Concerns

[Skeen]

I have a few semi technical questions, regarding the design;

1. On printers using timing belts alike the GT2 belt, we sometimes have the issue of backlash. Using a rack and pinion approach to the X,Y axis must incur a lot of backlash, by the mechanical nature of the gears? - I also imagine the backlash is going to increase, as the gears wear?

2. How is rapid acceleration and deacceleration to be handled, without skipping teeth on the spur gear? (On the x,y axis) - I fear that printing operation speed will be quite limited by the acceleration limits, which follow from this design.

3. Moving the entire printing bed, which is undoubtedly a lot heavier than the hotend, will limit printing speed quite immensely? - On the 100mm/s video it seems like the printer overshoots quite a bit?

4. What is the printer displacement? - Possibly as a ratio, to the build volume, because the printing area seems super tiny, compares to the printer?

5. It seems like the axises are controlled via servos, rather than steppers, which kind of position feedback is used? - hall sensors? - Also why did you go with servos?

6. The two z-axis smooth rods, seems to be positioned very closely, why is this? - In terms of stability, wouldn't it make more sense to separate them?

7. I imagine the printer being very sensitive to vibrations, alike, as the bed is only held down by gravity. Have you done any research in terms of how much vibration the printer can handle before skipping steps and thereby ruining prints? - I would hope it's able to handle a kid bumping into the table, while printing?

8. Are the parts going to be consumer replaceable, alike with the reprap printers? - Also is the printer going to be extensible in terms of alternative heads? - Possibility of 2 hot ends, for say, PVA support?

I'm actually a little sceptical about the x,y drive. All attempts of the same thing, within the reprap community has failed, and your solutions does not seem to address the major inherent issues?

[NewMatter]

Lots of good questions:

1. On printers using timing belts alike the GT2 belt, we sometimes have the issue of backlash. Using a rack and pinion approach to the X,Y axis must incur a lot of backlash, by the mechanical nature of the gears? - I also imagine the backlash is going to increase, as the gears wear?

Because the table is gravity loaded against the pinion gears, the backlash is inherently removed. As the racks on the build platform wear, the table simply lowers by some minute amount. Eventually (that is, after many many hours of use) the build table may need replacing, but it is just an injection molded part that will be fairly inexpensive. (Having a spare table is probably handy to have around anyway.)

2. How is rapid acceleration and deacceleration to be handled, without skipping teeth on the spur gear? (On the x,y axis) - I fear that printing operation speed will be quite limited by the acceleration limits, which follow from this design.

We've tested accelerations up to 5000mm/s^2, print speeds up to 100mm/s and rapid moves up to 200mm/sec and skipping teeth was not a problem. You actually have to run into something pretty hard before you skip teeth. We've never seen an instance of teeth skipping during normal printing. Check out this video of one of the early prototypes: http://youtu.be/eqscOkfifoE to get a rough idea of the speeds and accelerations.

3. Moving the entire printing bed, which is undoubtedly a lot heavier than the hotend, will limit printing speed quite immensely? - On the 100mm/s video it seems like the printer overshoots quite a bit?

The print bed only weighs a few ounces, less than most direct-drive extruders. This was part of the motivation for the design - minimize the moving mass. There is no overshoot evident in the print quality. If you print a pretty heavy object (say 250g), then the mass does become more of an issue, but we've tested the build platform with adding quite a bit more mass and still achieve decent accelerations.

4. What is the printer displacement? - Possibly as a ratio, to the build volume, because the printing area seems super tiny, compares to the printer?

The print area is 150 x 100mm. With this design, by necessity, the footprint of the printer must be at least 2x the X and Y print dimensions. For a large print area, this becomes problematic, but for a smaller print area printer, this is manageable. For instance, the MOD-t is has only a 16% bigger footprint than the Makerbot Mini, but it has a 50% bigger build area.

5. It seems like the axises are controlled via servos, rather than steppers, which kind of position feedback is used? - hall sensors? - Also why did you go with servos?

The reason for using servos is that you never skip steps. Even if your motion profile exceeds the capabilities of the motors, you may get larger positioning errors, but you will never lose track of where you are altogether like with steppers. And the servos use high resolution optical encoders for position feedback.

6. The two z-axis smooth rods, seems to be positioned very closely, why is this? - In terms of stability, wouldn't it make more sense to separate them?

The smooth rods in the earlier prototypes were close together for minimizing the dimensions of the Z-axis carriage. They will be a little wider in the production units. If you get them too far apart, you run the risk of the entire Z carriage cocking. Also note that because you don't have the dynamic loads of an extruder moving back & forth on the Z carriage, the Z-axis structure does not have to be nearly as stiff as with a conventional printer design.

7. I imagine the printer being very sensitive to vibrations, alike, as the bed is only held down by gravity. Have you done any research in terms of how much vibration the printer can handle before skipping steps and thereby ruining prints? - I would hope it's able to handle a kid bumping into the table, while printing?

As mentioned above, it is actually quite difficult to dislodge the build platform. Whacking the printer with a hammer might do the trick, but a kid bumping into the table it sits on it won't be a problem (unless he knocks it over altogether!).

8. Are the parts going to be consumer replaceable, alike with the reprap printers? - Also is the printer going to be extensible in terms of alternative heads? - Possibility of 2 hot ends, for say, PVA support?

This printer is a commercial product with some complex injection molded parts. There may be a few things on it where you could print your own replacement parts, but many of the parts (like the build table) would be difficult to print yourself. Also, this is intended as an entry-level printer and we aren't planning on making it configurable.

9. I'm actually a little sceptical about the x,y drive. All attempts of the same thing, within the reprap community has failed, and your solutions does not seem to address the major inherent issues?

We shared your skepticism when we first came up with the X-Y pinion drive mechanism, but the simplicity and low cost prompted us to explore it further. The first prototype convinced us that it had some potential. Since then, we've been through 4 additional generations of prototypes and have done a lot of printing. We wouldn't have brought it this far if we weren't completely confident that it was a reliable design that produced good quality prints.

[Skeen]

Lots of good questions:

1. On printers using timing belts alike the GT2 belt, we sometimes have the issue of backlash. Using a rack and pinion approach to the X,Y axis must incur a lot of backlash, by the mechanical nature of the gears? - I also imagine the backlash is going to increase, as the gears wear?

Because the table is gravity loaded against the pinion gears, the backlash is inherently removed. As the racks on the build platform wear, the table simply lowers by some minute amount. Eventually (that is, after many many hours of use) the build table may need replacing, but it is just an injection molded part that will be fairly inexpensive. (Having a spare table is probably handy to have around anyway.)

I imagine a certain amount of backlash will always exist, due to the tolerances, in the interlocking between the teeth profiles of the to medias. This is even a slight problem for the threaded rods on the z axis of most printers.

Just for comparison, to timing belt driven systems; how much can one forcibly move the table, when holding the axises in place? - That is how many microns backlash is there; teeth to teeth.

2. How is rapid acceleration and deacceleration to be handled, without skipping teeth on the spur gear? (On the x,y axis) - I fear that printing operation speed will be quite limited by the acceleration limits, which follow from this design.

We've tested accelerations up to 5000mm/s^2, print speeds up to 100mm/s and rapid moves up to 200mm/sec and skipping teeth was not a problem. You actually have to run into something pretty hard before you skip teeth. We've never seen an instance of teeth skipping during normal printing. Check out this video of one of the early prototypes: http://youtu.be/eqscOkfifoE to get a rough idea of the speeds and accelerations.

The thing I dislike about this video, is that you seem to be running from end-to-end. If the system is indeed not relying on end stops of sorts, then this doesn't matter, however I'd love to see the system running a spiral or something non regular.

That being said, the actual speed is of no interest, if precision isn't there, and I'm quite frankly not completely convinced about this system yet.

3. Moving the entire printing bed, which is undoubtedly a lot heavier than the hotend, will limit printing speed quite immensely? - On the 100mm/s video it seems like the printer overshoots quite a bit?

The print bed only weighs a few ounces, less than most direct-drive extruders. This was part of the motivation for the design - minimize the moving mass. There is no overshoot evident in the print quality. If you print a pretty heavy object (say 250g), then the mass does become more of an issue, but we've tested the build platform with adding quite a bit more mass and still achieve decent accelerations.

This is something that bugs me; You have a conflicting design goal here; You want to maximize the weight of the table, to reduce backlash, decrease teeth skipping, and ensure reliable printing. While on the other hand you'll want to minimize it, to ease printing speed, acceleration and deacceleration, and to reduce wear?

If you did indeed find a good trade of, then I'd to congratulate you, it must have been quite some work.

What is decent? - 2500mm/s^2?

[Skeen]

4. What is the printer displacement? - Possibly as a ratio, to the build volume, because the printing area seems super tiny, compares to the printer?

The print area is 150 x 100mm. With this design, by necessity, the footprint of the printer must be at least 2x the X and Y print dimensions. For a large print area, this becomes problematic, but for a smaller print area printer, this is manageable. For instance, the MOD-t is has only a 16% bigger footprint than the Makerbot Mini, but it has a 50% bigger build area.

5. It seems like the axises are controlled via servos, rather than steppers, which kind of position feedback is used? - hall sensors? - Also why did you go with servos?

The reason for using servos is that you never skip steps. Even if your motion profile exceeds the capabilities of the motors, you may get larger positioning errors, but you will never lose track of where you are altogether like with steppers. And the servos use high resolution optical encoders for position feedback.

6. The two z-axis smooth rods, seems to be positioned very closely, why is this? - In terms of stability, wouldn't it make more sense to separate them?

The smooth rods in the earlier prototypes were close together for minimizing the dimensions of the Z-axis carriage. They will be a little wider in the production units. If you get them too far apart, you run the risk of the entire Z carriage cocking. Also note that because you don't have the dynamic loads of an extruder moving back & forth on the Z carriage, the Z-axis structure does not have to be nearly as stiff as with a conventional printer design.

To be fair, I was indeed inspecting a picture of an earlier prototype, when I made a comment on the distance of the z-axis smooth rods. It seems fine on the actual release candidate.

7. I imagine the printer being very sensitive to vibrations, alike, as the bed is only held down by gravity. Have you done any research in terms of how much vibration the printer can handle before skipping steps and thereby ruining prints? - I would hope it's able to handle a kid bumping into the table, while printing?

As mentioned above, it is actually quite difficult to dislodge the build platform. Whacking the printer with a hammer might do the trick, but a kid bumping into the table it sits on it won't be a problem (unless he knocks it over altogether!).

Again a place, where you'd want a heavy table, but assuming you found a good tradeoff, then that's nice. - How about moving the printer during operation, will this be possible?

[Skeen]

8. Are the parts going to be consumer replaceable, alike with the reprap printers? - Also is the printer going to be extensible in terms of alternative heads? - Possibility of 2 hot ends, for say, PVA support?

This printer is a commercial product with some complex injection molded parts. There may be a few things on it where you could print your own replacement parts, but many of the parts (like the build table) would be difficult to print yourself. Also, this is intended as an entry-level printer and we aren't planning on making it configurable.

I think it's a bit of a shame, to be giving up on the self replication philosophy. And having a 3D printer motivates one, to create replacement parts on it. - Out of curiosity, which material is the table made from?

Also it's quite a shame, that it isn't 'hackable' / configurable. Is the files/software going to be open sourced?

9. I'm actually a little sceptical about the x,y drive. All attempts of the same thing, within the reprap community has failed, and your solutions does not seem to address the major inherent issues?

We shared your skepticism when we first came up with the X-Y pinion drive mechanism, but the simplicity and low cost prompted us to explore it further. The first prototype convinced us that it had some potential. Since then, we've been through 4 additional generations of prototypes and have done a lot of printing. We wouldn't have brought it this far if we weren't completely confident that it was a reliable design that produced good quality prints.

Will you share the experience gained in each iteration, and maybe reflect on how the design has evolved?

[Feign]

I think it's a bit of a shame, to be giving up on the self replication philosophy. And having a 3D printer motivates one, to create replacement parts on it.

I kind of agree with this, but I think that self-replectaion has it's place, but this isn't it. I'm pretty sure NewMatter didn't come out of the RepRap community, they're showing what can be done when self replecation isn't even a consideration.

And I'm pretty well convinced that the RepRap project has a lot of room to grow with regards to true self-replecation, and when it gets there, I'm pretty sure it will look completely different from what most people expect. (personally, I think the RepRap GUS Simpson is the most likely candidate for getting to 100% self replecation... But that's just me.)

[NewMatter]

Regarding build table weight and skipping teeth - this is a mass independent issue. A heavier table will require more force to make it skip, but you also end up with higher acceleration forces trying to make it skip. Increasing or decreasing the table mass has no affect on the peak accelerations you can achieve. The primary factor in determining skipping is the pressure angle of the gear teeth (we use 20 deg. pressure angle pinion gears). With a 20 deg. pressure angle the maximum theoretical acceleration is well over 1G (9807 mm/s^2).

Regarding backlash, small errors in the teeth profile will create some lost motion (not actually backlash because the teeth are always in contact). In practice, though, these errors are small compared to all of the other mechanical errors of bearing play, mechanical stiffness, motor positioning errors, etc, etc. Most discussions of accuracy often miss the major sources of errors. For instance, people talk about using 16x or 32x microstepping and think they are getting 0.01mm or 0.005mm position accuracy, where in fact steppers running at speed or under a load are still only accurate to +/- 1 full step -- the equivalent of 0.15mm accuracy. Structural deformations under acceleration loading further degrade the accuracy, but these sorts of dynamic positioning errors are very difficult to measure directly. At the end of the day, what you really have to look at is the quality of the print.

[NewMatter]

Regarding self-replication, this is an interesting field of exploration (I come from robotics where people have been fascinated with self-assembling robots for decades). However, there's a whole world of things out there to print other than 3D printer parts for your own printer. The MOD-t is designed for people who aren't necessarily interested in 3D printers per se, or even 3D printing per se, but are more interested in having the capability of making their own customizable stuff at home.

[AndrewK]

Regarding build table weight and skipping teeth - this is a mass independent issue. A heavier table will require more force to make it skip, but you also end up with higher acceleration forces trying to make it skip. Increasing or decreasing the table mass has no affect on the peak accelerations you can achieve. The primary factor in determining skipping is the pressure angle of the gear teeth (we use 20 deg. pressure angle pinion gears). With a 20 deg. pressure angle the maximum theoretical acceleration is well over 1G (9807 mm/s^2).

I wouldn't have even considered this an issue. I love the idea behind the new technology that will drive the MOD-t. It's brilliant. I'd like to know which one of the New Matter people came up with the idea.

[NewMatter]

I'd like to know which one of the New Matter people came up with the idea.

Steve Schell and I came up with the idea. (My name's Jeff Kerr.) I've worked with Idealab on various projects for quite a few years, and I'm one of the many support people helping to bring the MOD-t to life. I've also been helping out answering questions on this forum.