Reliable core XY printer

Hi everyone,

After having bought and used plenty of Cartesian printers, i came to a conclusion that 3D printing is a long way from a consumer level. Not just in terms of being able to get the machine to do anything but more so on the mechanical aspect of it.

I find that a lot of design choices in printers are not suitable for low maintenance operation. There's always something to adjust, oil up or look after.
Not that i am that bothered with all that myself, but i do wonder what kind of machine could provide trouble free operation for at least a month without maintenance?

With core XY machines slowly taking over, i want to try one, but instead of picking the fastest and most feature rich I'd like a machine that is dependable.
Right now there are so many variations and designs, all seem valid. What feature would you guys mark as essential for a reliable core XY machine?

To me V groove bearings seem suspect of maintenance. Also belt tension seems a source of problems, anyone have experience with good designs?

My printer has been very reliable. I designed it so I didn't have to use any autoleveling. I level the bed once and don't have to touch it again unless I modify the Z axis. I never have problems with the electronics or the belts. I used an absolute minimum of plastic parts in the printer, and used quality Japanese made linear guides purchased used via ebay for about the same price as the crappy Hi-Win knockoffs. You can see details of it by clicking the link in my sig, below.

Corexy belt tuning

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

I usually end up having to repair something on my printer about once every 2 years. In particular, the adhesive that holds the heater and PEI to the bed plate is good for about 2 years, then lets go. I solved the heater problem by using high temperature silicone to attach it to the bed plate, about 3 years ago, IRIC. I noticed that the PEI is starting to lift at one corner again- it's been a couple years since I last installed it, so time for a refresh.

The rest of the machine doesn't require much other maintenance. I recently put some lube on the guide rails for the first time in a few years. I used a Duet controller board that has drivers soldered to the board, so they don't overheat, and current is set by gcode commands in the config file- there are no tiny pots to break. I used a reliable MeanWell power supply to power everything. I generally prefer to keep things simple because simple often translates to reliable. For example, I have a 5" 220V fan connected to 117V when the machine powers up, blowing across the controller board whenever the machine is powered on. The fan turns slowly and quietly and keeps the board from heating up. Some people want to put in thermostatically controlled fans, but that's an added complication for minimal benefit- the fan won't turn on until the board heats up. Why not just keep it from heating up in the first place?

Many people think they want auto tramming because they remember the bad old days of having to manually retram the bed every time they tried to print. It never occurred to them to ask why the bed went out of tram, and to fix that problem. They just threw more complication in to handle it with a combo of additional hardware and software. New 3D printer people see youtube videos of printers tramming the beds automatically and think it's cool, so they want that, often without knowing the purpose of tramming the bed. Now all the printer makers include it. Auto tramming is a chicken and egg problem. People think they want auto tramming, so they use multiple motors to drive the Z axis, but if you use multiple motors, you need auto tramming because the motors get out of sync and that tilts the bed. In the bad old days, people used multiple motors because they thought it was cheaper/easier than connecting the screws to a single motor with a loop belt. Then they struggled with bed tramming because the bed (or X axis) would tilt when the motors were powered up. It took a while to realize that the motors jumping was causing that problem. What people really want, whether they realize it or not, is a machine that gets the first print layer to stick reliably without having to retram the bed manually, over and over. If you never let the screws or belts that lift the bed get out of sync, the bed won't tilt and you won't need to keep retramming it. That is easily accomplished by driving all Z axis screws or belts with one motor.

In UMMD the bed is lifted with two belts driven by one motor. The belts can't get out of sync, so the bed can't tilt. If I take the Z axis apart for some reason (it's been a couple years since I did that) I retram the bed manually which takes 1-2 minutes. I think it's been two years since I last retrammed the bed. I don't have to touch it again, even if I stuff the machine into my car, drive it across town and take it back out. The mount is stable and designed using basic rules of geometry -3 points define a plane (not 4!) - and thoughtful placement of those points makes it very quick and easy to retram the bed. The mount is a kinematic type that allows the bed to expand when heated without causing anything to bend. It is very stable because the bed plate is supported on the tramming screw heads, with springs used to hold the bed plate down, not push it up.

Edited 1 time(s). Last edit at 11/03/2022 11:25AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Jooch, your post is unclear about whether you wish to buy or build a CoreXY printer. If you wish to buy, I don't know of any CoreXY for sale that are suitable for low maintenance operation.

Quote

What feature would you guys mark as essential for a reliable core XY machine?

To start with, you need the features are required for any reliable 3D printer: decent hotend, decent electronics and power supply etc, so we'll take all that as given. So, I'll outline the features that I think are specific to a reliable core XY machine.

1. Generally, favour a simple design. Generally simple means reliable, "the best part is no part" etc. Many CoreXY designs seem overly complicated.
2. Rigid frame built with 2020 or larger extrusion. Blind joints - they self-align and are more rigid than any kind of corner bracket or corner plate. All parts within the frame so that it can be easily enclosed. Minimise use of plastic parts, and no structural plastic parts.
3. Use of linear rods and/or linear rails for motion system. V-slot with wheels just don't cut it for long-term reliability.
4. Aluminium tooling plate with 3 point fixing for print bed. 3-point fixing so that tramming is easy, and also so the bed does not warp when trammed (as it will with a 4-point fixing). Also the fixing points should allow motion in the horizontal (but not vertical) direction, so that the bed does not warp as it expands.
5. Manual bed tramming. Fully concur with the digital dentist here. I want a bed that I tram once, and it stays trammed. My first printer was a delta printer, and once I had set that up, I never needed to adjust the endstops for the bed.
6. Following on from (5), for print beds up to about 250x250mm, I prefer a cantilevered bed. (I believe the digital dentist will disagree with me on this point). Properly done, a cantilevered bed works well for a 3D printer. Note that the Slant 3D printer (see below) and the E3D Toolchanger and Motion System have a cantilevered bed. The MC300 (see below) uses 12mm linear rods and SCS12LUU bearing blocks: I've placed a 5kg weight in the centre of the bed and have measured a deflection of 0.2 degrees. Along with this I'd look for a z-motor with an integrated lead screw. Using an integrated lead screw is much better than using a shaft coupling - no problems with alignment or z-wobble.
7. Printer available in a variety of sizes. There are some CoreXY 3D printers that are only available in large sizes. If you only need to print in smaller sizes, then a large printer is overkill. What's more a large printer (all other things equal) is more expensive, takes longer to heat up, uses more power, vibrates more, is noisier, and is more susceptible to coming out of alignment.
8. Support different types of hotend. There is currently a lot of innovation happening in hotends, I want the ability to change to a different type of hotend in future. I don't want to be stuck with one type of hotend and I certainly don't want to be locked into a proprietary hotend.
9. Parts accessible for easy maintenance. I've seen a lot of designs where you have to disassemble large parts of the printer to replace a part.

I've recently discovered a series of youtube videos from Slant 3D. They run "the largest print farm in the western hemisphere" with hundreds of machines, with a plan to expand to three thousand machines. They built their own 3D printer because nothing was available that suited their reliability requirements. Slant 3D talk through the design of thier 3D printer here. Anyone who runs a large print farm is worth listening to when they talk about reliability.

And finally, you might like to look at my own design of CoreXY 3D printer, the MaybeCube. It has most of the features above. A description of the MaybeCube is here. Build instructions and BOM for the MC350 variant are here.