Ground up Restoration - TAZ3 to TAZ5

I know this has been covered before, that being, people converting their old TAZ 3 to a TAZ 5, but thought I’d share my process with anyone interested in watching it.

A while (6 weeks or so) ago, I managed to get my hands on a well used TAZ 3 printer for $635.00 USD. Even living in Canada (with the nasty exchange rate these days), that’s a bargain in my eyes.

Once I received the printer, and got it re-assembled (got pretty battered from shipping), I tested it to ensure that it was working. Compared to my new TAZ 5, the prints that were coming off were sloppy at best. The linear bearings were severely worn, and the mechanics of this printer certainly are not up to par with what I got on my TAZ 5. So I started bouncing around the idea of upgrading this printer. With some help (parts lists, comments from others) and lots of reading on this forum, I decided to take a shot at upgrading this printer.

First I started off by printing all of the necessary parts on my TAZ 5. That went pretty smoothly.

While I was printing off the required plastic parts, I started sourcing the lead screws, lead screw nuts and all the heatset inserts, screws, nuts, washers etc etc that I’ll need for the project.

Just recently everything came together, so I’ve started on the rehab.

Looking over the TAZ 3 with a critical eye, I noticed that the bearings of course were shot and I mean really shot! LOTS of play in those bearings, so the decision was made that all 11 of them had to go. Also noticed that the bearing holders on the bottom of the heated print bed were of a different configuration than those of my TAZ 5. I’ve opted to change those out as well.

As anyone with a TAZ 5 knows, X-Axis end plates are aluminum. As those are two pieces that I don’t have access to, I’ve decided to print those as well. Will see how it goes. I was reading a thread on here that discussed the pros/cons of using plastic. It was written that the Lulzbot people said the printer would work just as well with plastic or aluminum X-axis end plates. As such, I’m going with plastic and see how it goes. Can always figure something out later if it doesn’t work well.

So, the plan is as follows, not necessarily in any particular order. I’ve printed all of the parts using black eSun HIPS, 0.3mm layer height, using 75 percent infill on everything except the leadscrew bolt mounting pieces, linear bearing mounts, and the X-axis end plates. Those I printed at 100 percent fill.

  1. Replace all 11 linear bearings
  2. Replace all linear bear holders
  3. Replace Budaschnozzle with new Hexagon hotend toolhead
  4. Replace heated print bed with Borosilicate glass and PEI
  5. Replace TAZ3 leadscrews and all associated plastic parts with TAZ 5 equivalent parts
  6. Replace/Install any/all 608 bearings with new
  7. Update machines firmware with stock TAZ 5 firmware

Future plans (following the hopefully successful rehab) is to build a Flexystruder that I’ll test on this printer. Possibly even build a dual extruder and/or flexydually. Will see. Got to get through this mini-project first.

If there are any questions/comments as I go through this build, please feel free.

If you are going to go with printed X ends, I recommend incorporating the X end Motor Mount and Idler mount. Chiphead designed an excellent pair, and was kind enough to post them here

I was lucky enough to source a set of metal endplates for my 3 to 5 conversion, but I did print the plastic ones for comparison. The plastic ones had a fair amount of side to side play in them. Before I got the metal ones, I was considering making a more rigid unit. Just thickening the center section on the opposite side of where the leadscrew mount attaches would probably do the trick without compromising anything else.

2 other things to add to your list:

  1. change out the Z axis endstop switch and wiring. It moves from the lower threaded rod mount to the upper X end motor mount, so you will need a longer endstop wire. The endstop itself also mounts differently so you will need some M2 cap screws. I think the 10mm long ones work. You will also need an M5 Cap screw thumbscrew cover for the new Z endstop target. ANd a different spring and bolt for that matter.
  2. You will need to run a fan cable out to the X carriage to power the always on hot end cooling fan that is required for the hexagon hot end. You can see where that needs to plug into the board, along with the other assembly options at

The bed mount for the Y axis to the frame are also different, but I did not swap those out on mine. The stock 3 ones still line up fine with the new Taz 5 Z and X hardware bits.

If you can find them inexpensively, you might also want to consider swapping out the stock Taz X axis rails with stronger hardened ones. That will eliminate most of the X carriage droop in the middle. That should only add $30 or so to your rebuild cost.

I wonder if we could convince the Lulzbot peoples to sell us each a Taz 5 style electronics enclosure?

Thanks for the link. They look like good solid pieces. Printing them off right now. Thanks!

Thanks for your comments.

Lucky you that you sourced some nice metal endplates. Going to try those beefed up printed plates from the link above. See how that works. If in the end I’m not happy with it, I’ll have to make a set for myself somehow. Would be nice if they were able to be purchased from Lulzbot.

Yes, I had noticed today that the endstop wiring is rather short! Heh. Will for sure have to lengthen that wire. I already have the M5 capscrew and spring, so that’s all set to go in too.

You’re also right about the fan wiring that needs to go into the control box too. Saw that when I was reviewing the upgrade of the hotend. Good to mention too, thanks.

I like the idea of the harder rails. Would you happen to have a source where those could be found?

Was looking last night on the lulzbot site for the TAZ 5 electronics enclosure. Thought for sure I had seen it for sale, but alas it’s the old one (same as I have). Would be nice eh!

So far I’ve got the print bed completed. Here are some photos of my work. I had intended on replacing the belts on the Y and X axis, but upon close examination they’re still in fine condition. I’ll just put the new belts I bought in the replacement parts pile for now.

Pulled apart

Starting with the heatsets

Love how those things go in

Completed from the bottom

And now from the top

And here is the tail end of what I accomplished late last night.

Pulling it apart

The junk pile begins to grow

Really pleased with how these parts came out

First leadscrew is in. Yeah! The next one looks like it might be a tight fit with all those cables coming out of the bottom of the electronics box.

Might be a little late on this… but I’d recommend hardened rods and metal linear bearings. A lot of folks have mentioned that it reduces backlash and allows for faster print speeds.

That reminds me, the side with the electronics box does take some finesse to get back on there. I ended uop having to route some of the cables differently.

Don’t forget a liberal coating of lithium grease on the threaded rods.

Thanks, yes. I have a tube of that ready to go as soon as I complete the assembly. Didn’t want to put that on too soon only to wipe it all off with my hair or arms as I’m working around the printer. :wink:

Got the second leadscrew installed last night, and you aren’t kidding the wiring is tight going into the electronics box. I did end up getting a bit creative with re-routing the wiring but got it all in so far. Still have to get the additional wire through there for the Hexgon toolhead. That’ll fit though, sure of it.

Certainly not too late. For now, I’ll get it up and running with what I have, then look at continuing with suggested improvements on the printer.


I printed off the first one of these pieces last night and a heat retraction (what I suspect to be anyways) problem reared it’s ugly head. Kinda stumped on this one…

Posted a thread here In hopes for some sort of workable solution.

POSTSCRIPT NOTE FOR THIS ENTRY: The belt routing that you see in the photos below is WRONG. The routing as shown causes the X-axis print head to go with the wrong direction. Reading further along you’ll see that I corrected this problem.

Got a bit of work done this weekend on the restoration. I was planning on using those beefy x-axis plates, but the retraction I’m getting on the edges put me off trying to use them. Ended up using the flat x-axis plates that I had previously printed for this upgrade. Once I got them installed, then seem reasonably rigid. Will see how it prints once I get the rest of the work done.

What I have left to do now, is install the upgraded wiring for the new hexagon hotend, lengthen one of the wires for the z-axis stop (if memory serves that’s the one), install the two limit switches, and hook up the rest of the wiring.

Then I need to make sure the z-axis is perfectly parallel to the heated bed assembly. Then I guess it’ll be time then to power it up and see if it even works anymore. Fingers crossed.

Made one mistake here. I put the heatsets on the wrong side. But, it doesn’t appear that the heads of the bolts are interfering with the belt so luckily I’m ok there.

The bag of TAZ3 printed parts, worn-out bushings and threaded rods from the original printer. Once I get the upgraded printer working, then I’ll pull all the heatsets out and discard the parts.

Tackled the wiring tonight and powered it up.

I’ve gone so far as to test the X, Y and Z axis controls and have run into a really strange problem.

The X axis is operating exactly opposite as to how it should. I connected the wires back exactly as they were originally connected to the steppers. The Y axis performs properly as does the Z axis, however, when you click “Home X”, the head moves in the opposite direction that it should.

It’s as if the wires have been crossed, yet I connected them as they were originally connected . Was VERY shocked to see that happen.

Was wondering if anyone has any thoughts on this.


Looking at your picture of the back of the X carriage, you put the belt on backwards. The belt wraps around the top and crosses underneath to the idler then back to the carriage.

Oi! So obvious yet do you think I could see that last night? Heh. Thank-you very much for pointing this out. Guess that’s what happens when you put in full days then expect to be clear in your work late at night. Glad that it’s going to be a quick and easy fix.

As it turns out I forgot to order some 2.5m screws and heatsets, so I wasn’t able to properly install the X and Z limit switches. Put a couple little pins in the holes and taped the switches on for now as I was impatient to get this thing powered up. Clearly not the permanent solution. Gonna have to see what other parts I need from McMaster and get those parts underway.

Will report again once I get the belt switched around and things leveled.

Some initial problems I had with the printer was with one of the Z-axis steppers. When I first tested that, only one side worked as it started racking the entire assembly. Caught it in time. The problem was that two of the wires to the right most Z axis stepper had pulled out of the pins. I’ll have to figure out how to pull that apart and fix them. For now, I just stuffed the wires back in the holes. Needless to say, I won’t do much printing the way that is till I fix it.

After I corrected the messed up belt routing fiasco, I leveled the Z axis with my digital dial gauge, then proceeded to level the heated bed. Got it within 0.01mm between all four corners.

The printer looks a little ghetto right now with the limit switches being held in place with masking tape and none of the wires routed nicely (and a bad connector), but I HAD to do a test print regardless.

This is a small 4 layer rectangle that I use when changing filament colors. Was small enough to test the printer. I think it came out remarkably well having just put the printer together, loaded the firmware and changed nothing else.

I won’t know for sure till I start doing some serious printing, but so far the parts I printed and installed seem to be working alright. Even the stock looking X-axis plates seem very rigid, and with the two x-axis steel rods firmly in place, I’m able to get the x-axis belt very tight (as tight as it is on my TAZ5).

So, now besides the obvious (getting the hardware I need to properly affix the limit switches, and fixing the bad connector, I think I’m going to turn my attention to the TAZ3 electronics case that’s on this beast. I really don’t like how tight it is with all the wiring, and the fan in there is HORRIBLY noisy. I think I could find a quieter fan if I had a bit more room in the case.

Been thinking today that I should see if there are any 3D printable electronics cases out there that someone else has done. If not, might be time to spark up my 123D Design and make one, similar to the TAZ5 box. Will see what I can find.

I’m very pleased with the upgraded printer so far, seems to be working fine just having loaded the stock TAZ5 firmware and running with it.

Wondering if anyone knows what these connectors are called, and who might sell them. The connector that has the wires pulled out has an MX 12L on it. That’s the only descriptor I can find. Would also like to know the trick of how to pull the pins out of these things. If I could do that, then I could solder the wire in and repair the connector. Perhaps pushing in on the little silver opening on the first picture could release it. Has to be locked in there somehow.

This is the offending connector. The clear and green wires are the ones that came out.

Another view of the offending connector

This is the receiving end on the right z-axis stepper

The connectors are most likely Molex, available from DigiKey. Support can tell you or send you a link to the TAZ 3 BOM.

You should be able to release the pins by pressing on the small metal tab that is visible in the small square window with the tip of a small screwdriver.


Thanks for the suggestion on checking out the BOM (and for releasing the pins). Found the following for the four pin connectors at Digikey after combing through the bill of materials for the TAZ.

4-pin.molex.CONNECTOR.female - 0050579404 Molex Inc | WM2902-ND | DigiKey

4-pin.molex.CONNECTOR.male - 0701070003 Molex Inc | WM2535-ND | DigiKey

4-pin.molex.PIN.female - 0016020086 Molex Inc | WM2510TR-ND | DigiKey

4-pin.molex.PIN.male - 0016020107 Molex Inc | WM2517TR-ND | DigiKey

For now, I pulled the pins out, soldered the wires on and did a bit of heat shrink. It’s not pretty, and you can still see the conductor, but it works and nothing is shorting or pulling away from where it should be. Should do me until I can get my Digikey order.

Also received my latest shipment from McMaster-Carr, and got my M2 bolts and heatsets. Got that installed and the masking tape is a thing of the past. Limit switches work great.

Outstanding issues are still that I REALLY don’t like the wiring as it comes out of the control box. The wire appears too short now for the heated bed such that I can’t bundle tie it to where it’s supposed to go.

Been looking through thingiverse a bit and haven’t noticed anything that I much like the look of to resolve these concerns. Unless someone reading this has any cool suggestions, I’ll keep hunting. In the end I may have to draw up something in 123D Design that will satisfy my requirements for a better box (that is unless the lulzbot people feel inclined to list the TAZ5 control box for sale on their site).

Well, it’s been a while since I’ve posted, and thought it was time for an update.

I had not done much printing on this printer, as a number of the wires needed to be re-routed for the printer to function correctly. The print head, heated bed and Z axis wiring was simply too short for it to be run out the bottom of the TAZ3 electronics box. If you look at the TAZ5, you’ll notice that they run out the back of the electronics box. Not having access to a TAZ5 box, and not finding any viable solutions out on the web for this, I went about the “dreaming” phase of this project, trying to visualize something that would work.

What I ended up coming up with is a 17mm spacer that allows routing of said cables out the back, while keeping the electronics box cover. It’s (I think) a very elegant and simple solution that has ended up working REALLY well. Here are photos of the deployment.

The spacer was printed on this printer (before the wiring was corrected). It was printed with black PLA at 0.3mm resolution.

Subsequent to this, I have been running the printer hard! Running LOTS of short and long duration prints through it, and I have to say I am SUPER STOKED about the quality of the prints coming off this printer. In fact, there is no difference in quality between this upgraded TAZ3 and my TAZ5. NONE! Another thing to note is that, those 3D printed plates (that the lead screw bearings and Z axis assembly attach to) that are made out of laser cut aluminum on the TAZ5 function perfectly on the upgraded TAZ3 simply printed at 100 percent fill. I see absolutely NO NEED to go to the effort of getting those made out of aluminum. This upgraded printer that I’ve done is absolutely rock solid.

So all in all, this has been a fantastic success. With the printer completed, and “in production” so to speak, I’m now running whatever prints I need to do with it, and they’re coming out perfect every time.

Here are some examples of the quality of prints coming off this printer, and I think you’ll agree that they’re pretty impressive.

Here is a Raspberry Pi Case I printed for a Minecraft server I put together for my son

Here is a Raspberry Pi TAZ5 replacement end case plate that I printed

This is the bottom half of the UZEBOX I’m working on.

This is the two half shells for the Adafruit Pigrrl project

I hope you’ve enjoyed following along my progress through the re-construction of my TAZ3 into a TAZ35 :smiley: