Designing my first CoreXY printer, would appreciate some opinions.

Currently using a Taz 6 that I'm relatively happy with, but a little bit jealous of you guys' fast and nimble coreXY machines.

Always wanted to build a printer from scratch, initially figured I should just do budget build of an already proven design to first understand each designs limitations and progress along. Considered D-bot, HyperCube Evolatution, and Voron1.5/2 (made a post in Reddit but only got limited responses) but ultimately none of them offered exactly what I want in a printer. So I decided to just go ahead and get a drawing going to figure everything out as I go

This is what I have so far:

• 9mm rail on X carriage, 12mm on Y and Z
• 3030 frame, 2020 X carriage and maybe print bed frame
• 300x300mm 3/8" thick Mic6, 110v silicone heater
• Print head removable with two forward facing M4 thumbscrews, interchange between Bowden and direct drive
• AB stepper motors in the rear to reduce front clutter, idler bearings next to the motors for tensioning
• Front top bar placed towards the middle of the frame for more forward accessibility
• Haven't design the joints yet, they will be printed in ABS for quick revisions, possibly mill out aluminum replacements once I nail everything down

These are my goals:

• 300 x 300 x 300mm
• Rigid frame and scale-able to larger sizes
• Interchangeable print heads - love this feature on my Taz 6
• Z=0 using nozzle contact - makes using different length nozzle/print heads easy.
• Fully enclosed (including wiring loom/Bowen tube) with active heater
• Would prefer to avoid auto bed leveling
• Good visibility to the print area and easy access to print head.

Would be great to achieve but could wait till future revision:

• More elegant 3 point leveling with thermal expansion compliance
• Nozzle wipe pad before homing - limited by travel range and footprint
• Magnetic bed with quick swap spring steel + PEI
• Easier belt tension system

Below are a few question I would like to get some feedback on before I progress any further:

• Please double check my belt path, my understanding is that there are certain section of the belt that needs to be kept parallel with direct of axis, do you see any problems with my tension method?.
• Corner brackets for 3030, between Misumi and hidden, what is your preference?
• Still unsure how to drive Z, for some reason I'm feeling a little iffy about lead screws. Not sure how much I can trust the straightness of them, where to get quality screws? My Taz6 has thick screws and are constrained both top and bottom, I know this design is usually frowned upon, but it works well. How about belt drive? how much reduction would I need to lift/hold my 12" x 12" x 3/8" Mic6 plate?
• GT2 belts, is real Gates worth the extra price over those from Aliexpress? Should I use steel core or larger size for driving Z?
• Regarding silicone bed heater, how much advantage is using a heating that is exact to the bed size vs one that is 3/4 (or 1/2) the size? Aluminum is a good heat conductor but if using an undersized heater does heat distribution becomes an issue on printing?
• Was initially planning to use a Ramps 1.4 (that I'm familiar with) and run 24v + TMC2100, but all research suggest I should just get a DuetWifi which is new to me, anything to look out for? (ex. I need special steppers for 256 microstepping?)
• Building an bed frame, an extrusion box seems the most obvious but are there more elegant bed solutions for 3 point leveling?
• What are you guys opinion on PEI thickness, I've installed a ~1.8mm thick PEI on my Taz6, it's basically indestructible and prints pop off easy, but it did developed very minor "ripples" after a year worth of ABS heat cycles, but I'm unsure if it was caused by it's glass bed's uneven heat distribution.
• Exturder, specifically when in Bowden mode, I've built many true and trusted Wade's and also done a few revisions myself, any other extruders I should consider? Direct drive, only choice will probably be a Titan for size/weight.

That's all I can think of for now. any feedback will be appreciated it! Thanks!

Edited 1 time(s). Last edit at 06/08/2018 10:15PM by chilicoke.

The DuetWifi makes your printer about 50 percent quieter right out the box because it drives the steppers better. It also has a snazzy web interface which means you’ll never have to use sd cards or tether to a pc....

I did not like the titan extruder, but maybe my expectations were too high. It feels a bit flimsy to me. I’m working on replacing that with a bondtech bmg... i’ve modeled my extruder mount and the two bags give me about 20mm more print space in x and y than the titan. they feel more chunky too (printed sls nylon vs injection molded abs).

Quote
Qdeathstar
The DuetWifi makes your printer about 50 percent quieter right out the box because it drives the steppers better. It also has a snazzy web interface which means you’ll never have to use sd cards or tether to a pc....

Exactly what I'm looking for, Ramps + TMC2130 + Raspberry pi for Octoprint is already getting close in terms of cost, but the speed limit on coreXY definitely make Duet sound like a worthy upgrade.

Quote
Qdeathstar
I did not like the titan extruder, but maybe my expectations were too high. It feels a bit flimsy to me. I’m working on replacing that with a bondtech bmg... i’ve modeled my extruder mount and the two bags give me about 20mm more print space in x and y than the titan. they feel more chunky too (printed sls nylon vs injection molded abs).

Thanks for your feedback, I don't have any experience with the Titan nor Bowden set up, as I've been basically using nothing but Wade which is why an interchangeable print head is a pretty important feature to me. Bontech BMG is also something that I'm eyeing however it's high cost also makes the decision harder when there are multiple print heads that I'm considering. As for Titan there are more affordable clones that I can experiment with.

300x300 mm heaters are readily available. I wouldn't mess around trying to use heaters that are smaller than the bed unless you don't mind not being able to print over the entire bed surface. Go with line power, at least 500W, for that size bed. Cast aluminum tooling plate is good, 3 point leveling is good. If you arrange it as a kinematic mount it will be rock solid and will allow the bed plate to expand when heated without causing anything to bend.

Early Z axis testing in my printer used a 160 oz-in NEMA-23 motor to lift the bed using a long belt. The motor was able to lift the bed without any gear reduction. So the gears aren't needed so much for lifting as to prevent the bed from dropping when power is cut. The final design of the belt driven Z axis in my printer uses a 30:1 worm gear drive to prevent bed-drop when the power is cut. With the motor and pulleys I used it achieves 50 full steps per mm. I have tested the load capacity to 4 kg (in addition to the 3.5 kg bed assembly)- it can probably lift more.

I used a piece of 1.4 mm thick PEI on a printer at the makerspace where people tend to abuse the machines. It has survived over 2 years of daily printing and is still working fine. On my home machine I use 0.7 mm PEI and don't expect any problems for years.

In this diagram, belt segments A-H all have to be parallel to their respective guide rails. If they are not parallel, belt tension (and steps/mm) will vary with extruder carriage position and prints will be distorted (especially as print position moves away from the center of the bed) if the error is small, and the mechanism may bind if the error is large.



I've been using Titans for a couple years, with V6 hot-ends and have decided that while they are generally reliable, they leave a lot to be desired. The mounting system is not great- the screws that hold the cover on the extruder also mount the extruder body and motor. That means that if you want to take out the hot-end for any reason, you have to pretty much disassemble the whole thing and then rezero the bed, after you put it back together. There should be separate screws to mount the extruder/motor and the front cover so the unit can be opened for cleaning, etc., without creating a lot more work to get it back together. The hot-end mount is considered sort of a "standard" now, which is a shame because it is big and has no anti-rotation features (a simple flat surface on the side would do the job). Likewise, there's no secure way to mount the heat-break in the hot-end because it has no flat surfaces to grab with a wrench and the heat-break itself is too delicate to apply much torque without destroying it. I've had the heat-break unscrew itself from the heatsink while printing, allowing the heater block and nozzle to flop around which destroys print quality. The fan is held on with a plastic bracket that snaps over the heat sink, and it, too, has no anti-rotation features, so the fan tends to swivel around and sooner or later the plastic bracket comes into contact with the heater block and melts. Finally, the V6 fits a little too loosely into the extruder body, and the extruder body is somewhat flexible.

---

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

Thanks for the detailed response the_digital_dentist, I do have your blog bookmarked and it is a good inspiration for wanting to make my own printer from scratch.

I was hoping to belt lift Z with same Nema17 steppers as the rest of the printers to keep part sourcing simple, 9mm GT2 belt to reduce belt stretch, and belt pulley for drive reduction to avoid any sort of toothed profile gears to show up in Z movement. Curious how are you driving Nema23 motors? Are you using external drivers or driving them directly from printer controller? I'm running Nema23 on my home brewed LMS X2 CNC conversion but am running them with a Geckodrive g540, would love to know how to possibly drive them on 3D printers.

A little disappointing to hear your experience with Titan and I too mirror multiple of your complain on hot ends available today. The downside of wanting to experience with these pre-built direct drives (Titan, Bondtech, etc) with interchangeable print head concept is the huge additional cost for each print head, especially when there are multiple hot ends, different filament sizes, etc. Maybe I'll just stick to Bowden and DIY-able extruders such as Wade's or Mobius for now.

Driving motors is all about current. It doesn't really matter what size the motor is as long as the driver can handle the current it requires, and is adequately heatsinked for that current. I run the NEMA-23 motor in my Z axis with the driver on the SmoothieBoard- actually, I just installed a Duet board, so now it's being driven by that driver. In SoM the X and Z axes are also NEMA-23 motors being driven by the SmoothieBoard controller in that machine. I used to use A4988 drivers on a RAMPS board to run them, but poor heat dissipation on those little modules meant it was always running at about the limit and I had to replace drivers once in a while.

A geared down NEMA-17 motor should be capable of delivering the needed lifting force if you gear it down enough (pulley and belts or gears). Worm gears ensure that the bed won't drop when power is cut, belts and pulleys might or might not depending on the motor and amount of reduction.

---

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

If you want to use nema 23's you can use the bigfoot stepper driver or upgrade to a duet ethernet/wifi with active cooling. the duet can handle up to 3 amps i think(look it up before trusting me) if you blow a fan across the drivers.

Stepperonline also sells nema motor breaks now for 50 ish dollars so you dont need to get a worm gear to account for power loss.

Good to learn that Nema 23s are viable options, will definitely consider it as I continue to evolve my design.

Now that I am trying to find the best Z belt path for my printer to make sure it will clear other stuff, I am starting to debate if it's really worth going with belt lift, or if should I just go with the more conventional lead rod method.
As I am still in the early stage of component placement, I do want to keep the overall size of the printer compact, which means I will have to router longer belts and multiple reduction pulleys will increase complexity and cost.



As I will be lifting a fairly heavy bed assembly, I would prefer slightly thicker/nicer lead screws riding on bearings than the typical Chinese screw integrated Nema 17 motors. Do you guys have any recommendation on where to find well priced quality screws? I've looked on Misumi but their lead screw and nuts are very pricey. Also considering Chinese ball screws as they are well priced.

Quote
warbunniex
Stepperonline also sells nema motor breaks now for 50 ish dollars so you dont need to get a worm gear to account for power loss.

That's very cool but a bit pricy, something like this might come in handy in the future.

Edited 2 time(s). Last edit at 06/12/2018 02:14PM by chilicoke.

I would say keep with belts if you can. Lead screws are overkill for 3d printers and introduce inaccuracies in layer alignment.

My build will be using one of these: [www.robotdigg.com]
Already ordered it. The 500mm one was 150 when shipping was included and it is very nicely built.

Not sure where you can get cheap lead screws. You could look at ingus. But you really dont actually want a thicker screw. The thicker screw will require a stronger z rail to prevent the wobble of misalignment. And there will be misalignment.

That unit looks very nice. I did considered linear guides briefly but it was screw driven instead of belt, it was was a used unit locally but I couldn't get much info on it (wanted to keep overall build buget reasonable) so I moved onto figuring out rest of the printer first.

Looks like that unit is using a geared stepper motor, I'd love to hear your feedback on if it has any backlash or not for Z use.

Making a printer is seem to fascinate for you. You have tried to maintain component placement that is very inspired me. I could give you opinion about printer repair to coinsult to Ricoh printer support.

So Lulzbot Mini 2 was announced and I've been wanting to see how they prevent bed drop since their prototype files first showed that its Z will be belt driven.

It appears they are using the same "brake" warbunniex had suggested.

[www.lulzbot.com]

After considering how much belt length and reduction drive I would need to minimize bed drop with my 3/8 Mic6 plate, I am leaning towards keeping the belts short keep the Z drive assembly simple.

A question regarding the brake unit, Nema17 seems to lose it's holding torque as soon as power is cut. Does the brake react fast enough to completely prevent bed drop instantly?

Update: After being linked to the assembly guide for Mini2 by the owner of Lulzbot himself, it seems like the "brake" used on the Mini is in fact an electronic brake. This is very cool!

Edited 2 time(s). Last edit at 06/20/2018 05:57PM by chilicoke.