Why Rostock?

What advantages does the Rostock type design have, vs. the traditional XYZ axis setup. It's a pretty cool looking machine I can honestly say I've never seen that type of design in any machine. It's quite ingenious. Smart use of math for sure.

One advantage is that the bed does not move, so it can be as heavy as you need for the glass and heating element. Also, the effector is usually quite light allowing for high speed printing.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

as fast in Z as in X and Y, no lead screw related Z wobble issues (although it probably comes with its own artefacts). only 3 motors for movement instead of traditional prusa 4. more commonality of parts. much taller build volume (possibly at the expense of rigidity).

Low moving mass, higher apped/acceleration.
I was in two minds about the geometry, until I built a RostockMax, the layer consistency is stunningly good, better than either of my other two machines, which I think is in large part because you have 3 copies of the same mechanics. Still not convinced it's the only way forwards, but mine is getting more use than my other printers.
The software still needs work, I just swapped to the development build of Repetier and it's better than the current Marlin support.

I'd wager the layer consistency is due to not using Z screws. 2013 prediction: Z screws will start disappearing from new reprap designs.

Edited 1 time(s). Last edit at 01/13/2013 10:45PM by billyzelsnack.

there's a bit of misconception about threaded rod as leadscrews, they can actually be quite accurate, what you can do if you want serious accuracy is to lapp the threads and put in a springloaded nut and the backlash is peanuts, it also helps if you use decent threaded rod, most people use rod that is rubbish and bent and wonder why their x axis wobbles

billyzelsnack Wrote:
-------------------------------------------------------
> I'd wager the layer consistency is due to not
> using Z screws. 2013 prediction: Z screws will
> start disappearing from new reprap designs.

billyzelsnack Wrote:
-------------------------------------------------------
> I'd wager the layer consistency is due to not
> using Z screws. 2013 prediction: Z screws will
> start disappearing from new reprap designs.

Yes I tend to agree, I'd probably use belts on a ZAxis if I were designing something today.
Lead screws can be relatively wobble free, but they have to be straight and the axis constraints have to be much more robust than the screws.
My mendel max is using ACME lead screws and is relatively wobble free, but the Rostock max is still has better alignment.

Cool stuff. Part of the reason I'm asking this is because soon I'm getting enough motors and electronics for two more printers. My second build I have already decided will be nophead's Mendel90.
Trying to decide what #3 will be.
My ultimate goal for my machines is the pursuit of the finest quality prints possible.
The mendel90 seems promising for use as a high accuracy machine, of course with some add-ons and money spent on precision machined parts here and there.
But machine #3 I was thinking Ultimaker, as I have seen some really really impressive prints. But now after seeing the Rostock I'm thinking it may be nice to have a big, fast machine.
What would any of you pick for #3? Rostock, Ultimaker, other?

thejollygrimreaper Wrote:
-------------------------------------------------------
> there's a bit of misconception about threaded rod
> as leadscrews, they can actually be quite
> accurate, what you can do if you want serious
> accuracy is to lapp the threads and put in a
> springloaded nut and the backlash is peanuts, it
> also helps if you use decent threaded rod, most
> people use rod that is rubbish and bent and wonder
> why their x axis wobbles
Exactly, but in the CNC world they have been doing that for ages.

Don't have the money for an ACME screw setup, or a ball screw setup? Then use a thread rod with two nuts and a spring inbetween them with a harness holding the two nuts. The spring keeps the nuts up againt the edges of the thread so the backlash is pretty much gone and gone on the cheap.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

DA. It's not that simple.

billyzelsnack Wrote:
-------------------------------------------------------
> DA. It's not that simple.


Indeed if you look at how screws are used on mills, they are under tension using AC bearings, it's not really feasible todo that on a reprap,.

I know it isn't that simple I was not trying to over complicate it but I remember seeing it done about that simple somewhere early last year. Basically a poor mans anti-backlash nut.

edit: Something like this but on an even less costly way: [www.aquickcnc.com]

I know I saw it but I forgot what was used to hold the two nuts but it worked on the cheap and managed to be pretty accruate too.

Edited 1 time(s). Last edit at 01/14/2013 01:32AM by Dark Alchemist.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

... I've used an elastic tube between two nuts - one fixed to the slide, the other free: [forums.reprap.org]

And one of the big advantages of the Rostock is the small count of parts used - I've built my sample with only 12 srtuctural parts: [builders.reprap.org]

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Isn't a disadvantage of the Rostock design that it has less resolution toward the edges of the build plate?

Twitchy Wrote:
-------------------------------------------------------
> Isn't a disadvantage of the Rostock design that it
> has less resolution toward the edges of the build
> plate?

While true, it's only really a disadvantage if the resolution is too low near the center of the build plate, and that hasn't been my experience so far.

30mm per step on the inside and 300mm per step resolution on the outside.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

it's 30 steps/mm on the original design, but it depends on arm length relative to the machine radius and the radius of the drive pulley.
I believe the original uses a 20 tooth GT2 pulley.

Polygonhell Wrote:
-------------------------------------------------------
> it's 30 steps/mm on the original design, but it
> depends on arm length relative to the machine
> radius and the radius of the drive pulley.
> I believe the original uses a 20 tooth GT2 pulley.

Yep, but it will always have a higher resolution on the outside than on the inside as that is the nature of deltas.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

[OT]
If you needed a basis for the most accurate builds, what would you go with, mendelmax, mendel90, other?

I'm sold on Rostock I'm going to build one, but still not sure if I want a mendel90 or mendelmax(v2) or something else as my basis for a high accuracy unit...
[/OT]

As for as the Rostock now I'm looking into the variants. SeeMeCNCs Rostock Max looks nice, just wish I could get the hardware without the electronics/motors/PSU. (I'd source it myself but I'd still need the laser cut parts).

I see a couple of variations on it, the Max probably is most interesting to me. If I did do the Max, it'd be interesting to mod it to use linear rods for travel (should be quieter and less friction than bearings riding on extruded aluminum while retaining the rigidity of the Aluminum Frame), like MendelMax 2.0 is doing.

xclusive585: Yep, Rostock is so elegant it almost looks like a form of Tai-Chi when it prints. I personally think more will jump on the band wagon for it with a variant that does not use the Bowden extruder. People seem to hate the Bowden due to a lack of retraction but there is one variant that is a light weight extruder that extracts so there is hope.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

Well xclusive585, funny you would want a kit minus electronics, lol.

We were just adding a couple options for the Rostock MAX on our website. One kit is the BYOE kit (Bring Your Own Electronics), which takes the steppers, power supply and RAMBo out of it, but gets you everything else.

We also just added the BYOL kit (Bring Your Own Laser) for those who have access to a lasercutter, or want to customize it out of another material we don't offer yet.

John

John is a good guy and I am not surprised he is doing this.

Awesome work John.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

Dark Alchemist Wrote:
-------------------------------------------------------
> xclusive585: Yep, Rostock is so elegant it almost
> looks like a form of Tai-Chi when it prints. I
> personally think more will jump on the band wagon
> for it with a variant that does not use the Bowden
> extruder. People seem to hate the Bowden due to a
> lack of retraction but there is one variant that
> is a light weight extruder that extracts so there
> is hope.

... My J-head's retraction sucks for some reason (it shouldn't with a wades), doubt I'll mind the bowden. :-)

I take it JohnOly is SeeMeCNC?

[Edit] Yes I see he his. Was just chatting with him on IRC two days ago.
John, I will be doing this in <4 weeks now. I look forward to working with you. Good call on adding the BYOE options.

Edited 1 time(s). Last edit at 01/14/2013 11:49PM by xclusive585.

Polygonhell Wrote:
-------------------------------------------------------
> it's 30 steps/mm on the original design, but it
> depends on arm length relative to the machine
> radius and the radius of the drive pulley.
> I believe the original uses a 20 tooth GT2 pulley.


How many teeth on the pullies on your RoMax Polygon?

15 tooth GT2 pulleys, I believe See MeCNC has them made.

Would someone explain something to me please? Look at a Prusa Mendel I2 and look at a Rostock Delta (the first model) can you explain to me why the I2 is less costly to build? I don't get it because the Rostock has less parts and even one less motor but costs more to source.

I don't get it.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

I you know it costs more to source just look at the BOM and figure out why?

Polygonhell Wrote:
-------------------------------------------------------
> 15 tooth GT2 pulleys, I believe See MeCNC has them
> made.


TY

Im asking because I'm curious about maximising the resolution on the outer ends of the Rostock, hoping to at least get the resolution of a mendel at the rostock's weakest area- the outside...

I'm going to do some math this evening. I know the number of steps/time has a limit on our pololus. That coupled with the motors speed limits, mean at some point to get that resolution the speed will have to be really slow possibly so as not to push the pololus to hard, even if I get 400step/revolution motors, the speed of the pololus could be the limiting factor, or use smaller pulleys which again cuts the speed.

Hmmm

Polygonhell Wrote:
-------------------------------------------------------
> I you know it costs more to source just look at
> the BOM and figure out why?
A Rostock Delta has fewer parts now remember a delta is not the Kossel or the other variation that uses T Squares etc... It really should cost less but ends up costs way more.

---

_______
I await Skynet and my last vision will be of a RepRap self replicating the robots that is destroying the human race.

I'm afraid I don't quite get the benefit of the Rostock concept. I keep reading about the accuracies of all the various machines, and for the life of me I can't see how any of them get anywhere near the accuracies claimed. My problem is that I look at a machine that starts out with a very stiff triangulated frame, then suddenly all the moving parts are hanging on long thin rods. I can't help but think that there is a lot more deflection going on during operation than everyone realizes.

Some years ago I built packaging machines that heat welded poly tubing into poly bags, and then filled the bags with product and sealed it up. They had heavy steel frames and a carriage that reciprocated by the length of the bag being made. An air cylinder pushed it back and forth and hydraulic shock absorbers stopped it at each end. 45 bags minute, filled with parts. On one machine we added a thermal transfer printer and labeler. The printer was doing a bad job, the print was smeared. Clearly vibration and deflection were a problem. Everyone made suggestions about beefing up the printer mount, all in different areas.

I got out a dial indicator and placed it against the printer and pushed gently on it with my hand in the direction of the carriage motion. It moved quite a ways, and we wrote down that number. Then we moved the indicator back along the mounting path and pushed again with about the same force. We were able to determine that 90% of the motion came from one part of the mount, and we beefed that up. The printer then worked fine. Do this on your printers and I think you may be in for some surprises.

The Rostock looks to me like extreme flex from the long vertical axis'. I would think that linear rails pushing against heavy non-moving vertical structural members would greatly improve the stiffness and real accuracy. A dial indicator would easily prove this theory.