A version of the ball joint that allows you to manage large angles with continuous modulation of the clutch and that maintains reduced dimensions and strength, combined with a modular system that can be completely configured in any size and shape, to cover the different needs that may arise.

Starting from the joint, which is the most elaborate part of the system, you can see that there is an evolution in the model, but each version has some features that are worth considering to see which one suits the needs best.
In addition to the structural and performance differences, there are also differences in the work required for assembly.
Versions 1.0 and 1.5 are similar for the assembly procedure and differ in load capacities, which are obviously greater in version 1.5 which is slightly larger.
Version 2.0 is an evolution in which I tried to exploit the greater resistance of the printed material in the direction of the effort it has to support instead of with the perpendicular layers and thanks to this system I was able to make a joint that is resistant as, if not of more than version 1.5, but with dimensions smaller than 1.0.
The price to pay for version 2.0 is more work required for assembly.
All versions need a running-in period, and in all versions the concept of tolerance is managed differently, but always aimed at ensuring that after running-in the necessary clutches are the ones needed, as well as being continuous through the load placed with the tightening or loosening the two external parts that will compress the balls on the two braking pads (as well as on their own walls).
At the heart of the system there is the element that keeps the clutch modulation high and guarantees a prolonged duration over time, after running in. this element consists of a rubber disk, which can be obtained from any inner tube of a wheel or from any other part that has that type of rubber (even a piece made from an elastic has the same characteristics and with an elastic as those of 5 or 10mm wide, disks are obtained for a long time).
The thickness of the rubber needed will vary over time and will allow you to recalibrate the clutch every time it becomes impossible to increase it simply by turning the bodies, but thanks to this you will always have an effective joint, and thanks to the elasticity of the rubber the thrust will continue. to exist even with the consumption of plastic requiring a less frequent recalibration, with the effect that will increase in duration over time in a proportional way to the coupling of the parts that rub.
Obviously, the fluidity of the movement will also increase over time.
To obtain the performance and keep the dimensions small, not only plastic can be used, but metal must be used: the axes that fit into the balls are 3mm screws without the heads, which can be obtained from simple screws of length suitable or by threaded bars, always cut into parts of the necessary and desired length.
The assembly work therefore consists in obtaining the rubber and the screws and in cutting everything to size and then screwing them on the spheres (which are cut in half for an optimal print).
The size of the screws cannot be less than the length necessary for the assembly of the two halves of the external bodies ringed between the two spheres, but it can be easily greater, for special needs (it must be considered that extreme angles are covered with a single joint and with more than one in series, contosional maneuvers are made, if necessary), so maybe for someone even just one may be enough, but with a longer screw, perhaps to be associated with one or more twins upright frontally on a truss made with forms. Or you can mount more joints in series for the third, fourth and fifth hands ... the limit is only the imagination, (see pictures
We come to the tolerances: as soon as they are molded they will be hard to screw and the balls will move very little smoothly and that is why I talked about the running-in phase.
For versions 1.0 and 1.5 I recommend an assembly of the two shells, to be screwed and unscrewed several times, until they become smoothly flowing threads and I suggest to clean the threads at each step, because the dust tends to seize and once cleaned to lubricate slightly before screwing it back on (it will only take until they have dissolved obviously, a matter of a minute).
Once the threads have been loosened, you can insert the spheres with the axes mounted and do a lot of gymnastics, progressively increasing the friction, until the steps of the print layers settle, enough, the rest of the fluidity will come with time. If you make chains of more joints, it will be more complicated at first, I have them broken in and then I coupled them.

Version 2.0 is a little more complicated to assemble due to the fact that the printing of the threads horizontally is more critical and to guarantee an effective seal it must be molded more "tightly", so you will find the work of first coupling the threads extremely hard. , but still feasible and which will guarantee reliable, robust and smooth threads when the procedure is completed.
Another problem of horizontal printing without supports is that the upper parts of the sphere (shell) are much dirtier than in vertical printing, so when you insert the sphere to be run in the walls to be run in, you will not be able to screw the sealing ring of the two half, on the side of the sphere. this is why the assembly and running-in tool comes into play which will settle on the pieces to be run in, even if at the end of running in, it will not be able to guarantee a good seal, but you will only need it for the assembly and running in of the ball and of the walls and then you will insert the regular ring nut, so even just one could be enough to print for all the joints you will make.
In short, as announced, the work is more but then the size / performance ratio will reward the sacrifice of those who want to face it.
Fundamental for all versions is to ensure that the two halves of the sphere are breaded straight on the screw, because otherwise they will not approach well and the sphere will not be regular, creating various problems and in extreme cases, compromising and preventing operation.
If in screwing and unscrewing they loosen too much, so much to turn on the finished arm, annoying the accuracy, you can put a drop of superglue that will restore the brake on the axle, but which will not prevent disassembly if necessary.
To manage the assembly of the spheres on the axes, I do not recommend the use of tools that can damage them, I use a piece of the same rubber that I use to make the discs to put in the core and my hands (perhaps folded if too thin), holding the other side of the metal screw with the pliers.
However, for the whole assembly it is intuitive, especially for those accustomed to using third hands and other related tools.
In any case, for any need, write to me and the same also applies to any suggestion or advice.
For version 3.0 the speech is separate: it is a completely plastic version, with some limitations in the bending angles that go down to about 25 + 69 degrees, but thanks to the printing made so that the resistances are greater where they are needed, the performance remain good. The great advantage of this version is the ease of construction and assembly, which make it more friendly for those who do not want to do too much post-print DIY, just be careful to correctly couple the two halves of the balls so that the thread is correct, because they are two couples and not four twins.
there remains only the need for the rubber disc for the core and the running-in procedure, but in a more minor way.
The threads of version 1.0 and version 3.0 in the external shells are the same, as are the internal balls, so that you can also make mixed couplings between the versions, if necessary, for upgrades or variations over time.
There will probably be upgrades and variants for this version.
As mentioned at the beginning, everyone can choose the version he prefers for the characteristics that he thinks are most important.

We come to the modular structure, I don't think there is much to say about what to do with it and how to do it, as you will choose it and choose to print the pieces you will need.
In the pictures, a small example of what can be done.
The tips for printing are to use the orientations used in the slicer images and not to use supports for the inside of the female threads, otherwise you will be in trouble.
As for the assemblable bases, there are three tolerance versions which will guarantee you can choose the coupling hardness you want and manage everything based on the characteristics of your printers.
Nothing prevents you from mixing the tolerances to find intermediate hardness values.
As for the elastic tweezers, the cylinder that is together with the print must be forced into the space in the tail and serves to increase the spring load.
If I see that it will be necessary or something else comes to mind, I will bring edit to this explanation.
I printed everything with 0.2mm layers in PLA without difficulty, but no one forbids using thinner layers (just in case let me know what changes).
A small final note: the logo you find on the joints is that of Blender, the program I used to make them, to which it seemed the least I could do to dedicate them to it, with the hope that it will also make a mini-mini contribution to its ever-increasing diffusion.

Greetings to all

Edit 20/02/2021
Given the good resistance of the horizontal print that is beyond my expectations, I added a new pair of axes and balls for the 3.0 joint version that extends the angle to 35+69 degrees, so that the difference in terms of bend with versions 1.0, 1.5 and 2.0 it is minimal and can be useful to many for less force demanding uses.
I have also added some couplers for the axes in plastic and metal, so that you can interface the 3.0 also with the 2.0 and doubles the possibility of interfacing with the versions 1.0 and 1.5.
Since the shaft is smaller I will have the possibility to increase the 69 degrees and get close to 90 degrees, with a small modification to the shell, and as soon as it becomes available I will update it, presenting the whole joint as version 3.1.

Greetings to all

Edit 21/02/2021
I uploaded version 3.1. The bend angles are now 35+85 degrees, so the difference with the "metal" versions in terms of mobility is minimal, but the advantage in terms of ready to use is a lot.
I suggest a little care and attention in the running-in phase of the balls, because it is possible that the necessary forces are very close to the maximum load that the axis can withstand, so if you proceed in small increments there are no problems, but if you if you want to go a little too fast you risk breaking or weakening the axis, compromising the final yield. In any case, if you have problems you can always reprint only the spheres with the axes, which are the ones I had put as an upgrade.

Greetings to all

Edit 21/02/2021
I added a new axis in version 3.2 (I will also put the stl bundle of all the joint in version 3.2) compatible with versions 3.0 and 3.1 (in this case the shells are just the same) which optimizes the distance between the joints.
It has the same features as version 3.1 (35+85 degrees).
I tried to make the most of the available space, in terms of workable plastic without breaking, by directly threading the inside of the hemispheres.
The thread guarantees an optimal seal, despite the dimensions, but obviously everything must be seen in the context of the possible forces on that type of axis.
For the first assembly (even if only in replacement of version 3.1 or 3.0 if you want), you have to couple and run in the thread (a bit like it happened in version 2.0 or in general) because it is very tight tolerance and to do so, if you have a shell already run in, I suggest to insert it and mount the entire joint well tightened, so that the axis has the thread well aligned.
once the axis is coupled and aligned, it is necessary to make the half sphere with the through hole turn over it, sometimes, at least until the threads are well run in. In this process be careful because the print on the long part of the axis seems to be very weak to twist, so you have to proceed in small increments and go back, cleaning and lubricating the thread.
When the coupling is complete, even with the second half of the sphere (but it will be easy, because the first will have already done all the work), you can insert the female sphere in the other half of the joint and you will be ready to assemble it.
This version will always appear as two halves ringed on the axis between the spheres, so to be mounted you will have to print a version 3.1 or 3.0, which foresees the interfaces with the rest and put the two halves at the end of the chain created with version 3.2 ( or with a single if you prefer).
I will see if to create a direct interface for version 3.2, but for the moment I think it is not a good idea, because I am afraid that it is too fragile if coupled with something rigid and not another joint.
But fundamentally it is not the few millimeters of a more robust interface, like the previous ones that make the difference, the important thing is to have optimized the length in relation to the bend, in the whole chain.
The limit of the forces that can be applied and supported remains present, compared to the previous versions, but also in this case by losing something you gain something else, so this is another possibility of choice and adaptation to the preferences and needs of each.

Greetings to all

link to Low Profile joint:

Link to Low Profile Joint