CoreXY

Yes I noticed that the heavy bed surely slows down the acceleration. I had to put something like 600mm/s2 but the x-axis works great with 1200mm/s2. I think that this can be the dealbreaker?

With small pieces and countless hole perimeters this slows my printer down, and offcourse theres this enclosing thing what is easier and better looking with corexy things...

I'm driving now with sav mk1 and in vase mode it clearly shows that it have to slow down. So controller must do great difference with this. I have mks sbase waiting to get to work and I think that its going to faster this.

/Offtopic

But man that vase mode looks wonderful even with 0.3mm layerheight.

Another thing that is with pla that my layerbond is great, parts are way stronger than with petg.

Easiest way to attach plate to rail is used countersunk holes and flathead screws. Word, metric stuff is typically 90 degree c-sinks.
Plate needs to be thick enough to accept the countersink, and have enough, yet not too much, screw length into the slot for a T-nut.
Considering most likely tap the plate for the rail bolts, it'll be at least 1x diameter of the screw diameter thick for a decent thread engagement. Anything more than 2x is a waste.

Comparing belt lenght and the resulting stretching/ringing, the CoreXY needs a light weight gantry and remote extruder to have an advantage over cartesian printers.
I modified my Prusa i3 with a Volcano 0.8mm nozzle. That's probably the best use case for slow_by_design printers.

Yeah right, offcourse it´s the easiest way. Facepalm smiley here!

As a designer, I didn´t realise that you could make threads to aluminium too... And to thinking about it, I can put thru holes to aluminium and t-slot nuts underneath...
I was thinking about to mill exact places to aluminium plate to hold the rails in correct position, and thru thruholes attach the plate and rails att the same time. This way it should be easy to have everything in correct places.

Now it turn to think about x-axis gantry, If I put rails on top of the extrusion, should the MGN12 be enough withous extrusion to hold the mass of hotend. How much does MGN12 rail hang if the span is max 400mm, if somebody know this allready... Offcourse rail is going to be vertically.
Whats the best way to attach rail to bearing vertically, Is DD:s structure great? I was thinking about I-beam piece? No I wont put another bearing to another end...

PS. Is there a way to edit the title of this topic?

Edited 1 time(s). Last edit at 03/05/2018 03:41AM by kapperi.

I went vertical with the X axis guide rail because is was 24mm wide and I wanted to minimize sag over the span. I had those convenient square tubes on the Y axis bearing blocks to mount it on. I don't think it matters either way if the rail is square or almost square. Do whatever makes it easier to mount the extruder carriage.and the X axis rail.

I don't think you can mill exact width slots into the aluminum frame members to mount the linear guides unless you can be sure that the frame pieces are going to be exactly parallel. I milled a slot that was wider then the guide rail, assembled the Z axis frame, put the linear guides loosely into position, then used a metal block as a spacer between the rail and the flattest surface I had available- a pool table- and set the guide positions and tightened their mounting screws. It worked pretty well. We have since acquired a granite slab at the makerspace that I would use to do that if I had to do it again.

I am re-examining UMMD's X axis and extruder carriage designs and may make some changes in the near future.

Edited 2 time(s). Last edit at 03/05/2018 06:59AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

MGN12 is 12mm wide and 8mm tall, so there is also big difference wether its mountet vertically or horizontally.

If I waterjet plate what is same size than my width*length and hole in the middle what is 30mm smaller from sides. It's going to be base for the rails. I can mill it with same settings so it should be pretty accurate. I could then put holes in the corners to put 2020 profiles to make top cage. Does it make any sense?

This way the corexy mechanism is in own plate at the top of the printer and accuratly milled in the place. I too have flat surface at the work but sadly its pretty small.

DD did you use belts in z-axis cause it eliminates banding?

Now I can make DD jealous! I managed to salvage real HIWIN EGR20 bearings from work, and did not cost anything. And theres some rail too to salvage...

Is there any way to change topic name?

I used belts in the Z axis as an experiment, and because they don't produce lateral forces that result in Z wobble in the prints. I am very happy with the results.

The only real problem with belts is preventing bed-drop when Z motor power is cut. Some people try to fix that with counterweights or bungee cords with varying degrees of success, and it's always ugly. I used an elegant and relatively expensive ($140 all-in) 30:1 worm gear reducer on the motor with a keyed output shaft. I have designed a much less expensive 40:1 worm gear drive and am waiting for parts to arrive from China to test it. My new design will be smaller, and use "standard" parts (no keyed shaft!), and will cost about $40 for similar performance, assuming reasonable gear quality (we'll see...). The gearbox/motor mount will be printed, and will use a NEMA-17 size motor to deliver 100 full steps/mm resolution. I am expecting it to lift 5 kg without breaking a sweat.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

I have to think about that too.

I have thinked that I could put ballscrews to lift one 2020 profile what lifts the bed thru somekind of teflonplates. This way there cant be any wobble too. If ballscrews are constrained from bottom and top, it would not need any linear bearings.

Only the bed would be running on two smoothrods.

I really have to get my designs on here to show what I mean.

Xy mechanism is now rigid enough with egr20 rails and bearings. These are the real Hiwin products so I expect that these are good. Rails are ordered too, cost was about 100e for 3x400mm rails.

Others have tried using screws without guides and the result is never pretty.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
kapperi

Quote
the_digital_dentist
I wouldn't let my dog (if I had one) use 2020 to build a 3D printer. Go with the larger stuff.

Ball screws will probably allow the bed to drop when power to the Z motor is cut, if everything is aligned well.

I definitely would not mix bearing blocks and rails from different manufacturers. Linear guides are systems, not parts.

So do you think I go with the 3030 or is 2040, you build your megamaxsuperhyperturbo with 4040? So I think with much less Z, I could go with 3030.

Ball screws will allow the bed drop, but could I use geared Nema to stop that?

Edit: God tha PLA is so easy to print. If only it could come in better strength..

I suggest that you use 4040 that is consist of 2 2020 if you want it more rigid. A lot of hardwares sold are based on 2020 mounting.
MIsumi 40X40

Edited 1 time(s). Last edit at 03/12/2018 09:30PM by Vigilant.

Quote
the_digital_dentist
Others have tried using screws without guides and the result is never pretty.

Whats the problem? Only the lift is using ballscrews as a guides and the bed is constrained properly with bearings.

Quote
Vigilant
I suggest that you use 4040 that is consist of 2 2020 if you want it more rigid. A lot of hardwares sold are based on 2020 mounting.
MIsumi 40X40

Yes I think that i have to use 4040 because Egr20 bearing is 34mm 42mm wide.

It doesnt help to use 4040 because bearing is still wider, but I allready have 4040 profile, and man that it´s going to be ridig, even DD likes it.

Edited 2 time(s). Last edit at 03/13/2018 02:01AM by kapperi.

Quote
kapperi

Quote
the_digital_dentist
Others have tried using screws without guides and the result is never pretty.

Whats the problem? Only the lift is using ballscrews as a guides and the bed is constrained properly with bearings.

I don't understand what this means. Ball screws are not guides. The bed is attached to the "lift", so if the lift shifts laterally, the bed does too.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Theres exambles in I3 Prusa clones where screwnuts are constrained to the gantry only from top side of nut. And nut rotating of nut is prevented by arm to the smoothrod.

This time nut would be constained to bed gantry only from top and arm is replaced by 2020 profile from nut to nut.

By the way I managed to salvage two more egr20 bearings and 2 370mm rails, shamefully too short.

Now I have 5 bearings and 2 too short rails.

I think the problem is terminology. You're still going to use linear guides or guide rails on to limit motion to vertical. The screws/nuts are providing the lifting force. You're going to "float" the nuts on the platform but link them to the guide rails to prevent them from rotating- they'll be able to slide a little and accommodate bent screws without displacing the bed laterally.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Yes something like that, but I'm thinking to link nut to nut instead to rod. Cannot link to rod cause Im going to use rails.

Maybe thats useless with ballscrews and these big rails...

Do not use MGN9 not because they can't be used but their dimensions does not suit the 6mm gap on the extrusion. Use MGN12.
I made the same mistake.

I use robotdigg Stainless steel linear rails for my delta and it's ok.
It's cheaper than hiwin equivalents but more expensive than the typical chinese rails.
To me that is better value for money.

I'm using salvaged real Hiwin Egr20 bearings and matching rails. Only that rail costs 70e chopped per meter. Expensive...

Quote
Vigilant

I suggest that you use 4040 that is consist of 2 2020 if you want it more rigid. A lot of hardwares sold are based on 2020 mounting.
MIsumi 40X40

Neat profile, but I think that that consist 4 2020 afterall.

That would be great profile, Atleast that could be bolted with 4 bolts to thru other profile. Could make super heavyduty corners too, no need to put extra corners with that.

That is rhe most expensive profile what I have seen...

50 dollars per meter, that cannot be true...

Quote
kapperi
That is rhe most expensive profile what I have seen...

50 dollars per meter, that cannot be true...

Which profile are you looking at? HFS5-4040 is $15.40 per meter, and prices include cut to length at no additional cost with near flawless accuracy.

I checked similar profile from motedis, and I checked again and it was 36e this time... I cannot order from misumi cause Im not corporate user.

What do you think, should I make corner profiles so long that those works as a legs too? Is there any difference with vibrations etc.?

[Offtopic]

Today I was in the seminar where was 3 leading professors in Finland who studies 3d-printing. One professor studies using 3d-printer to create lenses. They can actually print lenses that is suitable for camera use also. Roughness was under 5nm and vertical repeatibilyty was something like 3um. Second professor is studying metalprintings and third one was one of the laserfysician in the 3DSystems. There was huge amount of knowhow in the same room!

Couple companies showed their business, other was doing mainly 3d-printing of plastic from powder and another one was printing metals. Powder printed plastic were not amy impressing thing, parts was very rough and moving mechanism are very difficult to print.

Wicked!!!

[/Offtopic]

Ok, now i found supplier for mine extrusions. Its 45x45 rexroth extrusion for the price of 12e/m sawed. I could have ordered cheaper but this shop is near me so no postage.

Should be rigid enough? Frame members costs about 65e now. Rails what I have to buy are about 140e.

Total budget must be under 500e, so theres maybe room for duet too.

Now I had time to design something, there´s a quick sketch what I´m thinking of.

Any thoughts in this point?

Edited 1 time(s). Last edit at 03/22/2018 08:59AM by kapperi.

Hi, add some bottom space for electronics I forgot this on mine :-(

I found this picture on the web close to your design with space bottom

I see that sort of frame design everywhere. If you need to add 64 corner braces to make your frame more rigid, maybe you need a more rigid frame, and not a bunch of corner braces.

Compare the cost of that frame plus all that hardware to the price of the same frame made from larger cross-section t-slot (which will be inherently more rigid). I think you'll find the larger t-slot cheaper, and you won't suffer the tedium of mounting all those ultimately useless corner braces.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Actually, I really dont need those corners to my design. Its 45x45 profile all bolted thru profiles with M12 bolts so it should be rigid enough.

Theres going to be a little twist in design. I really need to put pictures here someday.

Yes, the electronics will be under there shielded with aluminium plate and litlle fans to ensure cooling.

Yes too much corner are useless, I'm looking for a design with 4040 extrusion, actually my D-bot is on 2040 and shake a bit maybe upgrade to 4040