Showroom Belt driven linear actuator

I'm using 16 teeth 3mm pitch pulleys everywhere including the motor; therefore 16* 3mm = 48mm circumference and a 15.3mm diameter.

When I'm happier with the overall design I'll probably create a new post and pull together more of the specifications I based things on but largely I relied on this design guide from Gates along with the more limited specs given by this Aliexpress vendor (really they only provide breaking and working tensions).

https://www.gates.com/content/dam/g...catalogs/light-power-and-precision-manual.pdf

https://www.aliexpress.com/item/326...o.store_pc_groupList.8148356.4.368551ce46qlvk

With a 15,3 mm pulley you get around 30kg force with 2nm peak force

All the belt pitch lines are parallel so I have as many teeth engaged as is possible; in most applications eg skateboards the pulleys don't match and the smaller pulley has fewer teeth engaged.

On page 19 of the Gates guide they give a table for Torque vs RPM. For the 15mm belt we have a multiplication factor of 3.03 and at 3200 RPM the table shows 0.83NM giving a torque limit with our multiplication factor of 2.5NM. At lower speeds higher toques would be acceptable.

https://www.gates.com/content/dam/g...catalogs/light-power-and-precision-manual.pdf

Yes approximately and then multiplied by 4 with the pulley system.

Edit: Also FYI the working tension on the 15mm belt I linked earlier from AliExpress is only a fraction over 30Kg and that is one of the obstacles I was looking to overcome through the 4:1 pulley system.

no

there is this way no 4to1 pulley system this way, its does nothing its no rope system where you pull 2 meter rope and it moves 1 meter and becomes 2 times stronger

Could you explain better?
For a 1kg mass connected to a 1:1 pulley system you can pull a rope for 1m with a force of 9.8N and the mass will move 1m. With a 2:1 pulley system you now only need to apply 4.9N to lift the mass but pulling 1m of rope will only raise the mass by 0.5m.

The back and forth system I'm using is similar to one used on yachts for controlling the main sail. I forget exactly what it is called there but as you get more powerful sails you can no longer handle the main sheet directly so they have a rope version of my system for magnifying the force you can apply via winch.

Edit: This is a main sheet traveler which functions the same as my design but with a 2:1 ratio

I did take a ook how you want to fit the belt again. This could work this way youre richt

I think I did reduce my pulley ratio by one when I rerouted the belt, being 3:1 would explain why the teethed pulley locations needed to flip from side to side. That will reduce the maximum actuator force and increase the speed everything else being equal. The blue side of the belt would have teeth but trying to do them accurately kills my computer.



Edit: This was bugging me so I kept thinking about this and moving the split to the centre created two parallel pulley systems, effectively like putting two ropes instead of one through a single pulley.

Ah crap I have been miscalculating my pulley ratios from the first drawing and I have drawn it too high. Gadget, is that what you were referring to earlier when asking about all the belt?

Good news in a way, I suppose, as far less belt will be required but I wish I'd not just worked out the visual cases for 1:1 & 1:2 systems and then extrapolated incorrectly. I'm going to give it a break for a while and come back to this.

I know you get extra losses on small sprockets with chain and it is something serious cyclists might think about. I think belt is much more efficient in these situations but there are definitely a lot of tight bends. The belt and pulleys are both very light and probably contribute less than the motors rotor inertia. However these are the only parts moving at high speed all other moving components receive the mechanical advantage of the pulley system.

On a related topic; the idler pulleys come with a pair of bearings installed and are available for a variety of bore sizes eg 3, 4, 5 or 6mm. I wasn't sure which would allow for the strongest bearings, I think the bearing OD stayed the same for all the options offered so maybe it didn't make a difference? I've been meaning to look into it or ask for advice and had just gone down the middle for now.

I do need fewer pulleys than I'd been drawing for the ratios I wanted. Looking at the sketches below a 3:1 ratio system with 5mm pitch belt would fit much better than I had originally assumed. The 5mm belt had much higher tension ratings so even at 3:1 ratio it will ultimately handle higher actuator loads. With this arrangement the motor could no longer be mounted directly to a pulley but connecting with a small belt gives you an opportunity to regain some gearing from the 3:1 ratio change and slightly larger diameter drive pulley; possibly 10-14 teeth at 5mm each.

To be honest i like this idea. You can also use a fuw bearings for a roller on not toothed belt side.
I am only convinced you will get better motion with a ballscrew for same money.
You can feel everyting with a ballscrew with a belt probably not ,its flex a little and will bounce a little with direction chance, a ballscrew is more direct it only makes more noice

I was wondering whether some configurations ever could be redrawn to make things more symmetrical like I tried previously but now realize it impacts the motor gearing significantly. In the very simple sketches it's easy to visualize the pulleys will be spinning slowest near the fixed belt end and fastest on the left where the belt crosses between the lower and upper pulley systems. My previous model was complicated enough I'm struggling to visualize the relative speeds each pulley will be spinning as belt crosses back and forth in 3 locations.

The smallest 5MGT GT3 pulley that gets a specification in the Gates design guide (pg59) is 18 teeth and then 15mm wide belt has a rated working tension of 138lbs in this combination. I need to get further in the design to have a better idea regarding the static tension but using a 3:1 pulley system the peak actuator force this belt could handle approaches 400lbs. Asides from any static value additional tension only ever comes from the motor so is inherently a very controllable parameter.

Using the 18 tooth pulley each revolution provides 18*5mm = 90mm of belt travel and 30mm of actuator travel. To avoid needing incredibly accurate position control of the BLDC motor I think this should be much lower and the value previously was supposed to be 12mm. Selecting and fitting appropriate motor pulleys to return this to a similar value should be possible.

Regarding belt stretch; the peak forces I'm targeting are slightly above the values that give 0.1% belt elongation. The new belt path looks like it will be about 3m long so at the highest peak loads the belt might be stretching by 3000mm*0.15% = 4.5mm. That will be distributed throughout the belt length and because every pulley except the drive one is idle there don't seem to be any potential issues with pulley synchronization and tooth misalignment.

Edit: This doesn't seem completely true. Only one side of the pulley system will have higher than any static tension at a time depending on whether the actuator is extending or retracting. The tensioned length will vary depending on actuator position but on average it will be approximately half the total length.

First a happy new year
I like the idea, i am only convinced you can get better motion for same money.
But youre on a DIY site and can build wat you want, go youre way!
A bigger drive pulley and lower belt 3/1 ratio needs a bigger motor to get the forces you want.
regards Ad

Happy new year back to you. Currently I'm just trying to balance all those factors out; if I want to keep the motor mounted very close to the extrusion the centre distance between the two pulleys will not be very large and is likely to restrict my available pulley combinations. I've just been mulling options over and I'm about to start drawing again.