/tacovr

An infento-based omnidirectional VR treadmill

Primary LanguageC++GNU General Public License v3.0GPL-3.0

TACO VR

An Infento-based omnidirectional VR treadmill (unfinished prototype!)

Screenshot

Links

Similar projects

Infento parts list

parts

order them at the infento parts shop

  • profile 560mm E12,71 (x4)
  • profile 440mm E10,01 (x4)
  • profile 280mm E6,41 (x2)
  • profile 240mm E5,51 (x6)
  • profile 160mm E3,71 (x3)
  • profile 140mm E3,26 (x1)
  • profile 130mm E3,04 (x2)
  • profile 120mm E2,81 (x5)
  • profile 100mm E2,36 (x2)
  • profile 80mm E1,91 (x4)
  • profile 60mm E1,46 (x2)
  • profile 40mm E1,01 (x2)
  • profile spacer 5mm E0,11 (x6)
  • axle 550mm E6,53 (x4)
  • axle 166mm E2,01 (x3)
  • axle 122mm E1,49 (x2)
  • axle 101mm E1,24 (x2)
  • axle flange E8,60 (x10) (ugh!)
  • spacer 50mm E1,62 (x2)
  • spacer 10mm E0,59 (x4)
  • spacer 5mm E0,50 (x5)
  • spacer 2mm E0,48 (x3)
  • spacer 1mm E0,46 (x8)
  • L-joint E1,35 (x42)
  • bearing block 40x80 hexa E9,52 (x2)
  • bearing block 40x40 hexa E9,17 (x6)
  • 4 inch wheel E4,58 (x14)
  • knurled disc 2 stripes E0,60 (x4)
  • multi angle joint E1,43 (x8)
  • axle clamp E1,93 (x8)
  • clamping part with M6 thread E1,66 (x20)
  • clamping part without thread E1,66 (x11)
  • adjusting bolt M8x35 E0,24 (x11)
  • M8-16 washer E0,01 (x58)
  • T-positioning block E0,17 (x250) (may not be enough)
  • profile nut short E0,30 (x120) (may not be enough)
  • M6x12 bolt E0,04 (x50)
  • M6x20 bolt E0,04 (x30)
  • M6x25 bolt E0,07 (x14)
  • M6 nut E0,02 (x8)
  • M8x25 bolt E0,09 (x3)
  • M8x40 bolt E0,12 (x14)
  • M8x45 bolt E0,16 (x4)
  • M8x20 bolt flat head E0,16 (x120) (may not be enough)
  • M8 nut E0,02 (x4)
  • M8 locknut E0,03 (x4)
  • toothed belt pulley 26T + protection ring E2,16 (x5)
  • toothed belt pulley 46T + protection ring E4,34
  • toothed belt 100T E8,79 (probably 118T works better)
  • toothed belt 118T E10,28 (x2)
  • profile connector B E1,46 (x2)
  • parallel clamp joint E0,92 (x4)

Electronics components

NOTE: stepper motors are probably not a great idea. for the next iteration I will probably go with (teknic clearpath?) servo motors. servo motors are more powerful, closed-loop so don't miss steps, come with builtin drivers, and are practically silent.. as well as quite a bit more expensive.

3D-printed parts

Other parts/tools

  • M6 thread reamer
  • M6 drill bit
  • M8x35 sunken head screws (x3), for center hole in platform/base
  • round base plate, diameter ~130 cm
  • small round platform ~18mm thick, diameter ~29 cm (x2)
  • tape to go between axles and gears, to avoid any movement (depending on 3d printer precision)

Screenshot

Cost estimation

  • electronics: E545,32 (excluding shipping)
  • infento parts: E760,83 (excluding shipping)
  • total: E1306,15, more realistically around E1500 including shipping, wiring, wooden parts..

Background information/calculations

measurements with the android app physics toolbox accelerometer show that the acceleration of my feet does not seem to exceed 1G for casual walking, and 2G for pretty rapid walking.

using a hanging weight and PTA again, I found that we may need on the order of 50Nm to rotate just the platform itself with 1G. this would mean that a NEMA34 servo motor would not even cut it.

from this video on youtube, we can approximate the required torque. the moment of inertia for a person of 80 kg is similar to a cylinder with diameter of 40cm: I=0.5 * m * r^2 = 0.5 * 80 * (0.2) ^ 2 = 1.6. the moment of inertia of the platform of about 10kg is similar to a cylinder with diameter 1.3m: 0.5 * 10 * (0.65) ^ 2 = 2.1. the rotational acceleration alpha for 1G is 9.8m / s ^ 2 or 39 rad / s ^ 2 if the feet are at 0.25 from the center. the required torque is therefore (1.6 + 2.1) * 39 = 131Nm.

131nm of torque, or something close to it, may be achievable with the most expensive teknic clearpath NEMA56 servo motor (about 12Nm) and a 10:1 gear reducer. this would be quite expensive though and we might get in trouble with the max RPM (goes down by a factor of 10 with this gear reducer).

Current status

Screenshot

the project is currently on hold, as I think about the required torque for rotation.

TODO

  • detect step-down (speed approaching 0?)
  • test rotation using servo motor
  • make video showing walking/turning in several directions :-)