So U want Magnetic Balljoints for $40 or less? No Probs!

[Flateric]

Well I was cleaning up and organizing my junk yesterday and I came across a few things that all just clicked together in my head and I figured out how to do something very similar to the Berry bot. For less than $40 bucks, perhaps even less depending on how much crap you packrat around like I do.

Thanks again to the creator of the Berry Bot this inspiration could never have happened without you posting your machine here and on youtube.

While my solution is not nearly as robust or amazing as your solution, it is cheap, does the job and results in very similar or perhaps even the same accuracy. It also requires no special tools and no fabrication other than cutting your carbon rods and JBwelding some magnets in the end. The only place it lacks in comparison would be the strength your system retains the carriage with.

Alright enough then.....ready? let's go!

The first thing you will want to source for this is 12 - 15mm Ballbearings like this one.

The great thing about these bearings is that they simply snap directly into your hotend carriage directly, no moddin needed. When you do that it will look just like this.

The existing recesses in the carriage very firmly hold the bearings in place and they are not going anywhere anytime soon. Just loosen the tiny screws a bit first so that you don't snap the carriage when they go in. Yes they fit that well.

Next up your going to have to raid your childrens toy chest or go to Toy's R Us and buy a nagnetic construction toy that contains these.

The toy kit will provide you with something like 500 of them. You can fiddle around with the leftovers since they are fun anyways. They are surprisingly strong and hold the carriage very easily even with fast sudden movements.

Lastly you need some carbon rod, hollow, mine is 10.16mm OD and 7.55mm ID (inner diameter, outer diameter) and you will want to cut these into very precise 246mm lengths, you'll need 6 of them.

Then you glue the magnets into the ends of the rods (JB weld works great!) being very careful to make sure they all are exactly the same length. You may want to hang the magnet out of the end of each end on each rod so that you result in a combined length overall of exactly 250mm (10"). Similar to this.

The next part is the installation of the bearings into the x, y and z uprights. This is nearly the same as the hotend carriage but I had to add slightly oversized washers leftover from my build to center and hold them in place. You could also jbweld them here as well.

The nice thing about the magnets is they already have a plastic bumper surrounding them to reduce friction which is ideal for our purposes.

I'll take more pics later, it's super late and I am dead tired.

Be sure and let me know what you guys think! I'll also be happy to answer any questions I can. Now for bed.
I'll post more pics later too.

[MorbidSlowBurn]

I was playing with my nephews yesterday and had the same thought. Only thing that was stopping me was how to ensure the arms were in line with the center of the ball. The berry not does this with the steel pieces that are milled with a spherical pocket. I was thinking of doing something similar but making it printable and have the toy magnet in a pocket in the center. I also came across some neodynium disk magnets at harbor freight over the weekend that may also work and still reasonably priced.

Let us know how a thin wall test prints and what is the layer alignment consistency.

[cambo3d]

do those ball bearings snap in, even after we've done all that sanding to get the stock ujoint to fit? or is a new platform needed for this to work.
nifty idea, post a video when its up and running

[Flateric]

The lining around the magnet keeps it locked in center. And yes, the bearings snap extremely tightly into the spots regardless of your previous sanding because there are what appear to be molded in spots for prehaps a nut? But everything locks in real right and the fricton is gone, it's so smooth.

[ApacheXMD]

I had the exact same idea.

I was tired of sanding and bought a set of Geomag toys. The steel balls in the Geomag set measure exactly 0.50inches so they wont just snap into the platform or carriages. They do however center themselves decently in the hexagonal holes in the platform and carriages, just off to one side.

You might want to check your arm lengths though. The magnets will rotate about the center of the ball, and assuming you have the balls centered at the stock pivot locations, you're going to want your arms to measure about 254 mm. Stock arm length is 269mm from hole to hole. Minus one ball diameter of 15mm gets you 254. This is still aproximate because your magnetic "cup" may vary and may not represent the point at which the rod contacts the ball. I hope that makes sense... Of course you could make the adjustment in software.

I also had the idea of a 3d printable version of replacement platforms and carriage pivot mounts with provisions for balls.

[mhackney]

Very clever. How is the side-to-side movement on the balls? Is there enough range?

[Flateric]

There is just as much as stock, but with not friction. Infact, perhaps too little. I'm having an issue with the solution and will elaborate shortly, hopefully after solving it of course!

[Flateric]

Well, I am not entirely sure what is going on here.

It seems to me that the carriage does not lose steps when on it's way to a commanded position but rather it sort of overshoots once it is at the position it was commanded to go to. And while the smaller cups on the ends of the magnetic plastic toy parts do allow for full range movement they also do not have enough of a width or foundation to them. Especially at accelerated speeds. My other theory is that they rock slightly onto their sides when coming to a stop, again due to the lack of width near there bases coupled with the fact they are not as strong as the Berry bot design.

So I have replaced the bearings in my above photos with 15mm magnetic spheres. Perhaps even the same ones that Berry has used in his design? They are available at my local teachers store for 3 bucks each and available at Lee Valley Tools for 4.95 each so a similar deal should be available near you likely.

My current step is trying to find a cup system similar to Berry's, but if at all possible do it without any machining at all. So that more people could easily access the mod without special tools.

It may be that the 15mm stronger magnets in place of the bearings would overcoming the "overshooting" or orcking issue. I'll keep you posted and feel free to contribute any thoughts you may have.

My first thought is that I should not have posted so quickly in my excitment, so sorry, but exploring a solution as a group my be a better idea anyways.

The 15mm magspheres still fit perfectly in the same locations on the hotend side, and I found metal cups for the xyz uprights the bearings snap into.

So at this point using off the shelf items all that remains is the interfacing of magspheres to rods on uprights and on the hotend side.

Some thing that easily is attached to the rod, holds the bearing, is influenced by the magnet yet still articulates smoothly. I'm having some success with hexcap bolts that thread into the rods and have a tiny piece of sliding drawer teflon tape over the cap, but too soon to tell.

[foshon]

It looks like he drilled out a SHCS for the rod ends.

[Flateric]

Not sure what a SHCS is?

No drilling of any type has been done on anything as of yet......

[MorbidSlowBurn]

SHCS= socket head cap screw.

And yes, he used a ball end mill to put a cup recess in them.

[geneb]

Ah, there you go. A 1/2" ball end mill in a drill press would do a perfect job.

g.

[cambo3d]

I think this might be a better idea, find one in 15mm diameter with a tapped hole for a rod. http://www.precisionballs.com/Kinematic ... php#page11
http://www.precisionballs.com/threaded_ball_2.php

[MorbidSlowBurn]

Cambo3d- they are interesting but at $20+ each I think that is a little more than most would be willing to spend. It translates to $240+ just for the ball ends not even counting any other parts that are needed.

[Flateric]

I have now sourced ferrous metal cups, a nice teflon liner that snaps into this cup.

Nice smooth movement, very strong hold. Required JB welding the magnetic balls to the end of the rods.

Currently solving the cup mounting to the carriage. Then I should be good to go really.

Stay tuned.

[avdflyer]

How about these jb welded to the end of a carbon fiber rod with ball bearings in the carriage?

http://www.ebay.com/itm/36-Neodymium-Co ... 460dfd636e

or these
http://www.ebay.com/itm/5-x-Strong-Neod ... 2a20420fea

[harley573]

What size rod is everyone using and where are you getting them?

[harley573]

So do you have full range of motion? I modeled something up in Solidworks and it looks like my range of motion is limited. Maybe because I drew the bearing cup at the same diameter as the rod, (although the rod I used is smaller than what you posted.)

[MorbidSlowBurn]

I am still refining a cup ball magnet setup. But I do plan on using carbon rods that I have on hand. I normally use them for building kites. Another option for rods are arrow shafts.

Diameter wise I will probably select something about .25". Should give sufficient stiffness while having a good surface for bonding.

Once I get a little farther along with my design I will post pictures. My target is to minimize machining (no special tools) and to keep purchased parts down to a minimum for cost.

[type]

Im really following this, I dont think I will be purchasing the Rostock MAX Kit until I see some further improvement in a few areas, this being one of them.

Keep it up!

I also think Berry will be making a detailed post with some more info on a few parts of the Berrybot. Hopefully we can get some more info that is usable there too.

[harley573]

I sat down and did some modeling this afternoon along with a little price checking and came up with a source for magnetic balls. I am planning to make up a set of rods with mag-ball joints at either end. Mine will have precision machined and polished contact points, but once I have the machine set up, it's no big deal to keep running it if others want a set. From what I can tell, a set would run in the $40 - $45 range complete. I am even going to set up a jig that will allow me to epoxy the ends in place to obtain a precise length. I need to gauge interest in order to purchase the appropriate material quantities, so if you are interested in a set, PM me and let me know. My design has a 15mm ball snapped into the lower carriage, so I need to confirm that size is correct, (flateric)? It mates to a precision steel part that will be epoxied into an aluminum tube.

[cambo3d]

you should probably test it before you try selling it. looks good. The balls arent going to snap into the top like you show though. there is no recess for the ball in those mounts on the cheapskates.

[type]

If you can get a good set up for 40 bucks you will for sure have interest.

Im already interested.

[Eaglezsoar]

I am interested if the carriage and the cheapskate attachments are included in the final cost.

[harley573]

My plan was to build a set for my machine and give it a test run, but it sounds like I need to re-think the upper end? Maybe the OP can post a photo of his setup for reference? I was planning on including tubes, bearing cups, and mag-balls with the bearing cups epoxied into place.