David Bunch published a new build:

Read more about this build...

 

I made some progress today. I did a test print of the basic assembly on the side plate with a 444mm length timing belt printed in plastic to do a test fit before ordering the belt. It fits quite well for the 1st test print, so think I have the length I need. I made the plate only 5mm thick, so the plate flexed quite bit. I will have to make it thicker & maybe add some reinforcement. Otherwise I just cut an aluminum plate for each side, but would rather have to do that. I made the test timing belt plastic print a little thicker than what the belt measures so it would print. I was surprised how flexible it was. I did use eSun PLA+, which has more flex than regular PLA. I am using 5/16" (7.9375mm) hex bolts for the wheel connections instead of 8mm since they are close enough & they have been well tested in the MPCNC machines.

I had an idea for a quick connect pen connection. Last year I had made a towel ring that clips on our stove & a towel goes thru the hole as you can see from this screen shot of the design.

That works quite well, so I decided to try that idea with a sharpie pen. It is held in there surprising tight & easy to clip in & out. I plan to use a SG90 servo as I have seen other plotter designs use. The electronics will probably be a MKS Gen v1.4 with DRV8825 drivers which I have laying around.

I will move on to the X-axis after I tweak the side plate reinforcing a little more. I will order the timing belt this week after looking to see what else I might need from China for this project. I am still waiting on a couple parts for the eggbot design that I ordered just over a month ago.

 

After pondering whether to make this plate a little bigger to accommodate both 60mm & 100mm wheels, I opted to leave the dimensions like they are since I would also have to move the motor up 5mm or so as well as making it 10-20mm wider. After looking at my dimensions again, I could put a 2060 across the center which I do have & mount the y-axis motors to them. Speaking of Y-axis. I could make either the short 1 meter width axis the X or Y depending on which way I want to draw. I made the long axis on my MPCNC the X-axis, but think I will keep my initial decision of having dual Y motors & 1 X motor. I did decide to use 2040's instead of 2020's as it does give more room to mount stuff & route the wires thru the middle from other side. I also have 5 of the 2040x1000mm length V-slot. I am kind of amazed how much stuff I already have on hand to use for this project. I am printing the 1st side plate today.

In case anyone is interested, I used this youtube video to help draw the gearing for the 88T GT2 PulleyWheel.

I did discover that I had to add a div variable to be able to use 2 variables in an equation. This was after a google search for the solution.

Here are what my dimensions for the side plate are. I tried to keep the key dimensions nice even numbers.

 

I got one side finished today & printed all the TPU tires. The other side should be finished tomorrow. I am using zyltech black TPU for the tires. The TPU played nice today & didn't give me any problems printing it with the titan extruder. I tapped the 8 - M5 holes in the 2040x1000mm V-slot using the method on my last build. The bearings in the wheels are not as tight a fit as they should be, so will have to redo those. Thought I had that tolerance correct after 3 small test prints, but apparently I did not. There does look like enough clearance for the belts so they do not touch the ground. I was considering making them a 2-4 teeth less & will probably look at that anyway. I ordered the 444mm length closed loop timing belt & the other parts I need for this build from China. I also decided to order the bigtreetech V1.4 board to possibly use with this machine. Probably overkill, but looks simpler to add a WiFi board to it. From the youtube video on adding it to the MKS board seemed like a few jumpers to add. I bought 2 of them, so may try it on that board. I started thinking today that just adding a direct gear drive to the wheels might be a better way to go than the timing belt. I do not know a lot about gearing, but looks like I could only drive one wheel like the other 2 similar printers on here are using. Looks like I would need a differential for this. I might play with it & see if it will work since it will probably take a month or so to get the timing belts. I added 8 ovals to the face of the wheels to be able to see them moving easier.

 

Here is a cool Gear Generator webpage with animation. Think I have seen this before, but forgot about it until google found it again for me.
Involute spur gear generator and simulator

 

This gearing might work. I would like to make it helical or double helical gears for a better connection. The drive gear is 13 teeth, the big wheels are 52 teeth(104mm Pitch Diameter) & gear between wheels is 22 teeth & looks like it will be a 4:1 ratio. The motor with this configuration would have to mounted off the inside of 2040 connection. Would have to mount the other gear someway off the side plate with a bearing in there somewhere. I will play with it some more to see if I can get a better configuration.
Involute spur gear generator and simulator

 

I see a couple of problems getting the X-axis belt drive to work now that it is partially together. The standard way of using this belt with this configuration seems to run it thru the middle of 2040. So I need to redesign the End plates for that. I printed a couple of 444mm timing belts in TPU rubber & was surprised how well they printed. There is a bit of stringing on them, but good for test fitting. Might be worth test driving those belts once I get it more together, especially since it will be a while before the real belts arrive.

The X-carriage hits the plastic under support for the 2040 before reaching the end, but when I redesign those plates it should give me a little more space. It is looking like I can get between 30-35" of drawing space with 1 meter extrusions. I temporarily placed the X-axis motor about where it will be to see how it looks. With this change I will only be able to use one screw on each end of that 2040. I will redo the plate to give support to the other 20mm & maybe mount a screw with a T-nut underneath. Looks like I have enough room for that. I was also looking at extension cables for servo motors & found a 1 Meter extension. Will probably order that rather than create one. I also used M8x45mm full threaded bolts for the wheels since that was the closest the local hardware store had. They didn't have 5/16"x1.75" bolts. The metric were more expensive, but at least all the hardware is now metric. I would rather have partially threaded M8x45mm as they would probably be a better fit going thru the bearings.

 

Glad I had not started reprinting these wheels yet. A friend just reminded about using O-rings & I was looking into that the other day but had forgotten about it. I have harbor freight assortment I had used on the wind-up toy vehicle awhile back & the #327 fit good. There is only 7 of them in the assortment so I need 1 more. I found them at Grainger in packs of 10 for $5.53 or 50 for $11.68. I might just go ahead & buy 50 in case I decide to make another one of these. Here is the link for 10 of them. Round #327 Very Hard Buna N O-Ring & here is photos showing the difference between using the O-rings & the TPU 3d printed tires. I will have to make the cutouts to better fit the o-ring tires now. I will make a test print 1st. It is amazing how all these little problems are just working themselves out so easily.

 

Thanks Mark.. I am learning a lot from this build. Fusion 360 makes it a lot easier to design than my previous workflow of Draftsight Pro for 2D & openscad for the final parts. Although there are still some aspects of the design I explore openscad with. While looking at changing to a direct drive, I found a good gear library with openscad that includes herringbone gears. I also just found a free app in fusion 360 for this, so will try that also.

After printing a new y-axis plate & new wheels, I found a couple of things to change. The wheels seem very tight, but loosening the bolts helps some. I think those wheels need a bearing on the plate also, so I redesigned that this morning to add the same bearings I am using in the wheels. Those 4 bolts now have to be 50mm for metric or 2" for English. I also added more of a cutout around the bottom middle idler tensioner slot to make it easier to tighten that nut without using needle nose pliers. While putting the parts together, I noticed that it is better to install the bottom 2040 T-nut screw before putting on the wheels, but the ballhead allen wrench also works when I didn't do this 1st. Here is what the plate now looks like. I might put a ledge between those bearing holders to middle & top to make it a stronger connection.

 

I adjusted the y-axis plate one more time. It took me 4 versions before getting it the way I want. It would have taken only 3 if I had put the bearing on the plate to the outside. The wheels can get pulled up against the plate & I didn't realize it until test fitting it. I am using 5/16"x2.5". It only needs to be 2", but got them by mistake when they were in the wrong bin at Ace. I kind of like the longer length bolts because the partial threaded portion of bolt goes against more of the bearing. I also bought a box of 100-5/16" washers which work well against the bearings if you use the finish side against them. There are 7 or 8 washers on each axle. I meant to test the 1/4" washers as they might be almost exact fit around those bolts. Next time I am in Ace I will look at that. I am just using regular nuts while test fitting.

Since I don't have long enough belt for the 1 meter width yet, I am going with 500mm. My 2040x500mm V-slot won't be here until Monday. While I am waiting for those, I tried some C-beam & just 2020's to see how it looked. I kind of like the C-Beam, but it does add quite a bit of weight. That might be useful if I want a longer Z-axis mount.

I found one more problem while assembling this. I have the X-axis motor mount about 2.5mm off center of V-slot & with the belt path it really needs to be centered. I am also using 16T timing pulleys, but might switch to 20T since they are a little bit wider & might work better with the X-axis Belt. I just finished printing the new X-motor mount & will see if that is sufficient to solve that problem. If I switch to 20T on the X-axis, I will also switch it to 20T on they Y-axis to make it simpler. I have a 30% off Harbor Freight so will get another o-ring assortment tomorrow so I will have the one extra o-ring I need. I will use some of those O-rings for another windup toy vehicle.

 

When I woke in the night, I was thinking of how I could easily get the pen up/pen down movement using a servo with gearing. Then got on the computer & searched for rack & pinion movement with servo & low & behold I found this design that looks perfect for that movement. He is using a continuous rotation servos like the FS90R, but for the short movement I have the MG90S seems like it should work.
Save A Few Steps On Your Next Build With These Easy Linear Actuators

I have a couple of MG90S servos which only have 180 degrees of rotation. I will see if I can get that to work before ordering a FS90R. Here are a couple of web pages I found useful for servos.

MG90S Micro Servo Motor Datasheet, Wiring Diagram & Features
TowerPro MG90 Servo Specifications and Reviews

This might be just making the movement I need more complicated & the movement I am going to use on the eggbot is probably all I need. I will try this 1st before playing with the rack & pinion since I have it about setup to test. Maybe a manual rack & pinion movement could be used to set the initial height of the pen with a set screw locking in the position. Maybe a design that uses part of this idea of the toy grabber, taking out the large outside gears with claw. Predator Action Pliers Mk2 by ecoiras

Think I will explore making a rack & pinion mount for this with a pivot point for the pen at the mount so the pen has room to give when drawing, but set at 90 degrees for the end of the pivot. I also have a100 N52 10x2mm disc magnets to use as a quick disconnect. That will make it super easy to make different mounts. This is going to be fun.

 

I worked on the setting up the Marlin 2.0.4.4 firmware to run with the MKS Gen V1.4 board & was able to get it to compile. While working on the firmware I realized I need a Y endstop also. I will have it hit a table leg or other solid object. I received my openbuilds order yesterday. I wasn't expecting that until Monday, yeah. I ordered some short 250mm extrusions to play with & decided to put a 2060x250mm extrusion in the center of this build. Will probably mount the 12v P/S there until I figure out a more portable way to power this. I tapped all the holes in the 2060, but only using 2 of the holes for now. I figure that extra hole might be a good mounting point for something else later, maybe a picam. I also need to change the orientation of the raspberry pi mount. I used a simple mount I found on thingiverse instead of designing one of those. I plan to have wireless on the board eventually when I switch to the SKR 32bit board, so that should only be there short term. Looks like it will be tomorrow or Tuesday before I can run the 1st motor tests. Here is what the current design looks like.

 

I hadn't really thought about there not being enough teeth to engage it. I put the idler on the bottom to keep the belt from dragging on the floor. I will keep the speed down & watch for a problem there. Thanks for pointing that out. Would a 30 or 36 tooth drive gear work better than a 16T? Guess I loose torque with more teeth, but don't think it is a concern with this application. I probably won't be able to really test for a problem there until my closed loop belt comes in. I did spend a little time working on a direct gear drive which would have 4 gears if I did it right. One drive gear on the motor, a gear on each wheel & a gear in between the wheels. I made them double herring bone gears for better contact. I haven't done a test print on any of them yet. Those will probably be long prints even after cutting out some material.

Thought I was going to be able to test this today, but had to resolve some of the final touches. Should be able to test it Monday or Tuesday. I am starting to think about changing the short axis to the Y & the long axis to the X. I should only have to change the limit switch mounts if I do that. I will leave that alone until I test the movement.

 

For some reason I was thinking that if I drove both wheels with the motor gear they would be moving in opposite directions. After thinking about it again, not sure why I thought that. That should be fairly simple to do with just 3 gears. Here is an interesting web page on planetary gears that gives a good visualization of that movement when you set the number of planets to 2. This page is more fun than a fidget spinner, especially moving the sun or ring speed from negative to positive numbers.
Planetary Gear Simulator

 

I am definitely going to follow thru with this setup before seriously think of switching to gears. Putting idlers on both sides at the top will be a tight fit, but looks like it would fit especially if I use smaller diameter idler pulleys. I just didn't want to reprint those plates again since they are 2.5 - 3 hour prints each. I did go ahead & order 5-20T timing pulleys from Amazon since I don't have any & looks like x-axis will give more clearance with that slightly larger diameter. I am printing the P/S mount now & then will be ready to start wiring it up. It should be only a temporary mount until I find a better way to run 12v or 24v to it. I don't want to fry something right off the bat, so want the PS in a sturdy spot. I saw one of the other rolling machines on here using a 18v battery that are used on cordless power tools which looks like the best idea.

 

I have everything hooked up except for power. Will hook that up tomorrow, check all my connections & see if it works. The X-axis might catch in the V-Slot. I will swap out the 16T timing pulley with a 20T when it comes in on Wednesday. I am going to see if I can move the path of that belt to the top of V-slot. There is a belt slot at the top of that plate that will work for it.I had a jam in hotend today, so I had to fix that before I could finished my last couple of parts. I found this interesting instructable on making an adjustable P/S from a Ryobi 18V battery. It might add up to $100 for all the parts. I have been thinking of buying another cordless drill as my 12v B&D needs some help sometimes. Maybe I will get one by them. Battery Adjustable Power Supply - Ryobi 18V

Here is my progress from today. I flashed the firmware with Marlin 2.0.4.4 setting the X bed to 375 & the Y to 1000. I just realized the locknuts on the back of X carriage will just barely hit the P/S mounts, so will trim them enough to miss it.

 

I got a little side tracked again yesterday with thermistor problem on 3d printer E3D V6. I replaced it once, but can't see a problem. The only thing I can think it could be is the quick disconnect. I am thinking about just soldering that & get rid of that extra connection. Anyway since I redid the belt path for the X-axis, I also changed the limit switches to have the Y-axis the short length. I am not completely happy with the belt path as it is not completely in line with the pulleys at each end, but doesn't catch like it did using the other way. I might have to wait to see how the Pen mount plays out before changing it further. I also designed a simple Raspberry pi mount using the other as a guide and made it horizontal instead of vertical mount. To fix my x-carriage wheels hitting the P/S mount, I moved the connections back 1 v-slot. The X motor & idler mount have 2 V-slot connections on the inside & 2 on the outside. This gives me a little extra travel. I reflashed the firmware for the X & Y swap & have all the wiring hooked up ready to test. I did check the limit switches with the 5v connection & they trigger correctly. I am going hiking this morning, so might not get anything done on this today.

 

Mark you were right about there not being enough teeth to grab the belt. I was able to get it working today & it did have a hard time grabbing the TPU belt. Of course the real belt will be better. When I pressed down on the belt where those new idlers would go, it did engage. Drawing the path of the belt against that gear, it looks like I have less than 2 teeth engaging with current setup. 35.1 degrees versus 115.6 degrees with your recommended setup. With 16T timing idlers that is 22.5 degrees per tooth. I looked at the perimeter length of the new belt path & I should still be good for the 444mm length I ordered. I attached a screen capture of the layout of the new path versus the old. If I can get my 3d printer to cooperate, I will print out the new y-axis plates tomorrow. I only had to make 14 changes to the standard Marlin configuration.h & 1 change to the Configuration_adv.h file. When I add the servo for the pen there will probably be a few more changes to those files.

 

Thanks for the suggestion & thought it was a great idea until I looked at my machine. Unfortunately that would put the drive pulley on the smooth side of the belt. I played around with Mark's idea this morning by just drilling 2-5/16" holes for idler pulleys on each side & just putting in 5/16"x2.5" bolts with the smooth shoulder acting as the idlers (I attached a photo of it). The bolt idea didn't work as they bend too much without any support to the frame. I might see if I can print a big supporting washer. I am not sure I have enough room to put those idlers in there & still use 444mm belt. The design I came up with would need 500mm belt. I will work on drawing the geared version & see what that looks like before deciding which way to go.

The Geared drive has a 36T drive gear with 22T wheel gears which looks like a 1:64 ratio by the Gear Generator. I have it drawn as a herringbone gear, but will probably make it a double herringbone. Should I make that ratio a more even number for the stepper motor calcs? I can get 33T gear(1:1.5) for the Drive gear by shortening my wheel spacing from 120mm to 100mm. That solves another problem of the Wheel screws being right under the bottom V-slot t-nut mount on each side.

 

Since I have come this far with the belt idea, I will give it one more version to see if it works. I have some smaller Diameter idlers (18mm OD) & worked them in so it still uses the 444mm belt. It took about 3 trial & errors to get these to fit with 444mm. By the geometry, looks like I should have about 4.5 teeth contact of the 16T.

 

I seem to have found my thermistor problem on my 3d printer. It was not the thermistor which I didn't think it should be since I just changed that recently. It seems to have be one of the MOSFETs on the SKR V1.3 board. After finding a problem someone else had with this, I swapped the board with the one I have on another printer that is not being used. The problem has gone away. It is nice to have a spare board for troubleshooting. I could probably still use that board on my rolling plotter which doesn't need that circuit.

Since I had to reprint these y-axis plates I made some more changes. I added another bearing to the Y plate or I should say Wheel plate since it is no longer the Y-axis, but the X-axis. Since I didn't need that access for the belt anymore thru the V-Slot, I removed the t-nut connection on the bottom which was a pain & added another screw to the ends. That removes 4-Tee nuts from the design. I redid the wheels to be sunk in the middle so the bearings are only 3mm apart instead of the 10mm that was there before. Here are a couple screen shots of the changes with one of them being a section cut of the wheel assembly. On each wheel there are now 4 bearings, 4 washers, a 5/16"x2" bolt & locknut.

 

Think I might have caused my electronics problem with static. I have been using a metal feeler gauge for years without any problem for adjusting nozzle height & noticed a spark at one point, so I am now touching something else metal before touching printer. I probably should switch to a piece of paper for feeler gauge to avoid that problem. Since I upgraded my wooden frame on that printer to all metal, I am now wondering if that makes it more prone to catch static. It doesn't help that my room has carpet.

That is an interesting idea with the idlers on those 2 screws. I had looked at getting the idlers with M3 screws to save space on the head & nut width, but seems M5 would give them better support. Maybe a fat plastic designed M3 spacer would make that not a problem. If I order some from China, I will add 4 with M3's to try it. Is this how you meant for those idlers. It does make the belt a little longer. 461mm instead of the 444. It does give more coverage angle for the teeth on the drive gear. 166 versus 106 degrees or 7.3 teeth versus 4.7 with 16T drive gear. I could probably shorten the wheel base to make it work with the 444mm. I will stay on my current track for now. Keep the good ideas coming.

 

It is alive & first movements with the TPU went surprising well. Here is short video I made of it.