hey everyone, just wanted to hop into the discussion. im starting this build, and as it turns out its going to be very similar to akeric's build in size and build volume, so thank goodness his blog is SO GOOD and well documented! i am however going to use a few of the adaptations from trublu832's d-bot as well! im also really looking forward to hoping into the development cycle and parts design/revision on this project, i think Carl is and has done great work with this build and i cant wait to get mine going.

ill be finished paying off my student loans at the end of feburary and i am currently [slowly] buying parts for the build. luckily i already have a small delta printer and a genuine E3d lite6 and a knock off E3d v6 so i can still print parts and test fittings and that sort of thing, but i wont be able to test it in c-bot context for a little while. i am an experienced game developer so modeling is certainly a strong suit for me, so hopefully i can put some of those skills to good use and ACTUALLY contribute to this build instead of just lurking like i have been the last month!

thanks again for all the great work you all have done so far!

 

Forum won't let me respond, says everything I write is spam. It's not! Thanks for the responses I will try to respond more in depth later when the forum lets me.



​

 

I don't see why the size of xy would effect belt stretch. The same continuous fiber runs throughout the length. The belts shouldn't care how long the axis is, just how much the gantry weighs. I would think that the critical point here would be the rigidity of the linear way. Is 20x40 good enough for a meter unsupported. I think it should be.

 

I guess, Jetguy's ulti-ulti-replicator 2, a 4 foot cubed build area used gt2.

 

When I get back to the states I'll find the links for his build thread, Ryan Carlyle posted some math for sizing unsupported linear ways for both round rod and extruded aluminum.


Vernon's big bot wasn't the largest printer, seemecnc had their "parts daddy" there, a 5 meter tall delta with a shop vac feed pellet extruder. The New Largest Delta 3D Printer on the Planet, The SeeMeCNC PartDaddy Measures 15 Feet Tall . I think that was day I decided to build the triple-c. Suddenly my rep 1 seemed, well, inadequate.

 

Somewhere on the openbuilds site is deflection data for the various v-slot profiles. I don't remember specifics, but 2040 should be good for most of what we're likely to build.

 

What happened is as I turned the screws on the corners of the rail with 1 Z screw the bed stayed and the rails twisted.
It seems to me that even if I were screwing into 4 solid columns, 1 of the screws will loosen up. When the mounting points r not rigid, those points r pushed down or released up. The only issue now would be if there was excess weight on 1 of those corners the frame might try to tip. There is a possibility something else is working here like the frame isn't rigid enough but it sure feels as tight as I possibly can get it.
Anyway it seems to work better with only 3 until I find a problem with that.

 

So i think my problem with my z axis is my leadscrews.. not fully understanding back when i bought them, i got some 4-start leadscrews, so 1 rotation = 8mm of travel. that means each step is .04mm of travel, so 0.1mm of movment is 2.5 steps. Depending on exactly where the bed is vs the motors step, i think the resolution means my '0' point is between steps.

Since i don't want to buy new screws i'm going to use different sized pulleys on the motor vs the screw to gear it down a bit.



Also i started using your direct drive extruder. Working awesome. I've found if i cause too much backpressure on my nozzle that the extruder will push the ptfe tube out of the little holder to the side.

I attached a little extended holder. I know it fits but i haven't had a chance to try it to see if it prevents it.

 

Your calculation for step size is correct, but you might not be aware of the ability of stepper drivers to hold the rotor at a division between two steps, therefore increasing the number of steps per rotation. (Microstepping).

You sacrifice holding torque (exponentially as you increase the microstepping resolution) for more steps.

This will be what most people in this forum will be doing to gear down their steppers.

Anyone with a triple-c bot style set-up (motor connected to leadscrews with a belt) can easily use gears/pulleys to get higher resolution rather than microstepping.

 

thats exactly why .2mm has become the standard in printing because most people dont realize the z motor microstepping can become an issue, and you cant just print at whatever layer height you feel like that day. however with new boards and stepper drivers making their way into the affordable market like smoothieboard v2 and azteeg v2, i think the 1/128 ustepping will start to take a huge leap when those motors can hold those higher microsteps with more torque, and make the smaller resolutions and complex prints much more reliable.

 

Here is what the Z endstop bolt holder looks like on the Dbot variant. You could just make it the default since is has the same functionality as before, it's just one less M5x10 bolt.

 

Is there an option to turn the bolt upside down?

I like to move my bed into the correct position and then just adjust the end stop switch to trigger at that height. It looks like the screw head touching the switch like in your design makes it difficult to fine tune that.

 

The bolt is just loose enough that I can just twist it by hand to fine tune the Z height near Z=0.
The threadless part of the bolt also prevents it from being installed the other direction, maybe with a different bolt you could flip it.
Or flip the bolt and move the Z endstop down I suppose.

 

Friendly reminder that the hot end generator incorporates both the Z min endstop and X min or max endstop into the hot end carriage. This provides you with auto bed compensation as well as reduced mounts on the V Slot. I still plan to add Carl's direct drive as an option to the generator, though we are deep in the FRC season right now and my time is nearly completely devotated to that. If you want to go Bowden, it's all set as is.

 

No worries, just offering up an alternative, not sure who knows about it and who doesn't.

Speaking of Autodesk, just started using Fusion 360 for our team's robot design. Enjoying it so far, let me see if I can get one of the stls pulled into it.

 

Let me add my 1¢
I have 3 screws and 1:1 gearing (same pulleys on screws as on motor). They are not trivial to get right. It took me a lot of trial and much more error to get them not perfect but close enough.
If the platform does not move down by itself or with very little weight on it without the motor attached it is probably binding which might explain the backlash u r seeing. If the belts r tight and the pulleys are solidly fixed then look at the following:
1. Insure ur screw falls thru the nut with only gravity. My screws were severely flawed (Openbuild) and I had to file the entire length of the threads to get the screw to fall thru the nut.
2. insure that ur screws are very perpendicular to the platform. It took a lot of tweaking to get mine there. Mostly it was at the bottom of the travel that it really bound up but going down was not a problem because of the weight.

Changing the gearing isn't going to solve backlash. U r pushing the platform up. Going down u r just removing the blockage and gravity is taking it down.
If u r seeing that backlash it means something is loose or sticking.

 

Your now the second person I have heard complain of faulty leadscrews, Ryan Carlyle had received two recently with incorrect pitch. I had no trouble with my zen toolworks screws and there is zero binding. Unhooked from the motor, my Z will freefall.

Nice work Carl, It looks like a good number of the parts have been beefed up a bit.

 

Yes, I give Openbuilds a 10 for customer support, they sent me 3 replacements and they were all the same. The leadscrews look like they were made in China 30 years ago, very poor quality. They will work ok with their Delrin nuts but not at all with brass nuts unless they r filed smooth. That is why they r cheap.

 

I bought a set, and they work very nicely with the delrin blocks. Now I just need to finish assembling the printer and find out if they're consistent.

 

I'd be interested to know if they improved any. If u look at the screw with a glass u can see if there r cut tool marks in the grooves. It looks like a dull tool was used to cut the grooves. They sent me a Delrin nut with the last replacement which is how I found out they work with them. They should state that they don't work with brass nuts if the quality is still the same. I would be concerned with wear on the nut from the rough grooves.

 

So it's not the best photo I've taken-- Shoulda gone F11, instead of F8, and should have dusted the thing first. I believe the red and black marks are actually reflections, as I can't find them without the camera.