I made a case for my Raspberry Pi 3B+ in this design to protect a little more. I used https://www.thingiverse.com/thing:4054637 which is for the RPi4 as guide to make one for the 3B+.
I changed the Carriage Idler mount & the carriage belt clamps to the new versions, but belt was not long enough to add the new motor end mount. I went from 375mm maximum width to 392mm maximum width (2040x500mm extrusions) with this change & when I change to the motor end mount should get another 12-17mm width. With these changes, there will be 9 less T-Nuts used. The only T-Nuts that will be used are for the electronics brackets & Z-axis Pen Mount (10-12 T-Nuts). The T-nuts in the design did make it easier to take the Wheel endplates off & on easier while adjusting the design. I have enough belt to make 1meter version, so will add the motor mount on that design. I should have over 35" of travel with 1 meter lengths. Here are a couple of photos.
Since I will have a few extra O-Rings after getting a pack of 50 from Grainger, I decided to look at using 4 O-Rings on each wheel with the belt still between the O-Rings. It moved the belt a little further from the end plate & the 20T drive pulley was extended out further that I liked. I pondered it some more & decided to try moving the belt path to the inside of the wheel. Since the teeth are now on the outer edge, I am not tied to making them 88T. After some trial & error on different teeth count, 76T looked like it would work well with the closed loop belt I have. Having the O-Rings on the outside of the belt path also allows me to change the O-rings without taking the wheels off the plate. I spread the bearings apart about a 1/4", so I will need 2 1/4" bolts instead of 2". 2 1/4" bolts are not always available locally, but could use 2 1/2" bolts with some extra washers.
Has anyone ever tried 8mm Shoulder Bolts with the 608-2RS bearings. The ones I need are $1.72 versus $0.32 for the M8 Hex bolt, but I only need 4 of them. It doesn’t feel like I am getting any play in the wheels with the 5/16" bolts, but just wondering if anyone has experience with these. https://www.boltdepot.com/Product-Details.aspx?product=14243
Not shoulder bolts as such, but the MPCNC is a mix of longer bolts (with a smooth section) and short machine screws (threaded all the way) and it doesn’t seem to make a lot of difference whether the bearing rests on a smooth or threaded bolt. It might if the load forcing the bearing onto the threads was enough to deform the threads, but it’s all pretty lightly loaded.
Thanks for the info. I had already leaned towards buying the partial threaded bolts just so the bearings wouldn’t have as many threads to rest against. I might buy one shoulder bolt just for reference when I place an order. The fact that they are M6 threads on the end of M8 shoulder bolts, I would probably have to make a plastic washer on that end to rest on the bearing as the M6 washers might not have a big enough OD & fender washers are probably too big an OD.
I’ve been passively following Dave’s rolling plotter project, so now it seemed a good time to fire up the Prusa and at least print the wheel plates to play with. I’ve already ordered the 444mm belts Dave specified but they’re stuck on the slow boat and I don’t yet have them to play with. I also don’t have, or print with, anything but PLA so can’t print belts as Dave did. I did, however, once purchase some 60mm skate wheels from Ryan for my FoamRipper project… and a set of extras for a possible LR2 build. So I’ve started piddling with what I have on hand…
I downloaded one of Dave’s wheels from Thiniverse and “borrowed” the GT2 pulley section using TinkerCad… in hopes the 444mm belts (when they get here) might actually have a chance of fitting. I then combined the pulley part with a passive hub for one side of the skate wheel and another simple hub for the other side… these sit where the wheel bearings normally go but now just “lock” the wheel and pulley when tensioned by the axle bolt/nut.
I know these wheels won’t have a lot of traction but they are pretty “grippy” (maybe equivalent to an O-ring?) and may be enough for light loads; i.e. pen/marker, stylus, needle cutter, and possibly a laser. Right now, however, I’m really more interested to see how straight/true these wheels run… they run quite nicely on the FoamRipper. I’m sure a guide/curb of some sort will probably be needed for actual use… sort of like a track saw. At the very least, I should now have something to roll around on the floor and start playing with. It’s not much… but it’s a start.
Glad to see you are making some progress with it. If those wheels have some grip to the, they should work as well as the O-Rings. That looks like a good way to use those wheels & nice to see another option. Think I have the wheels & plate the way I want them for the version 3 of this design. I put in an order from boltdepot.com today & added some metric screws missing from my inventory to pad out the order. I decided to go ahead & get the M8 hex bolts for the wheels so it will be all metric now except for the O-Rings. The M8 locknuts from McMaster catalog inserts in fusion 360 are a little thicker than I thought, so I ordered M8x60mm bolts. I will wait until they arrive before looking at fine tuning the thickness of the wheel. My calcs for 1st assembly of the wheels & plate of this version were a little off, so I made a drilling template to redrill those idler slot holes a little more rather than reprinting that 4 hour plate. My new tooth count on these wheels is 78T. The print slope coming off the 76T was still a little ragged since it was less than 45degrees. I also put a lip on the inside of the teeth to make sure the belt didn’t come off. Another advantage of this new design is the belt doesn’t ride on the floor at all. Here is what the design looks like now.
I have about 6.5mm of motor shaft extended beyond the 20T driver pulley, so I am thinking a cool looking extruder visualizer might be cool to add to it. Here is one Jeff will like. https://www.thingiverse.com/thing:3135172
Since I was on a flower of life theme, I had to make a flower of life visual. Think I have some yellow filament to print this with. I was originally going to use a magnet to hold in place, but then realized the motor shaft wouldn’t stick to it. I should be able to just press fit it on there with that slot.
Here is my latest video. I am using all the new parts except the Carriage motor mount and I couldn’t use that because my belts were not quite long enough. After getting the 1st 2 wheels with 78T printed, I couldn’t get a descent print of them without webbing in the teeth, so I used a 76T wheel on each side to which shouldn’t make a difference in the printing. Looks like I need a little rougher surface for the wheels to roll on. I have some left over 5"x6’ bamboo flooring which feels like it might have just the amount of roughness on the backside for the wheels to roll on. I will test that next. After the next test, I will put the 1 meter width version together. I added a motor visualizer to the wheel motors & since I was drawing flower of life designs, it made only sense to make it look like that. I think it does make it look cooler while it is rolling.
On a much slower boat than normal… my 444mm belts were finally deemed “lost at sea” by the shipper. While waiting on the belts, I had pretty much piled everything in a corner and gone on to other things… and it didn’t look as though a rolling plotter was in my future anytime soon. Then a few days ago, mysteriously, a small package arrived in the mailbox… and, bored as I was, I was looking for something new to play with.
So I started playing with the endplates I had already printed… and started modding them to take advantage of parts in my stash. Wth zillions of 608 bearings on hand, I elongated one idler bearing slot to allow a single 608 idler to replace the two-bearing idler setup of the original endplate.
With the belt routing looking reasonable for now, I then took a look at the extrusion to endplate connection. I have a couple of 1000mm 2040 extrusions that I’m reluctant to cut to length until I know more about how I might use this thing… and decided I’d really like to try setting it to the same width as my FoamRipper gantry and work surface. So I reworked the extrusion “shelf” to allow the extrusions to “pass through”, for width adjustment and even endplate reversal (wheels inside or outside) if necessary.
So, now I’m printing a second endplate to match and can begin to think about getting this thing wired up and moving. I’m hoping to honor Dave’s mechanical design for the remainder of this build as much as possible… until I know more about how I might use it.
As my eventual “use case” differs from Dave’s, I’ll start another thread as this thing starts taking shape and moving. I know Dave has designed and demo-ed the rolling plotter to run on the floor (or any flat surface) plotting on paper and drawing neat patterns in a sandbox… but there is so much similarity to my FoamRipper’s gantry – driven skate wheels vs. free-wheeling and LR-style belts – that I suspect I’ll be running it on the interior door work surface of my FoamRipper, as a new extrusion-based gantry.
Looks cool. You should be able to run them both on the same table that way. Put a pen on one and a needle cutter on the other, and then synchronize them to make planes with penguins on them
Glad you are making progress. My biggest problem with my design so far seems to be the wheels binding occasionally, but do not see when it happens. Initially I had really good results, so I am thinking those wheels just need to be tightened just the right amount. I tried some newer bearings, but not sure they were any better than what I had. Lately I have been focusing on getting better prints from my 3d printer. I am getting a little bit of cob web in the GT2 teeth on the wheel & adjust heat temperature or retraction has not helped. I just finished swapping my electronics board on that printer from the Bigtreetech skr 1.3 with TMC2208 & LV8729 on E0 & Z to the Bigtreetech skr 1.4 with all TMC2209 in Uart mode & not sensorless homing. I have the hotend fan plugged into E1 & Bigtreetech changed the +/- orientation of that plug between the 1.3 & 1.4 board for some reason & I missed that. Fan still seems to work, so hopefully I did not hurt that fan. I can now try linear advance with this board & maybe that will make a difference. TMC2208 drivers seemed to not work properly with linear advance.
I had the same problem initially but printed a spacer to go between the twin 608 bearings’ inner races in the round section where the bearings seat. The spacer allows tightening the axle without pulling the inner races toward each other… and thus binding the bearings up. I forget the exact thickness but will check if you determine that’s your binding problem as well.
If you scroll back to message 114 there is a section view showing, my wheel has a center portion that touches the outer walls of the 2 bearings, but keeps the inner races free. I still have my previous wheel design on my printer, so these newer ones might work better. I was just hoping to get my printer fine tuned a little more before printing the new versions. If I get motivated again, I would like to revisit the idea of the hamster wheel drive.
I think this is the problem. If you take thumb and forefinger and press heavily on the inner race on each side of the wheel, you will be forcing the inner races toward the center and eventually bind the bearings. The spacer I’m talking about sits between the bearings and only touches the inner races… and keeps them from being forced out of alignment with the outer race when the nut is tightened fully.
I appreciate you pointing that out. Holding 2 fingers on there was not enough force, but when putting a bolt on there & tightening that does appear to be the problem. I had seen those spacers that Ryan sells for the lowrider & wondered if I needed something like that, but never pursued that thought stream. Guess I will make a 3d printed spacer to see if that fixes the problem.
You’re welcome. The length of the bearing spacer has to be just right, filling the space between the bearings when fully seated in their pockets… not too long (keeps the bearings from seating fully) and not too short (same as no spacer).
The other endplate finished printing so I quickly assembled it and slipped it onto the rails and set it roughly in place on the FoamRipper, where I see it eventually going.
