Moving to MGN-12 rails - Carriage Advice
-
Hi,
I've got a set of robotdigg mgn12 rails for my kossel/Griffin hybrid. I was going to hold off on using them until the PCB effector and carriages are available.
I've since discovered a 0.5mm mismatch in the spacing of one set of arms that I think is contributing to effector twist. It's not something I can easily fix, so I'm looking to replace the carriages and move to the rails earlier than expected.
Has anyone got advice on a suitable set of carriages for mgn-12 rails and Hayden Huntley's magnetic joints? I started designing a set here http://a360.co/2p9rq82 but stopped because I assume someone has already done a better job!
Thanks
David
-
Haydn has published a design himself, see http://www.thingiverse.com/thing:2183355 and http://www.thingiverse.com/thing:2183347. Looks like they are designed for wheels not rails, but he provides the OpenScad source, so you could replace the 3 wheel axle holes by the usual 4 mounting holes on a 20mm square.
-
There are some parts on eBay that are cast aluminum that should work well for a kossel. They also have arms and an end effector. Not bad IMO
-
Doug and myself and a few others are using water jet cut aluminium carriages and effector. I'm not sure there are any left but there is a design I'll find it for you and post it. Or hang on for David's pcb effector and carriages coming soon.
Haydn's printed ones work just fine too.
-
May I ask why Robotdigg carriages are not mentioned? The ones I have are specifically made for MGN-12 style linear rails. The linear rails are not by any means genuine Hiwin linear rails, but they're good enough once oiled, don't forget that maintenance though, keep them lubricated because all the oil will just slowly make its way down eventually.
Just know the threads on the carriages on the MGN-12 rails are not very deep, I had to buy screws specifically for that application, I expected to be able to use some of my spares. They even have an integrated belt tensioner style, the kind I use, I just changed the cheap screws that cane with then go hardened or stainless steel panhead/cap or socket screws that use an allen head rather than Phillips head. I have used them for about 4 months, only problem is I didn't use thread locker and I probably should.
They come pre-tapped with M3, it should be easy to find magnets with M3 thread.
-
Its because they don't fit Haydn's magball attachments, and are not the right rod spacing for most of the Haydn-system compatible effectors.
-
Its because they don't fit Haydn's magball attachments, and are not the right rod spacing for most of the Haydn-system compatible effectors.
Bingo. I have a set of these and they are simply not suited to Hayden's joints.
I'll probably finish the work on my own ones and publish to YouMagine. Im not that good with OpenSCAD so it's probably quicker for me to do. I was trying to copy the new style robotdigg belt tensioner but I think I'm going to swap to the Griffin style instead, they work so nicely.
Knowing my luck I'll get them finished just as Duet3d/E3D have their new carriages/effector/hotend set ready, but not everyone will be upgrading so maybe someone can make use of mine anyway….
David
-
http://m.ebay.com/itm/3D-Printer-Kossel-Mini-Delta-Reprap-Aluminum-Slide-Slider-Pulley-SET-OF-3-PCS-/111949748948?nav=SEARCH these are good but I don't think they are mgn12
-
If your rail carriages are 12H then they have 20mm hole spacing and those carriages will fit. However they don't have adjustable belt tension unlike other similar designs and won't fit Haydn's magnets, unless you rather unconventionally attach the mag balls for the holes where the traxxas rod ends would be attached, ensure even spacing on each of the three carriages and then design an effector with that spacing. You couldn't guarantee the full range of movement as the angle between the arms and the balls will be too extreme.
-
If someone designs a style of robotdigg style carriages made for MGN-12 for Hayden's magnetic ball joint….thing...I'll try to manufacture them, fairly low volume but I could. I'm not sure how to make those silver half round, half gear teeth, tensioners that robotdigg sells...how do you clamp that in a vice to mill it?? Are they cast?? Thats probably what I would do, I could mill them but if it's possible to cast them accurately...that would be much easier and much less expensive for higher volume production, I'd imagine.
I don't have Hayden's magnetic delta arms currently so I have a hard time imagining the problem, and as a result...the solution so if I have a file to work with, I could probably do it. I am NIMS (National Institute for Metalworking Skills) certified in a bunch of different areas. I am new to designing so I'd need a 3D file. I don't think I can use an Openscad or an STL for Autodesk Inventor HSM CAM simulation...but I'll try when I get home in a general STL. Might just be easier to make a simple T-Shape, something easy to machine even if it doesn't look beautiful, it'll work, maybe the same recessed slot like robotdigg' s so those tensioners can be used....although, milling one side and flipping it over might work to make those tensioners.
-
I've been tweaking the Fusion 360 design linked in my first post along the lines I wrote of. The tension is now set by a bolt screwed in from the top into a captive nut in the sliding section, just like the Griffin carriages. I've got a couple of clearance issues as you can probably see but it's at the point now where it's ready to print and test.
The hole spacing is set to 48mm right now to match my effector but it's parametric so can easily be changed.
I'm struggling like crazy with a head cold but I'll try and get something running this weekend.
@NoSkillzEngineer - It would be a hack as the profile would be all wrong, but might be worth playing with the technique used in this video to create those odd RobotDigg parts if you wanted to have a go:
https://www.youtube.com/watch?v=s1K8pawI-9M
As you can see they drill around the periphery and then turn down the outer to make the tracks.
-
and, typically, I've just noticed a fairly large screw up just after posting.
That model has the wrong profile on the belt holders, it needs inverting.
-
Twist your belts, on a delta this works as they are long and it isnt going to cause a problem.
-
I don't thing I have them facing the wrong way, do I?
It's more that the profile itself is wrong. I modelled the 2GT belt profile into the part, where I should have used that profile to cut the part. I've fixed it on one side now and I can easily do the other. It's not that critical to get the profile perfect but I doubt a belt would have squashed into what I had previously.
-
Ah I thought you meant the belt retainer was facing the wrong way as they are in some carriage designs. I twist the belts over the top pulleys for now but I might buy toothed pulleys.
-
For the belt tension could you use something like this:
Then machine out a small rectangular space that it could screw into? Make it slightly deeper then you could pull the belt tension by hand and then as you tightened it into the machined out groove (and it got closer to the bottom of the machine out space), it should tension the belt a few more millimeters?
-
Apologies for the long delay in replying, work has been incredibly busy this last week or so.
The model at http://a360.co/2p9rq82 is now pretty much complete and seems to print quite nicely. I've a tiny bit of tweaking to do on tolerances for my particular printer but that's easy to do and the parts do go together right now. My only remaining concern is the possibility of the joints fouling the tensioner/belt at extreme angles. Again, easy to fix if so.
Also, and I know this is the wrong place to ask, have there been any updates on the PCB carriages/effector? I'm wading through a mountain of unread threads now and just wondering if I've missed some news.
Thanks for the link to those parts Saffi, that could be a neat way of getting much more strength in the design if needed.
adavidm
-
We are about to ship PCB effectors and carriage adapters to 3 beta testers.
-
David am I on that list? I suspect not
Doug
-
Doug, please send me a PM or email with your real name, email address and reprap.org forum ID, and I'll check.