Welcome! Log In Create A New Profile

Advanced

Which Kit? Railcore? SecKit? Other?

Posted by Doppler9000 
Which Kit? Railcore? SecKit? Other?
March 06, 2020 04:46PM
Because I am not now in a position to build from scratch, I posed a question about a commercial corexy that could be improved. I was thinking along the lines of the Ender 3, where for not a lot of money, but a fair amount of time, people get a pretty good machine. I hadn't appreciated that there was such room for poor implementation of the corexy concept.

I wonder about a kit as the better high-value approach.

The SecKit seems pretty good, as does the Railcore, but I have not yet developed the intuition to critique different designs, so I am asking the experts here.

I wonder, too, what tends to become the binding constraint on higher print speeds.

Presumably, at the limit, the properties of the melted filament become the constraint?

Below that, does a smaller machine have a practical speed (mass) advantage over a larger, otherwise identical machine?

What would the speed profile curve look like as you scale up in 3 dimensions?

TIA
Re: Which Kit? Railcore? SecKit? Other?
March 06, 2020 05:55PM
As near as I can tell, the SecKit is primarily a copy of my UMMD printer, so it should be pretty good if the guy paid attention to detail and used quality linear guides.

A lot of people go nutz trying to minimize moving mass so they can print as fast as possible, but in the end, moving the extruder around, even if it isn't super light, isn't the biggest problem. You have correctly surmised that the primary speed limitation in FDM printing is the extrusion. It isn't that you can't extrude quickly- that's the easy part. Extruding in a controlled way, as speed and direction change, is the problem. The Duet controller board running RepRapFirmware has some nice features for tuning nonlinearities in the extrusion.

One of the ways people try to minimize moving mass is to use a Bowden tube extruder so only the hot-end is mounted on the carriage. You can't print flexible material with a Bowden tube extruder. I find flexible material very useful, so I couldn't live with a Bowden tube set up. You may have other ideas.

Here's mine running at 200 mm/sec with acceleration set to 10k (yes, 10k), back when it still had a smoothieboard controller and a completely different extruder/carriage. The X axis+extruder carriage weighs about 1.5 kg. : [vimeo.com]
I don't run it that way normally. I typically have acceleration set to 2k, and print at 80 mm/sec or so to ensure high quality prints.

I'm generally not in a big hurry when I print, so I start the printer up at night before I go to bed. It doesn't really matter if the print takes 2 hours or 6 hours (or 20 hours or 60 hours). It's all the same to me.

I'm not a big fan of the Railcore design. Too many printed parts that extend beyond the frame, making it hard to enclose. The motor mounts look flexible, the pulley mounts don't look very good either. Standing a post with two pulleys up in a piece of printed plastic and putting tension on the belts is likely to cause the plastic to flex and tilt the post. I'm also not a fan of using multiple Z axis motors. They get out of sync when you cycle power and that tilts the bed so you have to have some sort of scheme to relevel the bed frequently. It looks like the design could easily be modified to use a single Z axis motor. It's hard to be sure from the CAD model that doesn't show the belts, but it looks like the A and G belt segments will not be parallel to the Y axis guide rails.


Son of MegaMax 3D printer: [www.instructables.com]
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Which Kit? Railcore? SecKit? Other?
March 06, 2020 09:34PM
Thanks, Mark.

Plagiarism Is the sincerest form of flattery...

I have been looking at the Zesty Nimble direct extruder with remote motor, at 27 grams.

At “overnight” speeds, does a corexy have inherent advantages over an equal cost cartesian with non-slung bed?

I have more to read, but that is the impression I have, thus far.

Edited 1 time(s). Last edit at 03/06/2020 10:30PM by Doppler9000.
Re: Which Kit? Railcore? SecKit? Other?
March 07, 2020 12:30AM
I posted a CAD model of UMMD and a lot of information about problems and fixes for them. That's what open source is about. I'm surprised anyone tried to commercialize it because it isn't the cheapest way to build a printer, and cheapness seems to be the main driver of printer sales.

Most of the non corexy cartesian mechanisms move the X axis motor along the Y axis. A lot of people worry about the mass that motor adds, but I'm more concerned about having to run a cable to it. It's bad enough you have to run one to the extruder, but running two multiconductor cables to two moving motors is asking for trouble. I'm sure it can be done reliably if you apply enough effort, but it isn't my preference. My older printer, SoM, uses a flex ribbon cable to take all the signals and power to the extruder carriage. It has been working with 100% reliability for about 6 years now. If I could find a cheap source of that type cable and tools to terminate it I'd probably go into business selling it to 3D printer builders. I hate running cables to moving stuff and finally took the X axis endstop switch off the extruder carriage in SoM.

Back to your question. The short answer is no, I don't think there's any special advantage of corexy over any other cartesian (except maybe H-bot) when it comes to print quality when you're printing at "normal" speeds. Just about any mechanism will work fine. Corexy is pretty easy to build well, once you understand how it works.

I don't think much of the mechanical complication that something like a remote-motor extruder adds. The whole point of moving the motor off the extruder is to reduce moving mass so you can theoretically print faster, but in order to make it work they have to gear it down 20:1 and that may limit extrusion speed which will limit print speed anyway. What is the benefit that justifies the added cost and complication?

The Duet people are working on a design that runs power and a serial wire to the extruder and that's it. It'll have a board on the extruder carriage that drives the motor and controls the heater and fans. That will be a real advance over the multipin cabling and connectors that are needed now.

Edited 1 time(s). Last edit at 03/07/2020 12:33AM by the_digital_dentist.


Son of MegaMax 3D printer: [www.instructables.com]
Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]
Re: Which Kit? Railcore? SecKit? Other?
March 07, 2020 02:10AM
Quote
the_digital_dentist
I posted a CAD model of UMMD and a lot of information about problems and fixes for them. That's what open source is about. I'm surprised anyone tried to commercialize it because it isn't the cheapest way to build a printer, and cheapness seems to be the main driver of printer sales.

Most of the non corexy cartesian mechanisms move the X axis motor along the Y axis. A lot of people worry about the mass that motor adds, but I'm more concerned about having to run a cable to it. It's bad enough you have to run one to the extruder, but running two multiconductor cables to two moving motors is asking for trouble. I'm sure it can be done reliably if you apply enough effort, but it isn't my preference. My older printer, SoM, uses a flex ribbon cable to take all the signals and power to the extruder carriage. It has been working with 100% reliability for about 6 years now. If I could find a cheap source of that type cable and tools to terminate it I'd probably go into business selling it to 3D printer builders. I hate running cables to moving stuff and finally took the X axis endstop switch off the extruder carriage in SoM.

Back to your question. The short answer is no, I don't think there's any special advantage of corexy over any other cartesian (except maybe H-bot) when it comes to print quality when you're printing at "normal" speeds. Just about any mechanism will work fine. Corexy is pretty easy to build well, once you understand how it works.

I don't think much of the mechanical complication that something like a remote-motor extruder adds. The whole point of moving the motor off the extruder is to reduce moving mass so you can theoretically print faster, but in order to make it work they have to gear it down 20:1 and that may limit extrusion speed which will limit print speed anyway. What is the benefit that justifies the added cost and complication?

The Duet people are working on a design that runs power and a serial wire to the extruder and that's it. It'll have a board on the extruder carriage that drives the motor and controls the heater and fans. That will be a real advance over the multipin cabling and connectors that are needed now.

Very interesting about the moving wires, makes a lot of sense.

Good point, too, about the gear ratio on the remote extruder - just moves the problem, literally.

The Duet product sounds very interesting.
Re: Which Kit? Railcore? SecKit? Other?
June 06, 2020 01:44PM
A lot of misinfo on the Railcore from the Dentist.

Quote

I'm not a big fan of the Railcore design. Too many printed parts that extend beyond the frame, making it hard to enclose.

It is enclosable. You can purchase kits or do it yourself. The motors, except for the extruder motor, are kept outside of the hot build area.

Quote

The motor mounts look flexible, the pulley mounts don't look very good either. Standing a post with two pulleys up in a piece of printed plastic and putting tension on the belts is likely to cause the plastic to flex and tilt the post.

These and many other Railcore parts are available in aluminum. Highly recommended for some parts. I believe you can purchase every part except the fan manifold in aluminum now. Not necessary for some parts, of course, but people like bling. Mandala Rose Works and 713 Maker have specialized in making railcore parts.

Quote

I'm also not a fan of using multiple Z axis motors. They get out of sync when you cycle power and that tilts the bed so you have to have some sort of scheme to relevel the bed frequently. It looks like the design could easily be modified to use a single Z axis motor.

This re-tramming of the bed happens automatically, doesn't take more than a few seconds and levels the bed better than most can do with manual screws. It also accounts for different thickness springsteel build plates automatically.

Quote

It's hard to be sure from the CAD model that doesn't show the belts, but it looks like the A and G belt segments will not be parallel to the Y axis guide rails.

They are parallel.
Re: Which Kit? Railcore? SecKit? Other?
June 06, 2020 04:57PM
Quote
the_digital_dentist
I'm not a big fan of the Railcore design. Too many printed parts that extend beyond the frame, making it hard to enclose. The motor mounts look flexible, the pulley mounts don't look very good either. Standing a post with two pulleys up in a piece of printed plastic and putting tension on the belts is likely to cause the plastic to flex and tilt the post. I'm also not a fan of using multiple Z axis motors. They get out of sync when you cycle power and that tilts the bed so you have to have some sort of scheme to relevel the bed frequently. It looks like the design could easily be modified to use a single Z axis motor. It's hard to be sure from the CAD model that doesn't show the belts, but it looks like the A and G belt segments will not be parallel to the Y axis guide rails.

That's interested to hear you say that. I don't pretend to be knowledgeabe, but I devoured your blog in 2018 (and so many thanks for logging your exploits) and I looked for a printer with :-
  • minimal printed parts (you can replace all the internal parts with aluminium by purchasing and there are, or will be soon, open source versions of all aluminium parts)
  • enclosable/enclosed
  • 3 point bed levelling instead of 4 point bed bending (automatic too!)
  • a Duet (as I was familiar with this board from the Ormerod 2 I'd had previously, you were on Smoothieware at the time, but moved to Duet later)

and so I chose the RailCore II, and have been very happy with it.

(I don't think it's too hard to enclose either - P.S. . this was a handy image with the front door of the top enclosure off, naturally that is closed off too.)


I would say, as a big fan of your blog (and a big fan of the RailCore admittedly) to at least give it another look - if you haven't recently.
Feel free to pop by the Discord if you are so inclined!
Re: Which Kit? Railcore? SecKit? Other?
June 06, 2020 05:41PM
The enclosure can also keep the steppers outside, belts entering thru a slot in the panels. A cleaner look than the rendering above imo.

See enclosure here: [www.mandalaroseworks.com]

Edited 1 time(s). Last edit at 06/06/2020 05:41PM by Wescherry.
Re: Which Kit? Railcore? SecKit? Other?
June 07, 2020 10:12AM
Quote
Wescherry
A lot of misinfo on the Railcore from the Dentist.

...
Quote

I'm also not a fan of using multiple Z axis motors. They get out of sync when you cycle power and that tilts the bed so you have to have some sort of scheme to relevel the bed frequently. It looks like the design could easily be modified to use a single Z axis motor.

This re-tramming of the bed happens automatically, doesn't take more than a few seconds and levels the bed better than most can do with manual screws. It also accounts for different thickness springsteel build plates automatically.

.....

And you NEED this re-tramming due to the motors getting out of sync !

And you NEED to have the bed mounted on flexible joint to the three Z carriage, here they use rubber washers and screws. Not very neat and not a very good kinematic mount either !

A properly designed, built machine DOESN't NEED that.

Edited 1 time(s). Last edit at 06/07/2020 10:13AM by MKSA.


"A comical prototype doesn't mean a dumb idea is possible" (Thunderf00t)
Re: Which Kit? Railcore? SecKit? Other?
June 07, 2020 02:57PM
ICYMI, Don't bother replying to MKSA.
Re: Which Kit? Railcore? SecKit? Other?
June 07, 2020 03:08PM
It's a question of whether you prefer mechanical complexity or electrical complexity. If you are more comfortable with mechanics, you will prefer a single Z motor driving all 3 leadscrews with a closed loop belt, and bed levelling screws to get the bed level. If the frame is sturdy and rigid enough, once levelled it should stay levelled. OTOH if you are more comfortable with electrical complexity, you may prefer three separate motors driven by independent drivers, and no bed levelling screws.



Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].
Sorry, only registered users may post in this forum.

Click here to login