Thinking of motor driven wheels. Any input or thoughts?

Was thinking about switching from the MPCNC to the LowRider. I have a table already that is 5’-10’ that would be awesome for full sheet cutting. I use that table for various other glue up and assembly though as well.

Has anyone put anytime into thinking about gearing the wheel shaft and having the LowRider motors drive the wheels on the bed, instead of using g2 belts to pull the gantry? Any thoughts on the downsides or possible issues?

The big advantage would be the ability to have a self contained CNC gantry that could be removed from table when not in use, without having to worry about undoing the belts when you take it off, and having to put them back together when you want to put it back. This would also allow you to easily slide the width if you wanted to use it on another table instead?

I did a search for belt driven wheels to see if anyone else has talked about it but didn’t see anything related to this idea.

 

Thanks!

1 Like

I haven’t seen that mentioned before. The trouble (I would guess) is that there would be too much error in wheels. You can be pretty sure if you’re 0.31mm past the 257th tooth on the belt, that you know where you are in X. I don’t think the wheels could come close in terms of accuracy.

Some quick release type of belt connector would be a good start. Or even just some kind of protector over the belts so you could use the surface without removing them.

1 Like

Yeah, I can see that issue. I was thinking it would still be “belt” driven actually, but the belt would just be between the motor and a gear fixed to wheel shaft. I can see now that if it was to have a hiccup, like a crash that would normally cause a skip in the motor on the long belts, the wheel would probably just keep going and end up making the gantry out of square. Hummmm… going to keep thinking, also about that idea for quick disconnect belts.

Thanks for your thoughts :smiley:

Perhaps look more into locking the mechanism out of the way and leave the belts in place while the rest of the table is in it’s flex time.

1 Like

I love the idea! There is a CNC unit that is on the market for $5,000 +/- depending on options. This unit is also adjustable in width. The max size of the work area is 4’ x 8’. It has powered wheels. Just set the width, sit it on the workpiece and run it. I’d have to look back through my Wood Magazines to find the name and web address. When I first saw the Lowrider I thought it was wheel driven, until I looked at it in more detail and saw videos.

Still the idea of modifying the lowrider to wheel driven is something I want to look into more. Here are some issues that need to be addressed.
1 - traction. Gravity mixed with a dusty environment = slippage. Solution is to design the driving wheels to drive from under the table edges.

2 - gravity. With the unit being so lite weight, the wheels will slip on smooth surfaces. - Solution is the same as above.

3 - Tracking! This is a biggie… regardless of the drive units used, the initial alignment would be critical. Additionally, the “Drive” wheels would have to be EXACTLY the same size, else the alignment will f up more and more as it moves along the Y axis. - Solution would be machined metal drive wheels, machined with some sort of tread to improve traction.

The unit would have to be “mounted” to the table with the bottom drive and coaster wheels in a quick release clamp with adjustable pressure pads for the drive wheels.

I have already started playing around with the drive unit design ideas. As I think it is doable. I personally like the idea of unclamping the unit from the table and placing it on a shelf, thus giving me the option of using the table as needed in my woodworking shop, or even folding up the table for storage.

BUT… it is also an idea that is on my “Hmmm that would be nice to do” list, as I haven’t even finished building my first Lowrider.

I’ll look through my Wood Magazines and find the name of that CNC machine and post it here, it might give you some ideas on how to do it.

 

Maybe the Crawlbot? https://diy3dprinting.blogspot.com/2015/10/printrbot-crawlbot-cnc-set-up-and.html . It is dead, unfortunately, and it also used belts.

There are a lot of creative packing solutions for the LR already. You can just remove the belts and move it to the side, or even pull it off completely. Barry uses his as outfeed for his table saw and the belts stay on, it’s just not in the way enough to matter.

1 Like

In reference to the Crawlbot " It is dead, unfortunately, and it also used belts.

  1. It might be the one, but I thought the one I was thinking of had an LCD Display. Hmmmm It was in an issue of Wood Magazine... a... several years back, (maybe 5 or 6 years if not more). I also thought I remembered it not having a belt, or chain for the Y-axis. But I could be wrong.
  2. As far as storage goes I've decided that the best thing to do is find a nice large room for the Lowrider... So that means that I will be kicking my two grandsons (ages 10 & 13) out of the large upstairs bedroom; into a tent in the backyard. Then if they survive the harsh, freezing northwest Ohio winter, come spring I'll build on to the garage making much needed room for all of my woodworking toys. However since I live with my Daughter and her two sons, if anyone moves out to the tent it would be me. - Damn brat!
Actually I was thinking of using my table saw outfeed table as well, it's 5' x 7', but since I also use it as an assembly table, sanding table, and everything else table, I would not be able to leave the belts attached.

My plans are…

1 - Get it up and running and mounting it to a remake of my other assembly table which will be cut down, given a face lift and sized to support 3ft x 5ft CNC work. Yes 4ft x 8ft would be nice, but only in two real situations, cutting a lot of parts out of a full sheet of plywood, or making some big ass signs. At 69, handling 3x4x3/4 sheets of cabinet grade plywood doesn’t happen without some major help. And the only big sign I want to make is the “FOR SALE” sign to go on this house when we get ready to move into something more larger… meaning with larger shop area.

2 - After it is up and running, then I am sure I will be looking at the Lowrider under a microscope, and from every angle for ways to modify it and maybe even make it better too.

I must confess that this is my second entry into small CNC, the first time was in 1983, when I was part of a team at the college I was teaching at… yeah, there was nothing to work with them… except and idea and six Apple IIe computers. 3 years from start and we did have a table size unit operational, complete with robotic arms to load and blanks from a hopper, robotically lock the blank in place, then take the “name” (up to 12 characters) from the list typed into the Apple IIe send the commands (since there wasn’t any usable code) to the CNC spindle (Dremel) and watch it carve a name into the wood blank. Then robotically vacuum the sign, robotically release the clamps, and then the final robotic arm would remove the sign and hand it to you. Once you grasped the sign and tugged the robotic arm released the sign, triggering the computer to play something that resembled music, and in a robotic voice say “Thank you.”

While that sounds ridiculous in this day and age, with “Plug and Play” components, gcode, and everything else; back then it was leading edge research and development. Also while it sounds like and engraving machine, it wasn’t. We could carve with it, as the machine featured X, Y, & Z axis, it also featured a robotic knuckle as the spindle mount, this giving us a 4th axis that was custom programed as needed. The other thing was it was true 3d carving, not 2.5d.

Here’s what it took to make something…

It had to be drafted in 2d (since no 3d CAD hadn’t been invented yet), with all views and all dimensions. Curves, arcs, etc. had to all be plotted. Then we took all of that and started calculating each point on an X, Y, Z matrix. We actually called them pixels, like in printing… since the robotics (stepper motors) were addressable through control cards designed and built by the school, we told each motor how many steps to take and in what direction. So moving from X5, Y10, Z10, to X8, Y15, Z12, and then to X5, Y15, Z11 was done without having to wait for the next layer. With that said, I’m not saying it was a smooth move or even a pretty move, but it could be done. There wasn’t any “models” to work from. OK… the old man is ranting…

Needless to say that today’s digital electronics and the world of Makers have made it a hell of a lot easier to turn ideas into reality. And I’m not even going into the fact that my first computer a “Micro” (since the word “Personal Computer” hadn’t been invented yet) was built from scratch.

I’ve been a woodworker off and on most of my life, helping, learning, then on my own as a hobby, and at the age of 69 pushing 70 I decided that add a 3d printer to my shop. Not to print all the crap, but as a tool, printing brackets, and various things to use and help in the shop, now I’ve decided to add cnc back into my life, this time I hope without all of the headaches, again as a tool and not a toy. I see using the Lowrider to do a lot that I find it hard to do myself, like cutting various joints, and doing things better than I can, like the repeatability of cutting out all the parts to a build.

I’ll be sure to keep you (Jeffed3) and forum members using a different subject and not piggybacking on Joshua’s thoughts and ideas.

In the meantime, I’ll also go looking through my old Wood Magazines (I never toss them out) and see if that CNC unit I was thinking about was the Crawlbot or not. And if Crawlbot is out of business, I’m sorry to see them go, and wonder why (other than price) that it happened.

Enjoy… ok… a… now which darn part do I print next?

In reference to the Crawlbot " It is dead, unfortunately, and it also used belts.

  1. It might be the one, but I thought the one I was thinking of had an LCD Display. Hmmmm It was in an issue of Wood Magazine... a... several years back, (maybe 5 or 6 years if not more). I also thought I remembered it not having a belt, or chain for the Y-axis. But I could be wrong.
  2. As far as storage goes I've decided that the best thing to do is find a nice large room for the Lowrider... So that means that I will be kicking my two grandsons (ages 10 & 13) out of the large upstairs bedroom; into a tent in the backyard. Then if they survive the harsh, freezing northwest Ohio winter, come spring I'll build on to the garage making much needed room for all of my woodworking toys. However since I live with my Daughter and her two sons, if anyone moves out to the tent it would be me. - Damn brat!
Actually I was thinking of using my table saw outfeed table as well, it's 5' x 7', but since I also use it as an assembly table, sanding table, and everything else table, I would not be able to leave the belts attached.

My plans are…

1 - Get it up and running and mounting it to a remake of my other assembly table which will be cut down, given a face lift and sized to support 3ft x 5ft CNC work. Yes 4ft x 8ft would be nice, but only in two real situations, cutting a lot of parts out of a full sheet of plywood, or making some big ass signs. At 69, handling 3x4x3/4 sheets of cabinet grade plywood doesn’t happen without some major help. And the only big sign I want to make is the “FOR SALE” sign to go on this house when we get ready to move into something more larger… meaning with larger shop area.

2 - After it is up and running, then I am sure I will be looking at the Lowrider under a microscope, and from every angle for ways to modify it and maybe even make it better too.

I must confess that this is my second entry into small CNC, the first time was in 1983, when I was part of a team at the college I was teaching at… yeah, there was nothing to work with them… except and idea and six Apple IIe computers. 3 years from start and we did have a table size unit operational, complete with robotic arms to load and blanks from a hopper, robotically lock the blank in place, then take the “name” (up to 12 characters) from the list typed into the Apple IIe send the commands (since there wasn’t any usable code) to the CNC spindle (Dremel) and watch it carve a name into the wood blank. Then robotically vacuum the sign, robotically release the clamps, and then the final robotic arm would remove the sign and hand it to you. Once you grasped the sign and tugged the robotic arm released the sign, triggering the computer to play something that resembled music, and in a robotic voice say “Thank you.”

While that sounds ridiculous in this day and age, with “Plug and Play” components, gcode, and everything else; back then it was leading edge research and development. Also while it sounds like and engraving machine, it wasn’t. We could carve with it, as the machine featured X, Y, & Z axis, it also featured a robotic knuckle as the spindle mount, this giving us a 4th axis that was custom programed as needed. The other thing was it was true 3d carving, not 2.5d.

Here’s what it took to make something…

It had to be drafted in 2d (since no 3d CAD hadn’t been invented yet), with all views and all dimensions. Curves, arcs, etc. had to all be plotted. Then we took all of that and started calculating each point on an X, Y, Z matrix. We actually called them pixels, like in printing… since the robotics (stepper motors) were addressable through control cards designed and built by the school, we told each motor how many steps to take and in what direction. So moving from X5, Y10, Z10, to X8, Y15, Z12, and then to X5, Y15, Z11 was done without having to wait for the next layer. With that said, I’m not saying it was a smooth move or even a pretty move, but it could be done. There wasn’t any “models” to work from. OK… the old man is ranting…

Needless to say that today’s digital electronics and the world of Makers have made it a hell of a lot easier to turn ideas into reality. And I’m not even going into the fact that my first computer a “Micro” (since the word “Personal Computer” hadn’t been invented yet) was built from scratch.

I’ve been a woodworker off and on most of my life, helping, learning, then on my own as a hobby, and at the age of 69 pushing 70 I decided that add a 3d printer to my shop. Not to print all the crap, but as a tool, printing brackets, and various things to use and help in the shop, now I’ve decided to add cnc back into my life, this time I hope without all of the headaches, again as a tool and not a toy. I see using the Lowrider to do a lot that I find it hard to do myself, like cutting various joints, and doing things better than I can, like the repeatability of cutting out all the parts to a build.

I’ll be sure to keep you (Jeffed3) and forum members using a different subject and not piggybacking on Joshua’s thoughts and ideas.

In the meantime, I’ll also go looking through my old Wood Magazines (I never toss them out) and see if that CNC unit I was thinking about was the Crawlbot or not. And if Crawlbot is out of business, I’m sorry to see them go, and wonder why (other than price) that it happened.

Enjoy… ok… a… now which darn part do I print next?

Crawlbot was made by printrbot (and they were the inspiration for the LR). Their main business is what tanked, afaik.

We often hear about the improvements and openness of the hardware, but not much about the software. If the original patent for 3D fdm printing was originally open, it still would have taken a long time to gain traction because of the software needs. Thanks for that perspective. Writing gcode by hand would be a huge pain. Then also converting it to stepper movements before running it is unimaginable. Other than a proof of concept, there’s no way that would be worth it for hobbyists like me.

I would think the biggest issue with motorized wheels is lack of friction. Especially with the roller-blade wheels. The whole idea with the wheels is to provide as little friction as possible between the gantry and the table. That way all forces exerted from the belt being pulled gets applied to the bit cutting into the wood. Skaters also hate friction in the direction of the wheels and only like friction when pushing sideways on them.

If you got different wider wheels with a stickier compound, it might work. But I’m pretty sure roller-blade wheels would just spin on the table top as soon as the bit hit the material.

1 Like

If there are wheels on the top of the surface to move around on, at a specific height, and wheels under the surface to drive the thing around (likely spring loaded to provide a bit of pressure) I don’t see a lack of friction type problem.

1 Like

David Walling… Yes but mainly no.

Narrow wheels have more traction than wider wheels! What is important about the wheels would be the hardness of the tire. You would need a narrow wheel/tires that are soft (sticky) enough to grip, and made of a material that would not collect the fine saw dust.

It works like this… the weight of the full gantry is carried evenly on all 4 wheels, let’s say the gantry weighs 40 pounds (for easy math)… that places a 10 pound load on each tire. Now you have to look at the actual tire contact area. Let’s say that the roller blade tires have a… hmmm 1 mm of contact, where a wide skateboard wheel/tire has 10 mm of contact area. The roller blade wheel/tire is carrying all 10 pounds over 1 square mm while the wide skateboard wheel/tire is only carrying 1 pound per square mm. (Provided the wheels are of equal diameter). Another quick example would be walking out on thin ice on a pond… as you are walking it starts to crack. If you turn and run, you’ll break through for sure, as running makes for a smaller load area on you feet, and the added forces of your body in motion. While laying down on the ice and scooting to the edge will spread your weight out over a wider area, thus decreasing the weight per square inch on the ice. We need the drive wheels to break the ice. :slight_smile:

Look at a farm tractor… when it is used to plow (a job needing the most traction) the farmer uses only the the thin tires, but when other tasks are being done in the fields and the farmer doesn’t want the tires to dig into the soil, they use dual tires, or equipment with very wide tires (like those used on monster trucks) as they provide floatation.

The wheels used on the LR2 (at least the ones they sold me) have a heavy crown, with very little table contact, but they are also a softer material. The crown on the wheels are not to reduce friction as much as possible, but are to provide control (memic ice skate blades). The quality of the bearing and axle is what does the most to reduce friction. I hardly think that serious roller bladers especially the ones that compete use 50-cent bearings and an off the shelf bolt as an axle. In fact the bolt doesn’t even fit tight enough to make the wheels use the bearing. The bearings are so loose that they are only acting as a sleeve bushing. If I’m not mistaken, the bearings actually take an 8 mm axle, or bolt. The 5/16th bolts provided in the kits measure out to an average dia of 7.73 mm, with the average bearing sleeve measuring out to 7.88 mm. Also you’d need to design a new wheel hub that would allow the mounting of the drive gear/pulley/whatever.

My concern is that the roller blade wheels on the LR2 will load up with fine saw dust, especially Pine, and MDF as both of those materials produce a saw dust that is more “sticky” than let’s say cherry or white oak. In the pine it’s the sap that makes it sticky, and with the MDF it’s the electrical charge it gets when cut by high speed tools.

I was thinking that a track system might work best, as it could be printed on a 3d printer, and provide the traction, while also not collecting the saw dust, or add a small brush to clean the track as it it is laid down (like the small brush in a band saw to clean the lower wheel). The track could be printed out of a soft flexible material, soft PLA might work too. The tracks could be designed to simply wrap around the drive wheels; or, print drive cogs.

Your concern with the wheels losing traction when the bit starts the cut, or in the cut is valid.

I believe that the loss of traction would not be caused by the wheels, but by the lack of weight by the gantry. I believe that the absolute best solution would be to have a 3rd wheel that runs on the underside of the table edge rails. That would be the drive wheel, and could be adjusted to increase the traction by pulling/holding the gantry down.

Still, that is a project that for me at least will be down the road. Once I get my LR2 up and running I will better know and be able to gain information needed to then think about modifications or designing a wheel driven unit.

 

I totally agree.

Hmm gcode didn’t exist, and the first motors we used were not stepper motors, as the only stepper motors that were around were large and very,very, very expensive. We used DC motors, sending small timed impulses to start the step, when a reversed polarity pulse to instantly stop the motor to prevent movement. Much like a cheap inkjet printer does today. Our “gcode” matrix was a real 3d cube that was woven with red, blue, and green sewing thread. You counted and measured to get the coordinates. Once that was figured out, we wrote an algorithm to make generating the coordinates easier. And like I said, it took use a little more than 3 years to get the first system operational, as we have to invent everything from scratch with no existing models to work from

 

I’m being pedantic, but I don’t think that’s true on a hard surface.

The friction parallel to the surface is independent of the area. It doesn’t feel right but it is. I think about it as being proportional to the area and pressure, but pressure times area is weight. So a larger wheel would have more places it would touch, so higher friction. But each place has less pressure, so lower friction. Those balance out. This is the physics 1 understanding of friction.

When your tires are digging into the dirt, that’s a much more complicated model because it’s pushing back along the surface, and not using much friction at all. Like if you skip a rock on a pond, as soon as it digs in, it stops immediately.

Bigger wheels would average the friction on the whole area, so if you happened to catch a slick spot, the other 90% of the wheel would still have good friction. But on average, they have the same friction as smakler wheels.

With the small amount of weight compared to their intended use, the wheels don’t flex at all either.

2 Likes

Depending on the circumstance, your statement is true. But it doesn’t apply to all situations otherwise drag cars would all by running around on big skinny tires. You’re applying a torque across a surface. While the force from the weight of the vehicle would increase on the smaller patch, your torque being applied at that point would also increase. Skinny tires with high torque have a higher chance of slipping on a solid surface. Our machines are no where heavy enough to overcome the torque and lateral force the wheel would be pushing against. The force pushing against the side of the bit is going to be higher than what gravity can provide down through a skinny wheel. In this case, you need a wider contact patch (within reason) to apply more torque across the surface.

All the farm tractors around my house have multiple wheels on the back to increase the contact patch. The reason they’re skinny is so that the wheels can fit between the rows of product being grown. The other reason they’re skinny is to try to cut through the mud and muck to get to the more solid ground underneath it.

No. Rollerblade racers don’t use cheap bearings.

The idea of having a second set of drive wheels on the bottom that can be clamped tight to the surface intrigues me. It would help increase friction and help keep the wheels away from any dust landing on the surface.

Whether you go skinny rollerblade wheels or wider skateboard wheels, you would want softer compound for the drive wheels.

1 Like

OK… I agree with you both. I didn’t go far enough down that road of thought. Buying new thinking cap tonight at Walmart.

 

1 Like

Get me one while your at it!

You could pretty easily try this out with the first version of the lowrider just use a track under the table and a spring loaded gear to drive it. Or driver the current version on the table (with a wheel in some sort of track) (maybe add a little weight to increase friction). Truth is is would work, minor tweaks here and there but it would be a great idea, slightly more calibration to account for a few things but nothing crazy. It would get rid of the larger rack and pinion style backlash. In the end I just think it would require careful weight, tire, table considerations. How do you overcome the mechanical advantage of the 16T pulley vs 60mm wheel direct drive, can’t use a gearbox, way too much slop?

I would only attempt it with some sort of gear, but then that brings it right back to belt. Belt is cheap, easy, the easiest/cheapest to ship.

Wouldn’t you now also have to account for the size of the wheel when calculating distance traveled per revolution of the stepper?

Yeah but it would be really easy to just calibrate it in the real world as well.

Sorry, they didn’t have one in your size.

However while buying mine, I also bought some cookies, and they provided me with the answer to to how to overcome the problems of traction, gears, tracks, etc. I’ve decided to obtain two Kebbler Elves, and train them to move the carriage +/- along the Y-axis X increments based on the amount of electrical shock they receive to the electrodes attached to their butts. If an Elf can’t do it, then it can’t be done. Damn, those cookies were good.

Truth being… IMHO… that while direct drive sounds like a great idea (and still is), it would require a lot of modification and as Ryan inputted the problem itself of driving the wheels, which indeed would have to be exactly the same size. This is all well and good and us a good project to work on… at least for me at a later date. In the meantime, my solution is much like I saw in the Youtube video about the Crawlbot linked above by Jeffeb3. Which was to use detachable belts. Hence, I have designed a detachable - adjustable belt mounts. So I plan on using them. Taking that a little further, I think I might leave some extra length on both ends, that way I can use it on longer tables if I want. While making the width adjustable is simple, the hassle of bet adjustments for different table sizes is more than I want to deal with at this time. So I am freeing the two Kebbler Elves I kidnapped and going out to the shop to finish that assemble of my LR2 carriage.

Hope ya’all enjoy the weekend.

 

2 Likes