Ideas for improving mechanics
I recently purchased this printer; it's my first experience with 3D printers. I'm observing the printing mechanics; when the printer makes sudden movements, the frame wobbles. This makes sense, as the laws of inertia are still in place! I'm starting to think about ways to upgrade the printer to compensate for this inertial component. It would be a good idea to attach a small weight to the belt that would move in the opposite direction of the extruder, thus compensating for sudden braking or accelerations that causes the frame to wobble.
For the X-axis, this certainly doesn't seem difficult, but for the Y-axis, it's a bit more complicated.
Am I right in my thinking? 😏
RE: Ideas for improving mechanics
I recently purchased this printer; it's my first experience with 3D printers. I'm observing the printing mechanics; when the printer makes sudden movements, the frame wobbles. This makes sense, as the laws of inertia are still in place! I'm starting to think about ways to upgrade the printer to compensate for this inertial component. It would be a good idea to attach a small weight to the belt that would move in the opposite direction of the extruder, thus compensating for sudden braking or accelerations that causes the frame to wobble.
For the X-axis, this certainly doesn't seem difficult, but for the Y-axis, it's a bit more complicated.
Am I right in my thinking? 😏
I wouldn't over-think this. The input shaper compensates for this already. One way to dampen vibration is to change out the feet. There are many mods on Printables for this. I use the Hulu feet mod.
RE: Ideas for improving mechanics
You dont need to do anything with this. Let it be.
If you think the Core One shakes a lot, you should see the H2D. Its on squishy feet and the entire printer moves - a lot - during high speed printing. Sometimes enough to concern me TBH! But I leave it and the print always comes out perfect regardless.
See around 9:40 here:
RE: Ideas for improving mechanics
I wouldn't over-think this. The input shaper compensates for this already. One way to dampen vibration is to change out the feet. There are many mods on Printables for this. I use the Hulu feet mod.
I can't understand how a shaper can compensate for anything? You have a mechanical component that has mass. Software solutions can only eliminate the problems caused by the mechanics, but they can't solve them.Again, vibration, legs. This is dealing with the consequences, not the causes. CNC machines, such as the small 3018, use algorithms within the trajectory planner that take acceleration into account. This is easier for a printer.
I'll write below about my thoughts on solving mechanical problems. This solution isn't new; it's already in use; it just needs to be applied correctly in this case.
RE: Ideas for improving mechanics
Let me try to explain what's happening again.When the print head changes its position, it does so with acceleration. It has mass, so a force acts on the housing in the opposite direction. This force causes vibration. This printer is good, made of metal, and very well designed. However, this doesn't eliminate vibrations, which, when combined, can lead to quality loss.To reduce these vibrations, counterweights are used. They move in the opposite direction of the main mass, thereby compensating for the force.I borrowed a picture from the documentation to illustrate the use of counterweights. Unfortunately, the belt is unsupported in that location, and the mass should be approximately the same as the print head.
RE: Ideas for improving mechanics
There are many trade-offs in machine design. I would much prefer linear motors than belts, but the cost today could be prohibitive. Machine tool vendors handle vibration with added mass. Some top end machines even pour a cement block in the base of the machine to add mass and vibration dampening. Some put the Core One on a heavy paving stone if the table is not stiff enough.
RE: Ideas for improving mechanics
And it's a core xy. Both motors/belts have to move to do a move in x or y direction. I don't think there will be a single belt anywhere which moves equally and opposite to the print head.
I guess could be done on a bed slinger, but you'd need to add so much weight (and linear bearings or whatever) to support that weight.... it'd just slow the whole thing down anyway.
RE: Ideas for improving mechanics
CoreXY kinematics are non-intuitive at first. What is confusing is Prusa calls the 2 motors and belts X and Y. I really think they should be labeled A and B. So in a cartesian plane where the print head is moving diagonally such that X = Y then only belt A is moving. If the print head is moving diagonally such that 250 - X = Y then only belt B is moving. If moving orthogonally in X or Y then both belts are moving but in the opposite direction.
RE: Ideas for improving mechanics
Doubling the moving mass in the CoreXY stage (or even quadrupling it, if you want to compensate the moving masses in X and then Y too) will be harmful: You will lose acceleration capability since the motors can only supply so much torque, so you will need to print more slowly. If you are willing to accept that, you might as well limit the motor accelerations in the slicer in the first place -- that will reduce shaking with zero hardware effort.
RE: Ideas for improving mechanics
Counterpoint (No pun intended): Don't overthink this. Let it wobble. When the stepper motors stop suddenly, or reverse direction, there is a force. This force makes the printer chassis move. If you brace the printer chassis with a solid steel I-beam so it can no longer move, guess where that force goes when the printer can no longer move? Right back into the bearings and motors, which would probably then have a little less life span.
Firmware algorithms and slicer details that are designed to account for this and minimize print defects are designed for the way most of the printers are being used; sitting on a table without being encased in concrete and welded to steel I-Beams 🙂 Let the machine sit happy and work.
Just because you can see it wobble and vibrate does NOT mean that it is a specific problem. Prusa and other companies have optimized their firmware for this specific usage case; the machine sitting on a table and printing.
Cheers,
Neal
RE: Ideas for improving mechanics
As another example of machines that move; My ink jet computer printer used to sit on a printer stand. The entire stand would sway gently back and forth due to the acceleration and sudden reversal of the print head while printing. Eventually, I got a larger office desk and got rid of the printer stand. Now the printer does not move, but there was NO change in print quality. The printer just did not care. It does what it does. It's mechanical, and as rightly pointed out, has mass and acceleration.
Neal
RE: Ideas for improving mechanics
CoreXY kinematics are non-intuitive at first. What is confusing is Prusa calls the 2 motors and belts X and Y. I really think they should be labeled A and B. So in a cartesian plane where the print head is moving diagonally such that X = Y then only belt A is moving. If the print head is moving diagonally such that 250 - X = Y then only belt B is moving. If moving orthogonally in X or Y then both belts are moving but in the opposite direction.
I completely agree with you here. I gave an example to give some thought to those familiar with physics and the underlying causes. I know that the belt movement doesn't correspond to precise coordinates along the axes. Theoretically, it would be possible to add two beams that are independent of the belt movement direction and connect them to the head's center of mass using additional belt rings.
RE: Ideas for improving mechanics
Doubling the moving mass in the CoreXY stage (or even quadrupling it, if you want to compensate the moving masses in X and then Y too) will be harmful: You will lose acceleration capability since the motors can only supply so much torque, so you will need to print more slowly. If you are willing to accept that, you might as well limit the motor accelerations in the slicer in the first place -- that will reduce shaking with zero hardware effort.
Yes, this adds weight and load to the belt when combining belt systems. Adding additional belts to pull the counterweight also adds load to the engine. However, the fact that this compensates for the vibration component should actually have a positive effect on wear.
RE: Ideas for improving mechanics
CoreXY has been around since 2011 so it's hardly new and there hasn't been the need to change the basic design in 15 years. Here is the original writeup: https://corexy.com/index.html
I'm still not sure what improvement adding a damping mass creates. To Jurgen's point I can imagine it causing more harm than good.
Anyway, as others have pointed out, all of this can be controlled by the slicer by limiting acceleration (stealth mode) and employing input shaping.
RE: Ideas for improving mechanics
Adding any weight to the moving parts is stupidity non plus ultra in this case! On the contrary, in this case the moving parts should be as light as possible, so that the inertial forces are as small as possible, thereby reducing the strain on the motors and increasing the accuracy of the steps. Balancing weights are only added to specific mechanisms, which is definitely not the case here.