-
Moving Encoders from Motor Shafts to Arm Links
02/12/2023 at 23:57 • 0 commentsThe revision 2 of the RR1 robotic arm is progressing. The experience from rev. 1 tells that having encoders on motors shaft does not give enough accuracy, so I decided to move encoders from motors to arm links as shown in pictures (the pictures show two main joints of the arm and the link between them). Fortunatelly the design of RR1 is quite flexible so this change could be smoothly integrated in the current design.
The encoders are fixed to aluminum extrusions representing the link above or below the corresponding joint. The torque is transmitted to the encoders via belts from the rotating part of each joint. There is a significant reduction ratio between the joint rotation and the encoder rotation to increase accuracy.
In addition to redesigning the encoders, there are multiple improvements in gearboxes of the joints. The rotating parts are fixed with more bearings around which slightly increses the weight of the joints but significantly reduces the backlash.
The precise rotation tracking with the new encoders will futher contribute to reducing any backlash in the joints. I also plan that the new design will allow for more collaboratives modes of the robot such as lead through motion recording.
-
Assembling the Planetary Gear Reducer
08/21/2022 at 12:53 • 0 commentsReducers are a key component of the RR1 robot arm. I use my own custom built 3D-printed planetary gearboxes. The following video shows assembly of the gearbox for the upper joint (elbow join) for the second prototype of RR1 ("rev.2"). There are a total of 6 similar gearboxes of various sizes across the robot.
As the gearboxes in RR1 use herringbone gears for smoother rotation, the assembly must be carried out exactly as shown in the video. All gears must be meshed together well so that there is not much friction and not much backlash at the same time which is especially difficult to achieve with herringbone gears. The video shows that the assembled gearbox rotates smoothly and has no detectable backlash. Such an achievement requires lot of experience and a very good 3D printer (I use Prusa i3 MK3S+).
And as you can see "rev. 2" will be orange.
-
Speaking about the RR1 Robot
08/14/2022 at 12:06 • 0 commentsThis is a talk a had at Prusa Lab in Prague in June 2022. It is in Czech so probably only for the Czech audience but at least it shows me how crazy I am about robotics.
-
Planetary Gearbox Published on Github
08/12/2022 at 12:09 • 0 commentsDue to some requests I started to publish the files concerning the mechanic construction of RR1. I started the Github repository of the RR1 project where everything will be collected: https://github.com/surynek/RR1.
Currently there are the STL files for the planetary gearbox reducer of the main shoulder joint (shown below) for RR1 rev.2 (the one under construction, not the black one from videos). I recommend to use Prusa i3 MKS+ for 3D printing the parts. In addition to this, 7 ball bearings are necessary (16mm outer diameter, 5 mm inner diameter, 6mm height), 3 long bolts for planets (48mm length, 5mm diameter), 1 short bolt for the sun gear (24mm length, 5mm diameter), and 7 small washers (5mm inner diameter), and some 3mm screws, many 3mm square nuts, and very many 6mm bearing balls between the actuator ring and the outer ring. Please note that bolts, screws and bearings are not shown in the figure.
![]()
-
Shifting the Arm Forward
08/06/2022 at 12:10 • 0 commentsThe rev. 2 will not have the arm mounted in the center of the base. The arm will shifted a bit forward which will make the RR1 arm look more like professional robotic arms. The figure shows the part on which the arm will be mounted, this part sits on the base that changes only little from rev. 1. The reason why the arm will be shifted forward is not only visual but most importantly it improves the reach of the arm and makes the working space more natural.
-
Switched from Completed to Ongoing Project
07/25/2022 at 19:02 • 0 commentsI decided to switch the project from the completed one to ongoing. Anyway it seems that is true. The first prototype of the robot, revision 1, or rev. 1, is completed, but the revision 2 is on the way, so it seems it is better to mark the project as ongoing.
I am printing new parts for RR1 rev. 2, everything is done more precisely to minimize backlash everywhere possible. The picture shows printing of the custom bearing for the base of the robot.![]()
Pavel Surynek
