I don't mind coding in Python. It's maintaining it that sucks. I started doing statically typed functional programming a couple years ago at my job... After completely refactoring a core data structure used by 30+ assemblies and having it "just work" as soon as I resolved all the compile errors, I never looked back

 

I have kept that slot. Seems like my printer printed it a bit small. Taking a drill to the holes fixed the issue. Though, I have to ask, how tight should the carriages be? Should the carriage move if I hold an extrusion vertically? The carriage being only the plastic parts and wheels and nothing else attached.

Here are the files. If anything looks sloppy or wrong, tell me. The naming is the same and I just appended _SOLID_V. Feel free to change that however you like.

 

So I designed the carriage sled adapter I described earlier, but noticed I still have a half degree trapezoidal error so I'm working that through before testing it out. If anyone wants to try it, I have included the STL, and the code is checked in to github. Just print 2 copies and put them together, then put the bolts through them. I'm hoping to test it tonight, but don't know if I'll have time.

 

I have 20 teeth on the stepper and 40 on each of the leadscrews. That works out to 800 (micro)steps/mm or 0.02mm per every full step which makes it easy to pick slicer layer heights.

Someone asked a week ago how fast I was printing with my direct drive extruder, sorry I was on vacation. I usually print at 70mm/s, 110% for infill, 2250 acceleration. I've run at 70mm/s as well for a while with the same results except I had to turn up the hotend heat a bit and I started getting some ooze so rather than mess with retraction I just backed the speed down and save 70mm/s for "draft quality". EDIT: Just realized I'd been printing at 70mm/s all the time for at least a few weeks! I do get some vibration lines in some smooth print areas.

Also: Thanks for the links to the short body steppers, I didn't realize they made NEMA 17 motors so small!

 

@CapnBry Hey Sir I vaguely recall seeing a pic of what I think was your setup and recall seeing a bubblewrap "tent" of sorts over it. I'm assuming that was for ABS. I was wondering how well that setup worked out for you? Were your electronics or motors overheating ect? How long did you print with it? Any issues to speak of?

 

For some reason I found the 4x6 inside corners would not fit in the space available between the extrusion and the Z axis motors so I have remixed these smaller versions. I also found the two pairs of screws closest to the corner interfered with each other, so I removed one pair.


I have published these on thingiverse (Thing 1568176).

 

10"x10" build plate. 450mm legs, 390mm horizontals, 370mm bottom sides. The frame feels very solid so probably didn't need them but they won't hurt will they? Waiting for bits to arrive so drawing them out kept me busy.

 

Haha yes that was me!


The bubble wrap was just temporary covering over the dowel structure. I thought I'd give it a try because, hey, a lot of people insulate with the reflectrix bubble wrap stuff. I happened to have a bunch on hand thanks to Mouser. Even with packing tape sealing all the edges there was a lot of heat loss. I can't remember the numbers but it got up around 40C and didn't make much difference in ABS curling on large flat prints. I then cut cardboard panels and mounted them with some 3D printed loops. This worked somewhat better but I needed to redesign the back brackets to allow cabling to come up so the side cardboard pieces didn't quite fit right any more. I then just threw a fleece blanket over the top and man that worked like a charm. Temperatures climbed over 50C. The X and Y motors were still outside of the blanket so they were normal temperature. The Z motor and electronics are on the bottom so it isn't enclosed either. The extruder stepper definitely ran hotter but I only was running it at 0.7A out of 2.0A so it was well within spec and I could hold my finger on it for a few seconds before I felt it was starting to be a bad idea. I also printed one of the power supply air ducts because when that fan kicked on it would lower the enclosure temps at least 5C. I have a spare EPCOS 100K thermistor hooked to the extra thermistor input on my smoothieboard which just dangles off the top of the frame to measure the temp and show it on OctoPrint.

I need to re-cut the top dowels a little longer then I'm going to get some hardboard and cut it to fit with an acrylic front piece. The printed mounting loops act like hinges so the front can be flipped around 270 degrees to be open. The whole top piece in its entirety also tilts back, hinging on the back corners, so if you have enough space behind the printer you could theoretically tilt the whole thing open for maintenance. I do not have this kind of space though.

Although that seems like a lot of work so it is likely I'll just use the blanket for the next year or so before I get around to it :-D

 

Hey @CapnBry. I see you are running the two leadscrews off of one z stepper. I'm sure you have explained how you've set it up, but I can't seem to find it. Could you link me to it? Also I had to ask, I am also thinking of going the same way and will be using a 300mm×300mm bed. How powerful should the motor be? I already have a 5.5kgcm Nema 17. I dnt know if that'll be enough.

 

Yes it should be enough. I can't find a post where I talk about it fully, all I can find is this post and these pictures. There are just modified motor supports with a F608ZZ bearing sitting in it, with an 8mm shim then GT2-40 gears on top. There is a 2:1 gear ratio because the Z stepper has a GT2-20 on it so really almost any stepper will work. I run the Z motor (an official Kysan 1124090 5.5kg*cm) at 0.5A and it doesn't skip even moving 10mm/s. The belt is now a 976mm because I've spread the two leadscrews apart further so that there's no longer a chance of the hotend running into them unless it is out of bounds.

The flange bearing mounts I remixed myself, but Carl cleaned them up and I'd recommend his version because they're prettier than my squares. Z bearing mount carl.

 

@pushpreet Didn't see your original message asking about this, but since I did my setup based on @CapnBry setup, I thought I could share some pictures of it.

I used three idler pulleys to keep the angle around the toothed gears to the maximum contact with the belt. I have seen it done this way and the way CapnBry has his and both seem to work. I just feel more comfortable with this setup.


Here is one of the screws pulled up. You can see the bearing and the shim that sits between the bearing and toothed pulley.


Another view of the above picture. You can see where the flanged bearing sits in the mount.


Here is a view of the bearing in place and the screw pulled out.


Hope these help with your questions.

EDIT: The bed is 300 x 200mm. I used a 1220mm GT2 belt.

 

They run better at 24v input to the stepper drivers, there is still a slight high pitch noise when the motors are stationary but overall it is much quieter when printing.

 

@CapnBry Happen to know where I could pick up those dowel bracket files? I think I might like to give that a crack. Extra thermistor input on a RAMPS work pretty much the same way for ambient?

 

Thank You @CapnBry and @Elmo Clarity for those detailed pictures. Clears everything up. I was thinking more on the lines of what @Elmo has done, albeit with two pulleys screwed right into the extrusions. This would also help tension the belt.

I'm glad that I somehow ordered all the right parts when I ordered everything back when I had only @adamcooks design to refer. 608 bearings, GT2-40 and 20/16 gears and even the belt sizes I have are around +- ~100mm. Only problem is the bearings don't have a flange, I am thinking I will change the printed parts to have one on the underside so the bearing doesn't slip down? And second, the GT2-40 has a 5mm bore as I was thinking about turning down the Lead Screw to 5mm at one end, but now I think it'll be easier if I just bore out the Gt2-40 to 8mm.

Yet another question. As the lead screws are sitting lower than they would if I were to couple them to steppers, does that mean I should cut the vertical extrusions shorter? Currently I have two 300mm screws.

 

Here you go, they're mostly just simple sockets, pretty ugly really. The sockets are designed for 3/8" dowels but you can edit the OpenSCAD file and change that to whatever you happen to have on hand. The angle of the back slope is 30 degrees which will help you calculate the lengths of the dowels needed for the front to back run given the height of the enclosure and the depth of your printer.

The hangclips work OK but when I make them for real I'd probably make them a bit bigger by the nut trap, as they can be easily split when screwing them together. I haven't had one break apart completely, but definitely could use a little revision.

I think you can just recompile Marlin with another thermistor defined and it will just start showing up. Not entirely sure about that.

 

It would be great to just print a bottom on the bearing mount so flanged bearings wouldn't be needed. However, that would be a pretty large bridged area which you'd have a hard time printing a flat surface for the bearing to sit on. I'd also be weary of trying to bore out the GT2-40 from 5mm to 8mm just because if the new hole becomes slightly off-center then your leadscrews will wobble back and forth.

I am not sure about the height of the leadscrews. I bought the "250mm" screws, which are 290mm and the overall height of my printer is ~450mm. You can see the sizes of all my extrusion here, but I know the calculator has changed since the original version.

 

I just thought about how the mount will be printed, and yes, printing a bottom doesn't seem feasible. Let me see if I can figure something out, or I'll just get the flanged bearings.

I had the same concern about the GT2 gears, but that just confirms that I should bore them on a lathe, I've got access to one.

EDIT: Forgot to say. Thank you for the input on the lead screws. Seems I just need to input the length of the screws I have and it should work.

 

@Elmo Clarity @CapnBry A very much "thank you" to the pictures and info on your setups...Im finally putting all the pieces together..

 

They come out flat now with a 105C build plate on hardware store glass, unless the bed isn't level and one of the corners doesn't make good contact in which case you can definitely see that it isn't flat, but not so much it is like a plastic taco. The heated enclosure definitely makes a difference, and also the heated bed uses maybe 10-20% less power to maintain temperature so it is an all around win.

Another thing to consider is that we really shouldn't be breathing ABS fumes, and because I am often working in the same room as the printer, this does a fairly decent job of containing at least the smell. Not sure if I'm breathing less butenol fumes or microparticulates, but it definitely smells like I am.

@pushpreet you could always make 4x M3 holes or holes with nut traps on the top of the bearing mount and print a second piece of flat plastic that screws onto the bottom. It isn't elegant but it would work. Considering you can get 11 bearings for $10 shipped from US it might be worth it to just eat the $10 and save yourself the extra inventiveness. But since you have access to cool machining equipment, I'd be interested to hear if you come up with an inventive solution!

 

So by way of an update. I have been super stupid busy lately. But I finally managed to upgrade to RADDS, install my Flex3Drive, and upgrade my fan to the @AK Eric fan duct and an 18 CFM fan. Here are some pictures for those of you that are interested.

I threw in a picture of the carriage and of the first #3DBenchy I printed with it. Overall the quality is impressive considering I really haven't done any tuning on it at all. I feel like its under extruding slightly which could just be because I was too lazy to measure the actual filament and left it at "1.75" in S3D, its also leaving a few blobs here and there which could be because I have my retraction set super low. A tiny bit of ringing here and there but considering the lack of tuning I'm super happy.

If anyone wants more details about the Flex3Drive and stuff let me know.....its pretty **** slick.

 

@CapnBry Hate to pester but when I open up the OpenSCAD file in there, I'm not sure if it's right. Perhaps with my lack of experience with that app is to blame. I'm trying to re-do them for 1/2" dowels ( they were on sale and seemed a little sturdier than the 3/8" stuff) I apologize in advance for my ignorance with this app.

 

No, but its sitting 2 feet from me. What would you like to see ?

 

Here is the STL of the front and back, the holes for the belt clamps/extruder mount are a bit larger than usual because I use heat-set inserts rather than M3 nuts just for the fun of it.

 

Can people point me to a link, or maybe post some pictures of how they are implementing their x endstop. There are some pictures on page 22, but they are not as clear as they could be.

 

Hello @Matt Mathias,

I am currently building Spaud01 D-bot (D-Bot Core-XY 3D Printer by spauda01)
I will install a Flex3drive which I have already ordered.

Your prints looks very good !

Do you have any advise about mounting and setting it ? I guess you had to change some values in Marlin concerning the extruder ?

Which HEAD do you use ?

Thank you !

 

Hi @IanT - It really depends on which carriage you use. There is no shortage of varieties here, mine included. Below is an example of a carriage produced by my generator which includes the X and Z endstops. Keep in mind that my Z endstop is an inductive Z probe that also acts as the Z endstop. I choose to incorporate the X endstop into the carriage itself as it reduced the number of wire paths I needed. I run nearly all the non stepper wiring through cable chains to the carriage.

These photos are from the assembly phase, wiring was cleaned up and run through cable chains since.

YL-99, modified for NC as X Endstop on right side of the photo.




Additional bumper added to the XY bar end for the X endstop. I believe more recent versions of the XY bar end have this built in. The generator still produces one in case it's needed.

 

Very nice! I guess my main interest would revolve around the particular advantages gained by using the Flex3Drive over either a pure direct or bowden setup once you have some hours on it. Items of interest would be increases in usable print speed, reduction of ringing and other artifacts due to the offloading of the stepper and overall accuracy of the system. I assume there is no noticeable spring "delay" in the shaft on retracts/advances? Any idea of the weight of the portion of the assembly that mounts to the carriage vs the typical Nema 17? I almost bought one when I was doing my build, but decided to wait until I had everything else close to perfect. Getting close! Anyway, once you have some time under your belt and a free minute or two, let us know as I am sure there are others very interested in this setup. Thanks!