Opinion on Core XY belt layout
-
I'm throwing 4kgs around without any problems. Nema 17s at 1.8Amps so on that basis, Nema23s at 2.4Amps looks like overkill but hey…......
That's encouraging! Overkill is a bit par for the course for me tbh, but I also have some loose plans to chuck my Foredom toolshaft around with it eventually, possibly as a 3D print finishing experiment. Anyways, I like OP's idea of a tooling plate sandwich, as I understand it, the Railcore guys have been struggling with bending bolts on the pulleys over time as well as movement in the threads, though I guess it's a general corexy issue, relatively high torques involved.
-
…........................... the Railcore guys have been struggling with bending bolts on the pulleys over time as well as movement in the threads, though I guess it's a general corexy issue, relatively high torques involved.
I hadn't really looked at the Railcore so after you mentioned it, I just took a quick gander. Yes, I can see that would be a problem. Personally, I would never fix an axle at just one end. Apart from the motors, all my idlers and carriage wheels are on axles supported top and bottom (or at both ends).
-
Food for thought, you've got me thinking now actually (sorry for the hijack Zakfarias, but might be relevant to your build too) I've got a fair bit of 4mm aluminium sheet, with 4 brass stand offs in a nema17 pattern might make a quick and dirty idler sandwich.
-
@bikingviking
: No problems with the hijack, it is all very relevant to my issues. I am hoping for higher quality performance over the railcore(which is in the list of top 10 favorite designs). A lot of what I am doing with the sandwiched tool holding plates and using shoulder bolts instead of regular bolts for pulley guides makes up for its longevity issues.@all
Looking at diamond, I could run a 3:1 but its air cooled and I would basically run a gantry system stacked similar to @deckingman. However, I noticed reprap.me has a 5:1 that is watercooled…a little too many inputs (weight on secondary gantry) in my opinion but it is water cooled...Running a monomaterial would be great but multi-material is limited. Not a bad thing but at this point the design has so much $, time, and CNC work done on it that you may as well go for over the top performance.
I could run cartesian IDEX and just reconfigure my frame accordingly, however, with my minimum build volume requirements it would be a massive megaton of weight to sling around 2 AQUAs on a heavy duty large gantry.
Dual coreXY seems to be all fun and games, of course the machine footprint becomes massive to maintain build area with 2 dual rail gantries. Hmmm, belt routing with the current setup might become even more conflicting though, plus would need a helping hand on the firmware config.
Using the filament monitor and having it pause/resume/even notify when there's a runout would be nice and simple. However, what happens if the new roll of filament has a different diameter say 1st roll is perfect 1.75mm and the second is 1.73 or 1.72?
@DC42 I would definitely be interested in a few of the monitors, even if they don't make it into this build I would definitely use them on something.
Side note; I use S3D for all slicing mostly because of how many machines I manage and easy UI. Wanting to reference this since it does effect some of the optional routes to go.
-
…............................. and using shoulder bolts instead of regular bolts for pulley guides makes up for its longevity issues.
In my day (and still to my old fashioned way of thinking), a bolt by definition is partially threaded. If it is threaded all the way up to the head then it isn't bolt, it's a Set Screw.
Whatever you call them, it's best to run idlers on an unthreaded (smooth) shaft.Ref the water cooled Diamond. I've looked at them. What puts me off is that there is no data to prove that they work (or otherwise). The reason that puts me off is that I've done an awful lot of work with Diamond hot ends (see my blog), including the 5 colour (air cooled version). The air cooled version has (undocumented) issues with heat creep which causes filament to swell and jam in the area of the heat break. I have a sneaky suspicion that RepRap.me knew about filament jamming but were mistaken about the cause because it ships with a massive 80Watt heater that is simply too powerful. The real reason for the jamming issues was that, in order to get 5 heat sinks in the same footprint, the lower fins are reduced in diameter. This makes them less efficient so a much bigger fan is needed to solve the problem (not a bigger heater which does not solve the problem). My evidence is all here https://somei3deas.wordpress.com/2017/10/14/the-diamond-5-colour-part-2/ and https://somei3deas.wordpress.com/2017/10/23/the-diamond-5-colour-part-3-finally-tamed/
So given the fact that the stock 5 colour air cooled diamond has issues which are not made clear, and the lack of any information about the water cooled version, I'm reluctant to part with any money in case it too has issues. I'd happily test one if Peter Bogely (he's the CEO of RepRap.me) supplied me one but I doubt that'll happen.
-
From a molding/tooling background; when I say shoulder bolt I mean a precision shaft with a smaller thread at one end. Say a 5mm shaft with a socket head cap screw head and m4 threads at the other end. Amazing how many things have the same/similar reference.
Just what we have always called them around the shop, i think we are supposed to call them stripper bolts.Anywho, I agree on the diamond findings (I've followed your blog for a while now) which is why I have always been leery of going to a mixing hotend. Not sure if I have the same level of virtuous patience to handle those headaches as you.
-
I have a masochistic streak - essential with a mixing hot end
-
Hopefully DC chimes in here again, I am wondering if the main use gantry (call it Tool #1) can be true CoreXY and then have a second smaller carriage (Tool #2) ride along with it that has a small Nema motor to drive it in the "x" axis? Kind of a hybrid coreXY and cartesian. It would use the same XY axis but the second carriage is driven in the "X" independently. I don't know if I am making sense right now.
-
Hopefully DC chimes in here again, I am wondering if the main use gantry (call it Tool #1) can be true CoreXY and then have a second smaller carriage (Tool #2) ride along with it that has a small Nema motor to drive it in the "x" axis? Kind of a hybrid coreXY and cartesian. It would use the same XY axis but the second carriage is driven in the "X" independently. I don't know if I am making sense right now.
There is already a CoreXYUV class for CoreXY IDEX machines, but it uses an extra motor compared to the system that I think you are describing. I think you could use M584 to add a U axis for the second carriage to regular CoreXY.
-
I'm using the single rail on my build ala' Jetguy's design. It is a flat 2"x1/8" aluminum bar with a 420mm MGN12 rail on it. Makes a very stiff gantry esp. in the X/Y plane. I'm using the Rep2 single extruder kit from Bondtech. Is a very simple and compact arrangement using a MicroSwiss heat brake.
-
I'm curious to know why it seems like these designs use smooth bearings against the toothed side of the belt instead of a toothed pulley as an idler? Preferably the same size as the pulley as on the motor.
