Tl;DR : any motion system either Cartesian or Core X Y whatever the case up top and then mirrored in an equal and opposite bed motion system action. Of course each of them would only move 50% of the total distance... also with some extra steps either through software or mechanically reducing you could possibly do the bed as 25% or 33% of the total movement or even have it adjust as the print got bigger and the bed got heavier the bed would move less and the head would move more ending with the printhead doing 100% of the movement at the top of the print and only 50% at the bottom. So just to clarify in its simplest form if you needed to print x + 10 the printhead would move x + 5 and the bed would move x - 5 (or x+7 and bed x-3) so each motion system would be complete yet mirrored in the bed as well as the gantry, Whether it's Core X Y or Cartesian. Also I have thought of what I believe to be some reliable methods to do this without changing the firmware or adding stepper Motors just with belt path and gears and rods. As far as I know I have never seen this and I can't imagine it's not because I'm the first person to think about it and although it does introduce some definite complexities that could make it more harm than good I believe none of them are difficult to overcome and in fact I have been obsessively running it through in my mind for days and I believe I've come up with a way to do this with good constraints for minimal slop and lash and also minimal additional parts. I'd really love to run it by somebody who is experienced with motion systems and also Being a genius would be a plus. Be that as it may, any and all input would be greatly appreciated other than the inevitable naysayers who don't give a reason why it wouldn't work but they just say no. please point out problems and reasons in pairs Thank you!
ANYBODY? I am going to build this once money allows and since I have enough spare steppers I may just use two additional to limit extra motion and plan to do that with Klipper but I'd love any input if you've heard of anything like this or thought about it or just what you think now? In a best case it would quite literally double print speeds and I'm planning to do each as corexy... Call it coreXYYXeroc or maybe XYYX OR CoreeroC for short... I believe I will beat the record for fastest non delta benchy... What do you say? You know, you could also improve speeds while also improving accuracy by slowing each side down a bit but the combination would still be faster.... I'm surprised I've never seen this... Also, since I'd like to have independent motors I would REALLY appreciate help on how to take the calculations, invert them and spit then out to the second xy in the firmware... Since the calculations would already be made for the first motion system it wouldn't increase the computing by much, ... Wait I could just wire the other motors backwards and duplicate the outputs huh?
You will probably not see a speed gain since the bed is heavy (relative to head), this means it accelerates slowly. a CoreXY/Delta head can accelerate much faster than the bed ever could, so coordinated moves will be always limited by the bed speeds.
I hadn't even seen this, but I'm assuming most people were put off by the stream of consciousness wall of text? Maybe organise your thoughts a little more clearly on the page so that people can actually parse your theory and proposal. Ellipses are used to signify run-on thoughts or speech, not separated ideas, so you're effectively trying to get people to read long single-sentence blocks, which is mentally taxing.
Thank you for the reply! So even if the bed only moved at 10% of the speed of the head, and you set the steps accordingly as to never slow the head, you would still see a 10% speed increase, right? I suppose I need to figure out exactly how much faster a core xy head is than a bed and see if that reduction would be worth the cost of an entire additional motion system... I've never had it used a corexy and only really know of them visually speaking from the screening hevort so would any of you with experience in more average corexy and bedslinger be able to give me a rough estimate? Thank you all for the replies... I hope my whitespace is more enjoyable btw I was literally speaking voice to text so it was actually a wall \stream of consciousness because I'm on the road and don't have my laptop but I'll take care to go through and make it less crazy lol
If the bed is slower than the head then when they move at the same time the head can never move faster than the bed since they have to be coordinated. The resulting move will be limited to the slowest acceleration and slowest max rate of the 2 axes. Look closely at the hevort and voron systems, they are being tuned for maximum speeds. A very high max speed without a very high acceleration is not useful on a 3D printer since many motions are too short to accelerate to the max speed. Very high acceleration requires very low weight. I saw one recently with a carbon fibre X axis tube that did a 7 minute Benchy. no way a moving bed can keep up with that head! Increasing the motor power might help, but at some point the limits of accelerating the bigger motor itself will come into play and you are back to the beginning. Sadly the art of effective dictation has been lost along with the typing pool and young ladies with shorthand (-: My mother is 88 and still uses shorthand. I certainly cannot dictate having grown up with 'write it/ edit it' and since 1986, type it/edit it, I just do not think in the way required for dictation, and reading stuff without punctuation is something I normally avoid
Yeah I like the idea- and you could probably test it by modifying the HAL on a basic LinuxCNC config, so mostly in software; 3 axes, 5 joints. Reminds me of a dual-rapier loom or something. But I agree with David that it won't result in a significant increase of speed, because of the acceleration problem- and even if you were somehow able to move the bed that fast, I suspect you'd end up shaking the print off. Belts, ultra-light gantry materials, and servos are the way to go- basically, the laser cutter folks have already solved this problem of momentum, it just has to be adapted to the heavier head of a printer. I tried voice-to-text once, but formatting things like "period" and "new line" or whatever just got written in as text, so I gave up.
