Good time of day, dear geeks and sympathizers!
This publication is a continuation of the description of the design of my self-made 3D printer. The Z axis is one of the most controversial printer nodes.
What to choose – ultimate accuracy or good scaling? Move X axis or printer desktop?
Two approaches – two solutions.
I could not look at the first 3D printers without a shudder: the designs were immature, many nodes were used in violation of the specifications, due to the general fluctuations, constant adjustment, minor repairs, and working area dimensions were small. I decided to solve the problem of internal contradictions, simply by crossing a hedgehog with the snake of a portal milling machine and a 3D printer.
Was designed and twisted together the skeleton of the 3D monster:
It consists of forty-millimeter aluminum system profiles connected by thick corners 45×45 and bolts M8. This design has dimensions of 60x40x40 cm and is absolutely unshakable during normal operation of the printer. The size of the working platform was 45×22 cm, the maximum height of the objects could be 28 cm.
The Z axis is formed by two vertically arranged 20 millimeter linear buses with precision carriages. The carriages are driven by precision trapezoidal screws with polymer nuts. Why the polymer: there are no high speeds / loads, and the polymer nut is not so demanding for lubrication and is much easier to install. In this construction, the height of the model is increased by raising the X axis above the table, and the Z axis is used as a movable support for the X axis.
The screws were installed on the angular contact bearings, each screw Has a drive from the stepper motor through the belt transmission 1: 3. The upper ends of the screws are turned and inserted into the needle bushings so that the axial displacement of the screw in the bushing prevents its jamming during thermal expansion. Perhaps this measure is excessive, but then, under the impression of reading a lot of materials on CNC machines, it seemed to me extremely necessary.
Video of the work:
This axis Worked without any problems until disassembly of this printer for spare parts.
Disadvantages of this decision:
1. The price. Precision components are expensive.
2. Complexity of construction
3. Poor scalability.
When I started building a second printer, experience and stinginess took part in creating the design along with an innate desire to go its own way, not expecting favors from nature. Accordingly, the new printer was to become not only simpler, faster, more universal, reliable and maintainable, but also much cheaper.
For the Z axis of the new printer, a non-screw drive was chosen, and a cable structure similar to the hoist of freight cranes. It consists of the actual drive mechanism with a belt reduction gear and two blocks on which the entire mass lies along the Z axis.
Here is a photo of the printer as a whole:
For the stability of the spatial orientation, there are two round ten-millimeter steel tires, along which the pair of plain bearing bronze sliding bearings move.
The following two videos show the construction of a winch, there is nothing complicated: there is nothing complicated there:
Power calculations: The drive drum has a radius of 10 Mm. Accordingly, a torque of 0.3 Nm (a normal Nema 17 motor) on the lever of 10 mm is 30 N. The belt reducer with a gear ratio of 2: 1 doubles this number.
Because of this, the maximum force that this winch can develop is about 60 Newton, respectively, the maximum mass of the Z axis, together with the masses of the table and the object, should not exceed 6 kg at rest.
Now we determine the losses for acceleration and deceleration of the Z axis: to accelerate 1kg of mass with an acceleration of 1 m / s², it is necessary to apply force to 1 Newton. In fact, acceleration of 1 m / s² for the Z axis is quite enough, and each accelerated kilogram will cost us 1 N of applied force.
The heaviest element in the design is a heated table, it is a duralumin plate 350x350x3 mm weighing 980 grams with glued heating elements weighing 150 grams.
The remaining elements of the construction, including the thermal insulation from Basotect, weigh about 900 grams.
The total weight of the structure is about 2030 grams, which, when rounded up, will require 21 N for holding and 2.1 N more for acceleration. Total, once again rounding to higher values, 24 N.
If you add a kilogram model to the Z-axis mass, you will need 34 N, which is more than half of the rated power of the winch. It would seem that the design is redundant in power. But the devil, as always, lies in the details. The fact is that maximum currents must flow through the motor windings to achieve maximum torque, which will inevitably cause it to overheat and premature failure.
For this reason, I designed the design with the condition of a large power reserve, and the motor current was established by the experimental way a little more than the minimum required. The motor was heated to 50-60 ° C, which is quite acceptable according to the specifications.
In this video, the winch easily juggles with the Z axis without a desktop, but with two filament coils weighing 1300 grams:
So, the issue of power is solved. Now let’s talk about accuracy. Taking into account the parameters of the winch and motor components, with microstepping 1/16, the Z axis can be moved in steps of 0.02 mm. Now consider the accuracy problem for a winch with single-layer winding of the cable. The radius of the drive drum in my design is 10 mm, respectively, the circumference of the winding will be 62.8 mm. In order to raise the Z-axis by 44 cm, you need about seven turns. When using a cable with a thickness of 1 mm, the axial displacement of the winding is 7 mm. In this case, the distance from the point of contact of the cable with the drum and the lower support block is changed.
Calculate how bad everything is: the leading drum is in the center of one of the diagonals of the square formed by the bottom of the printer body. Accordingly, the drum is removed from each of the lower support blocks by 320 mm. By simple calculations it turns out that when moving the table by 440 mm, the drive geometry will change by 0.077 mm.
Qualitative screws with trapezoidal thread ensure accuracy from 0,07 to 0,4 mm at such distances. Simply put, do not provide any meaningful gain in accuracy. If for some reason you need to print models taller than 44 cm, you just need to proportionally increase the diameter of the drive drum (to save the small number of revolutions required to move the axis) and gear ratio (to maintain the rated load on the motor). At the same time, the cost of the mechanism grows insignificantly, in contrast to the design with a screw drive.
One of the test objects:
In conclusion, I can say that although experience is Derivative of the overcome difficulties and mistakes made, but sometimes the process of acquiring it gives more pleasure than the results achieved.
There will be no 3D models, because I can not find them on the disk.
Published under the WTFPL license.
Well, and traditional: Have fun!
We go down – grow up, or the Z axis for inexpensively / Geektimes
