sized Speed Challenge

      Pocket-Sized Speed Challenge
    • Super-Size Speed Challenge

      Super-Size Speed Challenge
    • Metalworking Contest

      Metalworking Contest

    6 Discussions


    1 year ago

    Hi, well written article.

    I have an UP 2 Plus and although working, it's not working the way I want and is closed source firmware/software. They have a A$399 CPU upgrade to upgrade to 32 bit and wifi but for that price they can get nicked as would rather get a complete 3rd party 32 bit board with wifi and learn something while doing it.

    I basically taught my self electronics/Arduino a few years ago and have done a lot on Arduino programming (including full frame rate 128*128 16 bit video with sound on an Arduino Due - SSD1351 driver with DMA)

    *** The other option I have is getting a board and minimal devices so I can just test compiling the code and controlling the wifi, steppers, limit switches and other sensors.


    Reply 1 year ago

    I have been where you are now. I haven't checked but does the UP 2 have a proprietary hot end or extruder? If they are straightforward units, then you should be able to update the electronics. I am a big fan of Smoothie based boards and in my example here I am using one from Cohesion 3D and their boards are 32 bit. I can totally vouch for C3D, they have excellent boards and their support is exemplary. If you prefer, an MKS board with a touchscreen is also an option and you might be able. For an $800 printer, replacing the board would effectively be halfway to buying a replacement machine. On the other hand, if you replace the electronics you effectively are rolling your own printer nonetheless. Between the new board and all your time (and anything else you might damage in your experiment), you can expect to spend about $300. Honestly, for $300 or so, you can buy a much bigger JGAurora A5 with a 300x300 bed (I have 3 of them), touchscreen and filament runout sensing. I have serviced mine as well with minimal parts and there's a YouTube video I noticed that adds OctoPrint and Wifi. If you want WiFi control you can do all that with an OctoPrint setup on Raspberry Pi versus having to buy a whole printer. If you want a machine that has built in WiFi, the Robo3D R2 has that (guess what, it is using a RasPi 3 and Octo onboard). My project was originally meant to test some boards for C3D, and I didn't want to do it on a working printer so I did do it on my mostly dead Solidoodle. So the short answer is you have these options:

    1) Take their upgrade, stay on the 'safe' side. Yes, I know, I know. It's not my thing either
    2) Go with a Smoothie board with onboard Wifi (there are a few out there, C3D's is/was working on it, the new one from Smoothie I believe does), but caveat emptor, you are totally on your own here
    3) Sell this printer and buy another, using the proceeds to finance the replacement
    4) Give up on the 32 bit for now and run the printer off an OctoPrint setup so that you can run it over Wifi
    5) #2 and #5 above, because WiFi doesn't mean it does it well natively

    Feel free to ping me back with ideas you want to toss over the fence. I have torn up, ruined, built , modified, upgraded and otherwise worked on a lot of printers over the past few years.


    Reply 1 year ago

    Hi, thanks for your reply. I looked at the Cetus 3D Mk3 that seems quite good and at ~ A$500 is madly close to just the CPU upgrade of the UP I have. Main bummer of the Cetus is once again on non-open source hardware/software.

    Thinking about making my own "ramps like" shield for my Arduino Due as it's more than capable. Have made a Due shield before so no problem populating the board differently and realistically the complexity is not that bad. Worse case scenario I'll have a Due compatible driver board to experiment with.


    Reply 1 year ago

    You are far braver than I. I do not have the chops to create my own board, so hats off to you that you can. I usually draw the line at open source software vs hardware for my machines.


    Question 2 years ago on Step 1

    I'm trying to convert an Afinia h800 to a RepRap/Mega/Marlin open source printer, but I'm looking for some support. This is the closest I've found to a starting place, but I'm afraid it's still over my head. Any suggestions where to look?


    Answer 2 years ago

    I'm not sure where to start. I am assuming you already know all the settings for the original printer in terms of dimensions and all that. If you are doing the conversion, does the Afinia have a proprietary extruder or anything like that (like the XYZ DaVinci line)? If so you might have to retrofit your extruder too. So on the assumption the only thing proprietary is the board and maybe the filament, I would say you have to load Marlin with the settings for that printer (dimensions, number of extruders, type of thermister, and so on). Once flashed, park the board in or on the printer and go to town. In my case, I have started to move away from RAMPS and gone down the Smoothie route just because there is no flashing of configuration settings (those are all on the SD card as a text file) and the firmware updates are self loading. I had a DaVinci Duo that I wanted to convert too when its motherboard died a horrible death, but the expense and time spent outweighed just buying a cheaper Chinese knockoff printer and dropping a decent board in it. But if you are hell bent on doing this start by identifying all the firmware settings on the board and what is or isn't propietary (I do not know anything about this printer as of now).