Neil Rosenberg published a new build:

Read more about this build...

 

Hi there and thanks for the message.

I want to be careful not to criticize CoreXY, as there are some VERY fine printers of this type. Take for example the Fusion series (i.e. F306) -- their printers use string drive in a CoreXY configuration and put out truly amazing print quality very quickly. They also cost over $4k each.

My comment is based on experiences with belt-driven printers. I have seen symptoms of unwanted flexing of the arms/carriages caused by (I think) interaction between the two axes. It was most pronounced in the H-Bot (a known issue), but I also saw it on CoreXY. In short, many of the movements exert unbalanced forces on the various bars and bearings and thus create the opportunity for positional errors. The Cross-Beam (and Prusa-style for that matter) does not have this issue. In fact the Dash-X Cross-beam goes one step further by having active synchronous drive on both ends of each beam.

I'm talking about really fine distinctions. I was able to get good parts from CoreXY, but they didn't quite rise to the quality my newly acquired Creality Ender 3 and CR10s Pro. BTW, the main drawback of these two fairly inexpensive Creality printers is their speed, or lack thereof.

Early results with this Cross-Arm system in Dash-X suggest that I'm getting quite close to Creality quality, while still keeping the speed up where I like it.

Of course your mileage may vary.

I'll report more as I dial it in and see how good it can get.

Neil

 

My build volume is 300mm cubed. Might be a few mm short of that in the z axis.

Love to see what you're building, keep in touch!

 

Very interesting!

I designed/built a 24" cube printer a few years back, but when the reality of printing times became apparent I decided it wasn't really suitable for making parts in my lifetime. In that printer I used the Volcano hot end plus a .8mm dia nozzle. Even with that, very large parts took an eternity.

Best of luck with your project.

Neil

 

I have completed and published a new book "Designing 3D Printers: Essential Knowledge". This may be helpful for those wishing to make their own custom printer. The book is available from Amazon: https://www.amazon.com/Designing-3D-Printers-Essential-Knowledge/dp/1082381861

Questions or comments are welcome, please leave them here and I'll respond asap.

Thanks
Neil

 

STLs, IGES and STEP files now uploaded and available. Thanks all for your interest.

 

I have updated the parts list

1. A few item links needed refreshing, they are corrected now.
2. 24v bed heaters are becoming scarce, everyone's using 120v with external relays. New links are provided to both in the list.
3. I neglected to include the aluminum for the bed, it's in the list now.
4. I also neglected to include the print surface. I am linking to my current favorite, the Anycubic Ultrabase.

Neil

 

Not at this time. The design has been pretty stable for a while.

 

Hi there and thanks for the message/request. Looks like you're having fun with 3D printers!

Unfortunately, I don't have the time just now to put further effort into the Dash-X printer, my teaching schedule is very heavy.

If you have specific questions or concerns, please post them here -- I'll do my best to answer.

Sorry
Neil

 

I have created a more generic version of the electronics enclosure, now supports MKS Gen L, MKS Gen 1.4 and SKR 1.3 controllers. Please see:
Case for MKS Gen L, MKS Gen 1.4 and SKR 1.3 Controllers by Narosenberg

Neil

 

Hi there, and thanks for the thoughtful comment.

You are correct, in theory it would be fine to move the motors to the other ends of the respective drives, and yes this could make belt tensioning a bit simpler. In hindsight that might have made more sense, I suppose it's just how the design evolved. I do remember fussing quite a bit with location of components versus carriage movement and bed offsets, there's not a lot of symmetry. Among my other design constraints: I wanted to have the motors near each other and close to the controller (for wiring purposes), away from the operator, and out of the way of the filament spool and Bowden extruder.

Looking at the current layout, you would probably have the least issues by moving the X-Axis motor to the rear-left and the Y-Axis motor to the front-right. The motor brackets may need some redesign. You may also have to relocate/redesign the limit switch brackets. If you have access to a cad package (i.e. Solidworks) I suggest you try swapping things in a test assembly to make sure everything clears.

In practice, belt tension is a set-and-forget affair. Unless you were planning on making this in production, the current design really doesn't add significantly to the setup time, but perhaps I'm rationalizing it

Good luck with the build, and do post photos once you've got it together please!

 

Congrats on getting so far along, you're well on your way! I can't tell from your messages whether you have a copy of my book, if not you may find answers to a number of your questions in there plus links to other sites that describe things much better than I can. Also, a great source of info on Youtube is Michael Lawson, Teaching Tech.

Before you know it you'll have the basis of a great working printer and then comes the "joy" of dialing it in: a years-long process of learning, printing and head-scratching. It's more fun than it sounds. If you're like many 3d printing enthusiasts, you'll start to appreciate much better print quality and wondering how you ever put up with such shabby looking parts in the past.

Neil

 

All of my designs for the past few years have used the Openbuilds wheels wherever possible. They have MANY advantages over other linear bearing styles, particularly in light-duty applications like 3d printers. The biggest one is the ability to dial-in a perfect running fit (with the eccentrics), something no other type of bearing allows. That's very important to me.

 

The 32 bit boards (SKR 1.3, 1.4...) do require the 2.x version of Marlin. It's not that different in terms of what it does (yet), but it is a learning curve getting it going. It's no longer on the Arduino IDE, the method of editing, building and transferring the firmware is quite different. And most troubling of all, some important settings and constants in the source code have changed names. Since many of the names were (and still are) fairly confusing, it took me a while to locate the various changes I wanted to make.

At the end of it all, the benefits of going to 32 bit were minimal to non-existent for Dash and printers like it, where the processing overhead is not huge. If I was building a Delta, 32 bit is the way to go. For simple cartesian, the 8-bit boards are just fine. I'm leaving Dash at 8 bit for now.

If you're considering changing for the learning, or adding some fancy new features (wifi, strange motor drivers, scara, other?) I say "go for it". If you're looking for performance improvement however, you may be disappointed.

One last thing, the BLTouch has been updated (3.1) -- you'll need to be careful in any case to use the VERY latest version of Marlin or you may have trouble with it. You know engineers, they love to change things.

Search in the YouTube archives, you'll find lots of great info on all of the above.

 

8mm leadscrews should be ok, as you say the loads (and speeds) are not huge.

As re the controller, if you can afford it, you really owe it to yourself to get the latest version of whatever board. I would not hesitate to use the MKS Gen 1.4 or Gen L again. The price is pretty negligible, and you'll be assured of Marlin compatibility plus all of the necessary I/O.

Neil

 

Really nice looking machine! I particularly like the synchronous top-drive Z-Axis, very innovative.

One comment on that however, I usually like more wrap on drive or driven pulleys, at least 90 degrees. From the photo it looks like your Z-Axis motor only has a few teeth touching. Perhaps I'm missing something or the photo is misleading?

As re the SKR Mini, looks like a really nice compact and inexpensive solution with lots of I/O. I may pick one up and try it.

I agree with you on sensorless homing. I prefer actual switches, they're just not that hard to implement.

Thanks for posting, have you thought about putting this build up here on Openbuilds?
Neil