Klipper + Input Shaping on TAZ

Has anyone put Klipper on a Rev 1.3L board (taz5) and done input shaping?

Are there STL files for adjustable belt tensioners for the taz printers? Should I make one? What about adxl345 mounts on the bed and hotend?

I’m getting crazy numbers with adxl345 trying to run input shaping and it seems almost like a belt is loose. I have very little Klipper experience on bed slingers. We have Klipper running on CoreXY, CoreXZ and CroXY (crossed, Ultimaker style) printers.

Am I missing something with the stepping or hold/run currents in these 8-bit boards? They are all going to be converted over to IDEX CoreXZ eventually so should I stop wasting my time and just charge forward with converting them?

Part of me wanted to make sure they were printing decent before I completely disassemble them and rebuild from the ground up.

Any input shaping advice for the taz 5 would be greatly appreciated. Including telling me to give up and walk away from the Lulzbot platform for a printer. I need something that can print IDEX and 300x300x200z.

Lulzbot: Do you have any metrics or numbers for igus bearings/bushings being out of tolerance? Should I look to igus to get those answers? I’m wondering if part of my problem is that some of the igus parts have been printed on a bunch and therefore need replacing. But also with how robust these taz 5 printers are, I question if the igus bushings are bad. (possible that prior owner --a school, could have replaced with who knows what parts).


The entire printhead of the taz is loose.

The whole sandwich goes:
Round rod, so the only thing keeping the head from spinning around is the second (round) rod.
Around that round rod is a (high-quality) Igus linear bearing. As it wears, it is less able to keep the head from twisting as there’s space for movement before it’s controlled by the round rod.
Those bearings are held by the sleeves made of mildly-flexible ABS plastic. Tubes are screwed into a carriage plate also made of ABS. That plate has another ABS plate with the bottom lightly wedged into it with a single screw at the top holding it in place. Attached to this plate is a metal plate that holds the hot end. On this plate is another piece of plastic, and attached to that piece of plastic is a sledgehammer of a nema 17 motor.

This is the bad news. The good news is that almost all the movement from this sandwich is in the Y-axis, which the toolhead doesn’t move in, but there is still significantly more flex in this setup than bolting an extruder to an adapter plate that is screwed right into a linear rail.

More good news? These were never meant to go fast, and at the speeds they are capable of with the heavy bed and head, input shaping really isn’t something necessary unless you’re looking purely for looks.

Now, if you equip them with a lightweight toolhead and are printing parts that won’t sway when moved quickly on the Y axis? Input shaping could be worth pursing.

If you want clout for being the first person to print a sub 10 minute benchy on a Taz 5, I fully support you. If you want a faster Taz 5, don’t waste your money. You will get print quality that will make you cry.

That’s absolutely not my goal. Input Shaping and Pressure Advance isn’t about printing fast.

I want reliable, quality parts (that finish printing before I die).

I get it. We’ve had these printers for a few weeks now and I’m just confused. Lulzbot seems like such a purist community --not open to mods in any way. But I guess if you buy a Rolls Royce, why would you mod it…

The single extruder (v2c?) TAZ 5 I have printing with 0.6mm nozzle is --actually not bad. All I did was blindly put Klipper on it (v 1.3L board) and I had to mess with the X-rod (idler side) grub screws a little because the belt was crooked. Other than that, I did nothing and it’s printing not terribly. And for full disclosure I’m saying “printing not terribly” with the printer having two cracked lower z-motor brackets and very worn igus bushings. I printed new upper Z caps, put Klipper on it and re-printed the extruder parts. Other than that I haven’t touched it. That’s why I’m thinking if it prints this decent as-is, then it’ll probably turn into a halfway decent printer if I put some work into it.

I don’t mind the detached PSU --I like it. It lets me move it out of an enclosure so I only cook the control board.

The glass bed: It just sucks. I need to figure that out.

I’ll put a lightweight toolhead on it and good news: These things are so god awful heavy that Lulzbot used giant motors. That means I can up the stepper motor run-current once I lighten up a few things.

I don’t get why everyone hates on linear rods so much. My experience with linear rails is that you either pay out the a$$ for legit rails, or you regret buying Chinese rails with crunchy bearings that you’ll be making a post about where to buy precision ball-bearings a few weeks down the road. On the flip-side, linear rods with igus will wear but for the most part, they “just work”.

I would have to completely disagree about this being a purist community, it’s open software, open hardware, open assembly instructions… Few of the regulars here run stock machines, as far as I know.

If the glass bed is flat (and you’re not printing at over 100c primarily) I’ve had great results with the textured spring steel sheet from fysetc as a print surface. It’s not quite as good as the octograb, but the price difference is enormous. Others swear by the whambam kits.

The linear rods with igus bearings are extremely reliable and quiet at low speed. The problem on the X axis is weight. Also purchasing low quantities of the Igus bearings is frustrating. I can get a linear rail (yes, from China) faster and for less than trying to get three 12mm igus bearings.

I’ve got fystec on a bunch of smaller printers and whambam on my 300x300 cr10. Do you stick the magnet to the glass? Doesn’t that make the overall bed a bunch heavier? Ideally I’d like to end up with something that is overall lighter than the stock glass, with a spring steel sheet (or the ability for other surfaces, if materials/temps require).

I want flat and a magnetic removable steel sheet. I also don’t want to pay $150-$200 per bed (for 6x printers).

Sorry about the “Purist” comment. I guess when you buy a $3k-$5k printer you don’t exactly go out and mod it like a Creality. I’m just shocked that I can’t find anyone who as CoreXY’d or CoreXZ’d a TAZ. It was hard to search and find someone who had done a Voron toolhead on a Lulzbot. I don’t mind the rods and I don’t mind bed-slinger. I’m not looking for speed, I’m looking for quality prints that aren’t painfully slow.

Igus is local to me. (No clue if they do local pickup) Not sure if it is technically called a “bushing” or a “bearing”. There are no moving parts so I tend to say bushing but who knows… Either way: https://www.amazon.com/Solid-Polymer-LM12UU-Bearing-shaft/dp/B074CCXZ4S That’s a 4-pack for $14. I hope you aren’t buying linear rails for less than $14. Re-packing ball-bearings in linear rails is a painful experience.

I guess I’ll just charge ahead making mounts and parts to adjust belt tension, mount adxl345, put on a smaller and lighter toolhead, and ultimately CoreXZ+IDEX these TAZ 5’s.

Yes. It’s stuck to the glass and does make it heavier, but these shouldn’t be going fast anyway, so I haven’t worried about the extra mass. You could swap the glass for some acrylic, G10 or similar if you want to make it lighter. So long as you have good ABL to compensate it should be fine.

When I was looking for the linear bushings before, it was a 3-4 week wait, with $10 shipping or so.

I would like to see a Taz Pro or Workhorse converted to CANBUS and coreXZ, that’d do something to get rid of that mess of drag chains and wires. I think you could get the belt path set up without major modifications.

I hope I don’t have to do a darn thing. I’ll just steal (and give credit where credit is due) IggyTheWolf’s work on the ‘TazWire’ github.

STL files should probably need some tweaking here & there, and he’s using rails while I plan on using rods, adding a second rod per side, different spacing, probably something similar to a belted-z like the voron0: Something like this, except with rods and stock frame. So some adaptation will be needed. But that’s the idea and the motor with reduction gear (like Lulzbot’s herringbone gears) is perfect for what I’ll need.

When I built my ‘Halfwire’ (DIY, I didn’t use Kingroon printer/frame):

I didn’t initially realize that the halfwire had 2x Y steppers. (look closely, there’s a Y motor in the front view, and in the rear view) I’m using skr3ez board with 2x btt toolheads. So idk, maybe I’ll just do 1x Y stepper first and then possibly move over to individual MCU boards for Y motors.

Again and I want to be clear on this: I’m looking for a super-reliable workhorse, not a speed-benchy printer. So I’m OK if it prints slow, as long as the parts are perfect. And then at some point, I’ll evaluate my cost of electricity and how much it costs to heat a 300x300 bed for 8hr vs. 2.5hrs on a faster printer…

I just need to lighten up the toolhead and hopefully the corexz thing won’t be too painful. All I need is to get one done then the next 5x will be easy.