What features do you like on other Deltas?

[Jimustanguitar]

I've been scheming about building my own delta for a while, and I've finally started putting down some of my concepts on NapkinCAD. I've only really gotten to see and inspect SeeMeCNC deltas in person though. What other designs and features do you guys like that I might find some useful design inspiration from?

I've owned an original Rostock Max and a V2, plus I've seen and used a bunch of Orions and have gotten to oogle the Eris up close as well. I know there are a ton of other creative designs out there, what are some of the best?


*edit*
I almost forgot this subtopics posting rule.... Sharks with frickin' laser beams!

[Nylocke]

Frickin lasers are cool, but that's not as necessary. :p

Apart of the lasers, a quick change mount is awesome. Dial indicator, different hotends, lasers, etc.

Traxxas and CF arms are the best things since sliced bread.

Beefed up towers (some are using 2 wide extrusions) are a plus.

Spacing the arms out a little extra (like Jfettig's effector/carriage mounts) helps with rigidity/helps keep the effector more level across the whole area. Also slightly longer arms helps as well (cuts down on build height though)

Easy alignment of the towers/fully constraining the extrusions in the base/top is good.

[626Pilot]

I feel SeeMeCNC's mounting for the towers is a weak part of their design. If you make your own delta, constrain the towers as well as possible, top and bottom. They should sit on a flat surface on the bottom, and the top plate should sit on top of them. All four sides of each tower should be fully constrained so that the tendency to lean is minimized. It should not be open on one side, or have the freedom for each tower to be a different height, as is the case with the Rostock MAX. Any sort of looseness about the areas where the towers are secured is problematic, so spend some time on designing that.

I think it would be ideal to fix the towers to four plates, two closely-spaced ones up top and another two on the bottom, ideally stacked and cut at the same time on a CNC router so that their profiles are identical. Each plate would be stood off from its associated plate using spacers, so you'd know for sure that they were set up right. Why two instead of four? Same reason the delta platform has two arms per axis instead of one. It just adds another layer of constraints, so there's less possibility for the towers to get squirrely.

I would use a Smoothieboard or a clone. Price is the same or less compared to a RAMBo, and the firmware does more things, including automatic delta calibration. Marlin and Repetier have basic bed leveling and calibration, but it's nowhere near as good. I would recommend the Smoothieboard 3/4/5X over the Azteeg X5 Mini, as it has big, beefy power FETs for the heated bed and hot ends. I blew the bed heat FET on an Azteeg X5 Mini, which is just a weak, tiny, surface mount part. That's no way to do a bed heater!

Put the bed heater on wood, and rest the wood on FSRs, one near each tower. That way, you can calibrate without needing a separate Z-probe, and it will be accurate to the exact position of the nozzle.

Consider using a fishing line drive. They can be a little fiddly, but the cost per inch on fishing line is perhaps 1% of what it is for GT2, and it gives you options for how the motors are mounted that you wouldn't have otherwise. The link goes to what I think is the best implementation, with the line being passed between the pulley on the stepper and another pulley close by. This provides enough friction to keep the line from slipping. The other (easier) design is to have the thread go to a single spool on the stepper, but that lets it wander up and down the spool as the carriage moves up and down, so you don't get constant tension. In either case, the SeeMeCNC tensioning solution (pulley mounted to slots in the tower tops) is a good way to tension the line - and it will need tensioning!

Use an E3D, and put three of SeeMeCNC's tiny squirrel cage blowers in fan ducts for part cooling. In fact, I would say to just use the SeeMe effector, since you'd have instant compatibility with dozens of existing mods. You'd also be able to use Trick Laser arms out of the box. STLs are out there, so you can just 3D print the effector if you don't want to buy it from them.

For my effector wiring, I used cable made specifically for robotics applications, like this. It's designed for continuous flex, is shielded, and can be had with many conductors, so there is just medium-thickness cable running up there, rather than a bunch of skinny ones and however many zip ties. This stuff is more expensive than speaker wire, but you can have 12 conductors in one SHIELDED cable - enough for hot end heat & thermistor, +5V, +12V, GND, +5V PWM (for part cooling fan), and a Z-probe signal return wire, all in one. That gives you enough to connect all the accessories you need.

Use some nice plugs for wiring up the hot end and other stuff on the effector. I like TE Connectivity (plug, cap, pins).

Take cable management into consideration from day one. Design a mast that you can run the effector and stepper wiring up. Could be a half-height extrusion sticking off the side, or a 3D printed piece that force-fits onto the structural plate below the print surface. You could also mount your extruder to it. It's important to get the extruder as close to the hot end as possible, and that would help. Alternately, you can use Trick Laser's Fly-N-Strude mod, or my TL arm mount mod, link in sig.

It's tempting to mount the filament spool up top, and if I was designing a Delta, I might do the same. But, if you don't mind wasting some floor space, consider hanging the filament off the side of the bottom of the printer. This makes the printer less top-heavy. It also relieves the top screws and locknuts from the weight of the filament, as well as the mechanical forces of it clunking and thumping as the spool slowly turns. If you do mount the filament spool up top, make sure to beef up the top support plates and tower mounting.

[KAS]

This would be a great opportunity to create a OneNote Notebook and share it with everyone to scheme up different idea's from around the web.

[Batteau62]

I think 626Pilot has hit most of the list. The one thing that's been on my mind lately is an enclosure. I think once you nail a frame design down, you have to consider an enclosure. I have a V1 with the original side extruder mount. It severely hampers designing a "clean" enclosure of the print area. Top mounting solves a lot of problems, but you still have to consider wiring, bowden tube(*or not), heat source & venting/fans in the print area, patents and so on.

*(or not) extruder placement- most agree that the shortening or elimination of the bowden is better. While I like the fly-n-struder concept, I think a direct drive right on the effector would be optimal, hard on a delta though. It's a tough problem to design something lightweight enough to do the job. Especially if you start to think about dual extrusion. All the mods seem to add mass to the effector. I like the "tilting" hot ends concept(can't find the link ) but that adds a servo.

edit: found it- http://3dprintingindustrynews.com/spide ... erfection/

[KAS]

Any benefit to designing a delta around three vertical steppers with lead screws similar to this? Granted you'd need to find a quality shaft with anti-backlash, or maybe that wouldn't even matter with the hanging weight pulling down?

http://www.ebay.com/itm/NEMA-17-Linear- ... 4190253607

[Nylocke]

I've been using Pilot's EZStruder mounts exclusively since I installed them. They work perfectly, I love them.

[JFettig]

The only one thing I don't like about 626's suggestions is the fishing line. Its a precision machine, not a food fetching device, GT2 belts and pulleys are CHEAP. One of the biggest downfalls of printers is precision and accuracy, especially deltas.

I'm also designing a printer for myself but I'm not skimping. I'll be using THK rails, which surplus rails are not much more expensive than round rail.

[Nylocke]

Are you going to post your results here Jon? Your work is great and I'd love to see what you come up with.

[bvandiepenbos]

Frame should be all aluminum.
No laser cut wood or plastic printed parts.

[626Pilot]

Frame should be all aluminum.
No laser cut wood or plastic printed parts.

This is good, and what I would do myself. I have a CNC mill, but no laser cutter. Obviously, plastic printed parts would be iffy. What are your reasons for not liking laser cut wood?

[Jimustanguitar]

I feel SeeMeCNC's mounting for the towers is a weak part of their design. If you make your own delta, constrain the towers as well as possible, top and bottom. They should sit on a flat surface on the bottom, and the top plate should sit on top of them. All four sides of each tower should be fully constrained so that the tendency to lean is minimized. It should not be open on one side, or have the freedom for each tower to be a different height, as is the case with the Rostock MAX. Any sort of looseness about the areas where the towers are secured is problematic, so spend some time on designing that.

I think it would be ideal to fix the towers to four plates, two closely-spaced ones up top and another two on the bottom, ideally stacked and cut at the same time on a CNC router so that their profiles are identical. Each plate would be stood off from its associated plate using spacers, so you'd know for sure that they were set up right. Why two instead of four? Same reason the delta platform has two arms per axis instead of one. It just adds another layer of constraints, so there's less possibility for the towers to get squirrely.

Have you checked out the design on the V2 very closely yet? I believe that the base at least does what you're talking about (the top is still funky, and something that I'd like to change). Each tower fits snugly through a 4-sided hole on the top of the base, screws between 2 parallel plates that go full height from top to bottom, and there are "stop screws" to make each tower's height consistent.

When you mention having the extrusion sit on a flat surface (and the same on the top), how would that work? Wouldn't that interfere with the belts and wiring for each tower, or would you cut out notches and holes for those things to pass through?


Great ideas, thank you very much for sharing your thoughts!

[bvandiepenbos]

Frame should be all aluminum.
No laser cut wood or plastic printed parts.

This is good, and what I would do myself. I have a CNC mill, but no laser cutter. Obviously, plastic printed parts would be iffy. What are your reasons for not liking laser cut wood?

Laser cut wood can be OK with the right design, BUT never as precision, stable and sturdy as machined aluminum.
Also laser cut wood designs are more complicated since you have to overbuild it to make a decent frame.
Machined aluminum is so much more efficient and elegant IMHO

[Tincho85]

+1 for the enclosure. The best upgrade to my Rostock so far.
I like the idea of getting the extruder as close to the hot end as possible. But a heated chamber of +/- 60ºc with a stepper motor in the same place is not a good combo.
I've read that it would be dangerous to run steppers in environments with ambient temperature above 50ºc

A solution would be water cooling. Like this one maybe:
[img]http://i743.photobucket.com/albums/xx78 ... b29991.jpg[/img]
Yup, that's a V5 water cooled (http://forum.e3d-online.com/viewtopic.php?f=7&t=66)

[626Pilot]

When you mention having the extrusion sit on a flat surface (and the same on the top), how would that work? Wouldn't that interfere with the belts and wiring for each tower, or would you cut out notches and holes for those things to pass through?

The top plate on my DeltaPrintr is solid where it sits on the towers. Notches are cut next to where the towers are for the fishing line to come through. All you have to do is poke a hole.

A solution would be water cooling. Like this one maybe:

You would need some seriously beefy arms to handle all that, I think, especially with the waterjacket and the hoses. I would try a Peltier cooler with a fan-cooled heat sink before I tried a waterjacket. With some thermal-conductive grease, it should keep the steppers well below ambient. Way less to leak or break, as well, and you wouldn't have to worry about stuff growing in the tubes if you don't use it for awhile.

[Nylocke]

I think he was saying if the extruder was stationary rather than moving. Water cooling on the extruders and a hotend would be a feat with your extruder mount design..

[626Pilot]

Oh yeah, I forgot most people are using static mounted extruders

[jesse]

Auto bed leveling on the Folgertech Kossel 2020

[BobH]

I haven't see anything better than a BerryBot at this point. That's what I would build if I had the time a tools.

[geneb]

Auto bed leveling on the Folgertech Kossel 2020

That's easy. It's auto-calibration that's damn hard.

626Pilot is the only person I've seen that has come close or even achieved good auto-calibration in a delta.

g.

[Jimustanguitar]

626Pilot is the only person I've seen that has come close or even achieved good auto-calibration in a delta.

MHackney's got a nice solution too. It's a python script that plugs into KissSlicer and post-processes your GCode to match your measured positioning inaccuracies.

[JFettig]

Are you going to post your results here Jon? Your work is great and I'd love to see what you come up with.

Yes absolutely. It'll be an enclosed delta printer made of 80/20, printed parts, and laser cut melamene(need to source). Capable of being heated.

[Polygonhell]

The biggest win in the kit space would be self alignment during assembly. But it's somewhat tricky to do with Aluminum extrusions because even with the best design the extrusions tend to crush/bend when fasteners are over tightened, which I would guess a large portion of people purchasing any kit would be likely to do.

I like the idea a quick change tool heads, but it's not that much of a win most of the time.
Depth Probes are a nice to have, but again they aren't something you end up using on a regular basis.