Separate names with a comma.
Some features disabled for guests. Register Today.
Discussion in 'CNC Mills/Routers' started by Stokrotka74, Nov 13, 2020.
This is a Sphinx variation, aimed to minimize footprint and maximize workarea.
Work in progress.
Stokrotka74 published a new build:
Read more about this build...
Nice, I like the drawing!
Not sure you will need the belt reduction on each motor, will make the axis slower than needed.
The force the screws can generate is amazing.
Clever lady - welcome aboard.
reduction adds to resolution too. I plan weak spinde (noise restrictions) so it wont be cutting fast anyway.
of course reduction can be 1:1, or maybe other way: trade torque (and resolution) for speed.
Nice job @Stokrotka74 welcome to OpenBuilds and thank you for sharing!
I immediately clicked on this as I had had the same idea when redesigning the xcarve.
This looks like a much nicer machine and begged to have the same mods done.
Love your Onshape document (but then again I might be biased as my name is on the Onshape patent )!!!!
My concern about this design is the loss of the direct screw drive when adding belts - I want a machine for some very intricate work.
I am not sure how far your build has progressed, but I am wondering if you would be able to comment on any loss of resolution?
With reduction like this project the resolution is increased.
8mm lead screws like these have a direct drive resolution of 8mm/1600steps at 8th micro step or 0.005mm per microstep.
I have resolution like 2400steps/mm, 800 from screw, times 3 from pulleys ratio.
inaccuracies on belt systems comes from stretch of belts which is a function of their length. Here should be negligible
Final precision is still determined by run-out on spindle and general flex of structure (i built 1500x1000)
Frankly i have no capacity to precisely measure it, for now.
Yeah, i shall put some updates to this project.
Ovservant eye will notice i developed my own endstops for this projec, proably will do some batch for sale in my shop.
For control board, i made ESP-32 6motors on i2s and 74HC595
1) There'll be a lot of flex in the structure at that size, probably more than you immediately anticipate, but you can dial back cutting speeds/forces or reinforce the machine as necessary- the torque to drive more weight is readily available.
2) The screws themselves will also be a source of imprecision; potentially up to around +/-0.1mm at any given point. Stated precisions (when they even exist) are often optimistic on hobby-grade parts, which often render attempts to build high-precision machines somewhat futile.
Rob - I am new to the 'OpenBuild' machines and note they are using Acme lead screws - is there a drop-in recirculating ball screw upgrade kit or are we pretty much on our own here? Do you think it would make any/much of a difference?
Many thanks - Philip.
Affordable ball screws are source of trouble and inaccuracies of their own kind.
Guys who are making HevORT have stories about that.
I don't think any "kits" as such exist, but ball screws will often claim "antibacklash C7" and still have 0.010" of lash and up to +/-0.3mm vs nominal readings across the deck. No different to lead screws, essentially- if anything slightly worse because lash-reduction systems are more complex. The problem is in the rolling method of manufacture, I think. Generally I think more highly of ballscrews, but they're more difficult to use with simple control systems like grbl which don't have backlash compensation. My only grbl machine is belt-based, which grbl excels at (all of these big routers were originally belt driven too, once upon a time. Didn't work all that well, but did spawn some ingenious belt rigidity systems)
The problem is, ground screws of either type typically cost in the low four figures, and will only serve to show up the inadequacies of the rest of the machine (eg. you can push fairly lightly by hand in the middle of a 1500mm C-Beam and get up to a millimeter of deflection).
Hobby grade parts produce hobby grade results, unless extreme care is taken and the appropriate metrology equipment is available. But at the same time, that works just fine for most people. It's just that "precision machining" is typically going to be fairly out of reach.
^^^^Pretty much sums it up.
Note, play in the screws generally comes from the end bearings and the ball/lead screw nuts and any backlash in the system will only increase with time as these parts wear. Attempts at significant pre-tensioning only increases the rate at which wear on the end bearings will happen. Hobby grade parts beget hobby grade results (which are actually good enough for most people).
Sadly, you can't avoid tesioning those screws, as they will sag and bend (especially that 1500mm one) under own weight,
and than cause scary effect when you start spinning them too fast, you end up having jumping rope not a leadscrew-)
This machine is not about resolution.