4 motor CoreXY?
Posted by PulsedMedia
|
Re: 4 motor CoreXY? August 30, 2016 12:55AM |
From reading the other post there seems to be some confusion about what you are referring to, or at least to what I think you are referring to.
If you mean putting one stepper motor on each corner of the XY gantry and paring the diagonal steppers I have recently done it:
Links to Videos:
[goo.gl]
[goo.gl]
[goo.gl]
Photos:
So what is the advantage of using 4 steppers? The biggest reason the have 4 stepper is to shorten the "spring" length of the belt. As you know in a core XY system you have a lot of belting running around so shortening the amount of belt that is springing will permit the user to run the printer at more aggressive accelerations. Another side benefit of 4 steppers is that they make really convent 5mm shafts to attach idler pulleys to.
What are the disadvantages of using 4 stepper? Cost and motor honk on rapid moves. Motor honk can be completely eliminated by driving each stepper with its own driver.
For me it made sense to try but given my experience building this printer I would not recommend CoreXY. The added complexity of routing and tensioning the two belts along with the "spring" length of the belt make CoreXY really tedious to execute effectively.
I am currently on V3 of my XY gantry:
Edited 1 time(s). Last edit at 08/30/2016 12:58AM by FrankPropaneTank.
If you mean putting one stepper motor on each corner of the XY gantry and paring the diagonal steppers I have recently done it:
Links to Videos:
[goo.gl]
[goo.gl]
[goo.gl]
Photos:
So what is the advantage of using 4 steppers? The biggest reason the have 4 stepper is to shorten the "spring" length of the belt. As you know in a core XY system you have a lot of belting running around so shortening the amount of belt that is springing will permit the user to run the printer at more aggressive accelerations. Another side benefit of 4 steppers is that they make really convent 5mm shafts to attach idler pulleys to.
What are the disadvantages of using 4 stepper? Cost and motor honk on rapid moves. Motor honk can be completely eliminated by driving each stepper with its own driver.
For me it made sense to try but given my experience building this printer I would not recommend CoreXY. The added complexity of routing and tensioning the two belts along with the "spring" length of the belt make CoreXY really tedious to execute effectively.
I am currently on V3 of my XY gantry:
Edited 1 time(s). Last edit at 08/30/2016 12:58AM by FrankPropaneTank.
|
Re: 4 motor CoreXY? August 30, 2016 03:03PM |
Dentist,
The print quality is excellent provide the speed and acceleration settings are correctly set.
But to entertain that idea for a moment the error will never be more than a full step, as if it were the stepper motor would lose many steps. In my setup we have 80 steps per mm so the maximum error we can have will be 0.0125mm or +- 0.00625mm really nothing to worry about.
The total belt length is not relevant only amount of belt able to stretch under force.
Let us take the red pulleys as linked together and the green pulleys as linked. Now let's think about how force transmitted by timing belts. If we were dealing with a frictionless pulley and a frictionless belt the tension along the full length of the belt be the same and any force on the end effector would transmitted along the entire length of the belt. Timing belts and pulleys are not frictionless infact for our purposes we can take them to have infinite friction. How can we make this statement? If a force is large enough to overcome the teeth of the belt and pulley we have lost control of the system so we can consider that to out of our design scope. The same is true for any force large enough to force us to lose a step, any force that can cause a lost step renders the system out of control thus we are again out of design scope.
So we have now set the boundaries for our problem let's apply a force to our X carriage:
We can now see how the force is transmitted along the belt back to the first driven pulley it encounters.
Now lets take a look at a stock coreXY config:
You see how the force on the blue belt is transmitted along a much longer belt path in this configuration than in the 4 motor configuration.
Strain is defined at the change length over the total length of a member under load, so it is best to think of it as a percentage. Most timing belts have about a 2% strain before failing so for every 100mm of belt you can see up to 2mm of stretch in the belt, you will be taking most of the stretch out of the belt when tensioning it but there is still a good bit present. By using 4 stepper motors we cut the amount that the blue belt able to stretch about 2/3 rds! (assuming a square/symmetric XY travel) I have not run the exact math but this will certainly reduce the effective belt stretch of a coreXY printer. I would guestimate a 4 motor coreXY system can accelerate between 25% to 50% faster than the same system with only 2 motors; holding end effector deflection constant ie. the same print quality.
Yes the first video was my first XY gantry, it was not great and required a correction table printer correctly. The belts on my current design are parallel.
All this being said CoreXY is not really suited for large printers due the the amount of belt involved even with a 4 Motor setup.
Never stop making -Frank
Edited 4 time(s). Last edit at 08/30/2016 03:16PM by FrankPropaneTank.
The print quality is excellent provide the speed and acceleration settings are correctly set.
But to entertain that idea for a moment the error will never be more than a full step, as if it were the stepper motor would lose many steps. In my setup we have 80 steps per mm so the maximum error we can have will be 0.0125mm or +- 0.00625mm really nothing to worry about.
The total belt length is not relevant only amount of belt able to stretch under force.
Let us take the red pulleys as linked together and the green pulleys as linked. Now let's think about how force transmitted by timing belts. If we were dealing with a frictionless pulley and a frictionless belt the tension along the full length of the belt be the same and any force on the end effector would transmitted along the entire length of the belt. Timing belts and pulleys are not frictionless infact for our purposes we can take them to have infinite friction. How can we make this statement? If a force is large enough to overcome the teeth of the belt and pulley we have lost control of the system so we can consider that to out of our design scope. The same is true for any force large enough to force us to lose a step, any force that can cause a lost step renders the system out of control thus we are again out of design scope.
So we have now set the boundaries for our problem let's apply a force to our X carriage:
We can now see how the force is transmitted along the belt back to the first driven pulley it encounters.
Now lets take a look at a stock coreXY config:
You see how the force on the blue belt is transmitted along a much longer belt path in this configuration than in the 4 motor configuration.
Strain is defined at the change length over the total length of a member under load, so it is best to think of it as a percentage. Most timing belts have about a 2% strain before failing so for every 100mm of belt you can see up to 2mm of stretch in the belt, you will be taking most of the stretch out of the belt when tensioning it but there is still a good bit present. By using 4 stepper motors we cut the amount that the blue belt able to stretch about 2/3 rds! (assuming a square/symmetric XY travel) I have not run the exact math but this will certainly reduce the effective belt stretch of a coreXY printer. I would guestimate a 4 motor coreXY system can accelerate between 25% to 50% faster than the same system with only 2 motors; holding end effector deflection constant ie. the same print quality.
Yes the first video was my first XY gantry, it was not great and required a correction table printer correctly. The belts on my current design are parallel.
All this being said CoreXY is not really suited for large printers due the the amount of belt involved even with a 4 Motor setup.
Never stop making -Frank
Edited 4 time(s). Last edit at 08/30/2016 03:16PM by FrankPropaneTank.
Sorry, only registered users may post in this forum.