Design of a timing belt actuator

I wonder if anyone has ever used linear motors?

http://www.aerotech.com/products/motors ... otors.html
http://www.youtube.com/watch?v=PrktRSFI ... re=related

sn4il

Linear motors are mainly intended for horizontal applications. For both the belt and even more for the linear motor you have to foresee a braking mechanism in case of loss of electrical power – or you will fall down ½ m with 1g acceleration
Additionally powerful linear motors are very expensive.

The beauty of the Quanser design is clearly the low vertical room needed and no obstacles in the way of the platform when moving.
Also if you fix the left clamp on the lower side of a belt and the right clamp on the upper side of a belt, you can drive both with the same motor.

Step and servo motors have a fairly stable torque over a wide range of rpm. One might still need a gear to achieve the high power on the belt (I calculate about 25Nm per axis).
I haven't found the affordable stepper yet

but you need to built your own controller before make it usable with x-sim i think.

Finding a cheap planetarygear wont be easy.
regards

I'm yet again finding myself to favor the ball screw design over the belt ones. I fear that even the Quanser design would need gearing as well. There really don't seem to be cheap planetary gears around that I would gladly bolt on to anything. On ebay there are some cheap Chinese electric scooter gears but I won't even go to that as they are really crappy looking sintered metal gears...

Have a look at Ideas about MK IV:
http://jippes.home.xs4all.nl/hi6sim/MK_IV.html
Regards

Hi douwe,

That was exactly the thing I first wanted to build. On my design the push rod would have been in the midle of the belts and motor would have been perpendicular to the actuator and not pointing on one side, eg no worm gear reduction. I'm still looking for a suitable planetary gear and if I find one I'll build a belt based actuator, but for now I'm thinking of doing one with ball screw.

Hi Jyrki
Youre last post made my think
make a quanser actuator screw belt driven
Make from leverarm clamp on lower side from belt ,ballnut housing to

I think I may be able to get my hands on some Yaskawa robot arm servo motors and I'm hoping they would be suitable for my purposes, either for the timing belt or for the ball screw design. The motors in question are rated at 400W and they have a peak output power of 2kW.

I've placed an order for one Chinese brand ball screw nut (3205) for prototyping purposes, let's see if the ball screw design will give better results. With the timing belt I needed a reduction of 1:10-1:20, but with the ball screw I need the motor to turn at 6000 rpm for the given 500mm/s speed... Well I'd most likely be happy for the 250mm/s speeds, but then again I could use 2:1 belt pulleys on the screws with 3000rpm motor, would still be cheaper than a reduction gearing or I could of course use different pitch ball screw but that would take more torque out of the motor and the 5mm pitch is a compromise of some sort. That and the fact that I all ready had them. Those ball screws I'm planning to use might be a bit of overkill as they are 32mm in width.

Those robot arm motors were not suitable as they were low voltage and high amperage motors. I'm not planning on rolling a +60v@100Amp power supply.

Ok... I got a bit further with the simulator design. I have dropped the idea of using timing belts because of the expense involved with the gearing and started prototyping the actuator with ball screws...

This prototype screw is still a bit long, for the final version and I'm aiming somewhere around 400mm for travel. I need to check the overall geometry of the thing so it will fit inside my house (2500mm wall height)

[img600px]http://www.anvianet.fi/jyrki.koivisto/simulator/pics/DSC05648.JPG[/img600px]

Screw is 32mm in diameter and the pitch is 5mm, quite beefy and should cope nicely with the forces involved. It will be driven by a 22 teeth timing pulley (480mm long HTD 8M 20mm belt)

I'm still trying to find suitable motor to drive the thing and I've recalculated the forces for a bit weaker simulator.

With 250mm/s maximum target speed and 1G acceleration (maybe too much) and with the weight of the moving part with passanger (200kg) I got these figures.

Force needed on axis (up) 1340,24 N
Force needed on axis (down) 163,44 N
Force needed on axis (stall) 523,02 N
Torque needed on axis (up) 1,07 Nm
Torque needed on axis (down) 0,13 Nm
Torque needed on axis (stall) 0,42 Nm
Required motor power (up) 335,06 W
Required motor power (down) 40,86 W

The upper bearing will be made to take the brute of all the forces and the lower one is just for the pulley and will not be subject to large forces. Since the picture was taken I've removed the shaft collar and substituted it with a circlip also there will be a collar welded to the screw where the thread starts as to transfer forces better to the main top bearing.

On top of the screw you can see my first version of sliding unit that keeps the screw centered in the axis tube, it will be replaced by a needle bearing and longer plastic slide. Remember that the screw rotates at 3000rpm (max) and I don't want the screw to twirl around inside the tube, needle bearing will take away any rotational movement in the plastic slide, it will only move up and down to reduce wear on it.

Hi
Are you gone use a dubble roll angularcontactbearing for upper bearing?
Ad

Yes the top bearings is double row angular contact type. Limiting speed is 9000rpm and static load rating 27,5kN (dynamic 41,5kN) should suffice for the application.

Axis will be 40mm inner diameter (50mm outer) aluminium tube, haven't decided on the outer skin for the axis, maybe plain sheet metal or another aluminium tube.

Top bearing will be sandwiched between the plate holding the motor (20mm thick aluminium plate) and cover made of the same material. I don't know yeat how I'm going to mount the outer tube (if using a tube) on it though. Metal sheet would be easier, but then I couldn't make the ball nut to have a slide surface as easily as with a tube. I'll post some hand drawings of the design (when I get to actually draw something down... )

[img600px]http://www.anvianet.fi/jyrki.koivisto/simulator/pics/DRAWING.JPG[/img600px]

Here's a better picture of the bearing assembly, upper bearing and the pulley will shift a bit more to the top.
[img600px]http://www.anvianet.fi/jyrki.koivisto/simulator/pics/DSC05650.JPG[/img600px]

[img600px]http://www.anvianet.fi/jyrki.koivisto/simulator/pics/CALCULATIONS.JPG[/img600px]

Here's the screw without bearings. I remind you that I grinded it on a milling machine (manually) so the surface is awfull, but it's quite evenly awfull so the bearing fits snuggly. This is old picture without the grooves for circlips. Screw is hardened and the hardening goes quite deep, those grooves for circlips took some time to make with Dremel.

[img600px]http://www.anvianet.fi/jyrki.koivisto/simulator/pics/DSC05653.JPG[/img600px]

Timing belt arrived and here's the screw as it is now. Shaft for the lower bearing will be cut down to size as at the moment it's a bit too long, it will be cut few millimeters (2mm or so) below the lower circlip groove. I was told on another forum that it may not be a wise idea to use circlips instead of locking nuts for the bearings, I may have to make few changes when making the final version if the circlips give problems...

Now I need to make the plates holding the motor (I still haven't got any...) and the upper and lower bearings in place. Maybe I'll do some wooden mockups again as they are far easier to make than making one from aluminium.

[img600px]http://www.anvianet.fi/jyrki.koivisto/simulator/pics/DSC05703.JPG[/img600px]

Coul you post plans? I'm interested in build a new one also. I'm working with a guys that can make this kind of spindle.

Thanks and nice job!! Regards.

There isn't any plans available as this is DIY project and a learning experience at the same time as well. I may post some plans and pictures as I go along with the project, but this will take some time as something like this is a huge undertaking and this will not happen over night. My plans have changed along the way as the topic suggest (belt actuator) just because I'm playing with all kinds of ideas to drive the seat.

Terve, (=Hello in Finnish)

I have gone also long way by considering different LOW COST diy actuator consepts. My first setup will be powered someday by 12V winches. Those winches have about 1kW DC electrical motor with coal brushes and gear ratios 1:100...1:200. You can drive these with 24V if selected amplifier support it. Ballscrews are more professional way to do actuator but costs and all other hardware design and buildup would be too heavy consept for me...

http://www.youtube.com/watch?v=syhtzadjdx8

http://www.youtube.com/watch?v=zojT1iox ... re=related

Prices starts from 70..100€ , in Finland for example Motonet has few cheap models, and also Keskinen has normally one good 69,50€ model ( which was sold out last week grrrr).

I started building the main axis frame where the seat will hang. I decided to go with the Force Dynamics design as I wanted the axis to intersect at the shoulder. I will use angular contact bearings for the axis joint. I will have to mill some reinforment slots at the bends, but I'm quite positive this will hold together... :?

[img600px]http://www.netikka.net/jyrki.koivisto/simulator/DSC06126.JPG[/img600px]

There's only the back bearing hole done as I still have to figure out how far will the other two bearings go. I'll try to get +-30degrees of movement and the front axis will dictate if the seat can actually tip forward and backwards that amount. I'm confident it will as the axis bearing shaft protrudes quite a bit. Those shafts are 25mm in diameter, strong enough I hope.

Great looking project there. :thbup:

Thanks bsft! Now if I only could someday pull it all together...