What Delta Variation?

Hi, first time posting but I have been following this forum for a long time. I am thinking of building my 4th printer and I want it to be a delta bot. I am stuck between 2 models I like the Rostock Max for it size that thing looks sweet a 310mm printing bed sounds great but I do not like the laser cut approach on the other hand I like the Kossel because of it's construction with aluminum extrusions but it is small, is there a machine with the best of both worlds? Suggestions for a delta bot variant? Thank you.

I was considering the same and ended up building a Kossel mini with 2020 extrusions, 360mm horizontal, 900mm vertical and 330mm diagonal arms.
That gives me a printable range of ~280x450mm.

I'm very happy with the results I get running Rich Cattell's Marlin with auto calibration.
Have a look around; there are many upgrades/improvements available already for KM, just pick, download and print the ones you like.

Edited 1 time(s). Last edit at 09/13/2014 05:56AM by forerunnert.

Whatever you do, don't use 15x15 extrusion, it's far too small for scaling to the size you want and you will end up with a wet noodle for a frame.
There are some upscaled Kossel parts out there as Forerunnert said.

The SeeMeCNC is actually nicer in person than it looks in pictures (nice guys too), but I too dislike the wood (the smell makes me ill).

Along with the Kossel, is the Wolfstock designed specifically as a large delta, and the Griffin, which was designed specifically as a scaleable delta. There are carry on luggage size Griffins all the way up to 1meter tall Griffins, with larger ones being built, though I don't really recommend it. If you want to go over 1000/500 (h/w), I would recommend a Wolfstock and it's 20x40 for a bit more rigidity. The stiffer your frame the better.

Even if you don't build a Griffin, there are a lot of resources there for those building and designing large deltas. From design considerations to available glass, heaters and diagonal lengths as well as some parts you can use on other 20x20 designs, such as extruder and electronics mounts.

Thank you so much for all the information guys, I was actually thinking on using 15x15 however I did notice that the seemecnc uses 1" extrusions, and now sheepdog43 just confirmed what I was thinking, I need something more rigid.
I think the Rostock Max size is perfect. One thing that I do want to use is HIWIN MGN-12H rails instead of the cheapskate bearings setup of the Rostock. I will check your recommendations and see where that take me.
I been looking for an extruder, have anyone had any experiences with this [www.thingiverse.com] I thought it was just a cool looking extruder but when I checked it closely the herringbone gears are smaller than standard gears, the big gear is slightly larger than the motor making it small it looks very well designed, it is in 1.75mm and I am leaning towards 3mm, I messaged the maker he/she said that quite a few people have asked him for a 3mm version and that he is working on it, hope he releases it soon.
there is about 5 extrudes under this person I would like to hear from anyone that has tried his work. Thanks

Don't use an extruder based on the pretty pictures, or even how many extruders they have designed. Not only could you could say they just haven't yet figured out how to do it right, but you also don't know how many extruders others designed and never released

3mm is not optimal on a delta, the extra thick rod doesn't flex as well, and on a delta, you need that. It also weighs more. It can be done, it's just not as good.

As for extruder, I've run geared, I've run direct... Nothing, and I mean nothing, is as simple as a geared extruder. No muss, no fuss. Plastic gears, especially herringbone are difficult to do, and if you get even a slightly warped herringbone, the side loads can actually destroy not only the extruder, but the bearings as well. I have a graveyard of them.

Geared can nip your fingers (or worse), gears wear out and strip. Direct has no torque and large steps.

Thank you Sheepdog43, I'll stick to 1.75mm for my Delta, I agree with you 100% geared extruders are the way to go much more reliable than direct, I have tried many extruders they always seem to work for me I have never had a problem printing gears I guess I am lucky, I do have a greaveyard of extruders but not becouse they do not work but becouse I am always looking for something better.
I did download and completed printing the parts for the Infill 3D extruder actually turned out to be good, I am still tripping out on the gears they are small everything is moving along well, now I just have to make a hobbed bolt and I can get to test it

Oops, I meant nothing was as easy as a geared stepper motor. All the power of a geared extruder, with the simplicity of direct drive.

Geared extruders look great, and on a cartesian are fine, on a delta, they can look clumsy and out of place, but the gear ratio is the best of the three systems. We solved that on the Griffin by using a larger hobb to dial back the gear reduction (it rocks).

I was doing ABS and PLA mixed gears (ABS pinion for heat), and it was a nightmare, some people have better luck than others. I still have a direct drive sitting as spare, as well as a geared extruder.

I am with you on geared extruders beaing big and they do look clumsy, that is why I liked the Infill3D geared extruder the gears are small is hard to see it in the pictures but once you hold it in your hand you can really see the difference in size to a standard geared extruder, it is not big and defintely does not look clumsy. I have not had the chance to make the hobbed bolt once is made I'll test the extruder and let you folks know how it performs.

Hello sheepdog43, what do you think about aluminum extrutions 25x25, do you think I should go bigger?

You don't really need 25x25 until you hit about 1000mm or bigger.

Before you do that, I would go 20x40 V-slot or Open Rail or similar with a built in track system, it just saves so much hassle and money.

Just a point about geared extruders. The slop in gears gets multiplied due to bowden setup. I would invest the cash on a set of pulleys and belt and build the extruder around those. You get nearly the same lash as direct drive. Also do not go crazy with the ratios, 2:1 or 3:1 seem to be the sweet spot otherwise you will be cranking the motors at high rpm.

I am looking at buying two sets of pulleys in the 20T and 40T flavor for a gt2 closed loop belt.

EDIT: forgot to mention belts with the pulleys.

Edited 1 time(s). Last edit at 09/22/2014 04:14PM by jaguarking11.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

Slop in the extruder isn't that big of a deal and is easily accounted for (add a little extra to retraction, done), what do you think we do for bowdens.
Slop in your axis drivetrain, is a different story.

Your hot end buffers imperfections, which is why you don't see the pulses as the motor pushes filament. A bowden can also buffer those pulses.
More importantly though, have you ever looked at how "perfect" a geared extruder (not stepper) actually is, since you use a bolt and not a precision ground shaft, the bearings, pulleys and hobb are almost never a perfect fit. This only gets worse if you print your gears/pulleys. My geared extruders always had more slop than my geared steppers ever thought of having, not to mention were a lot more trouble.

I do agree with you about the ratios though. 5:1 isn't great (retraction that suffers a tad) and 1:1 is terrible (too little control and too little power). That was why we built our hobb/extruder the way we did, it brings the 5:1 ratio back to a 3:1 equivalent. We get the simplicity without the drawbacks.

Because you can account for slop its still not ideal. You can can only go so far with software correction. The other point is that when you have gears that are constantly beating themselves to death your reducing lifespan and increasing user maintenance.

If you collect the following parts you can build a nearly slop free extruder.

1 - main gt2 pulley 20T (go for a 16T or a 12T version to increase torque)
2 - secondary gt2 pulley 40T
3 - 2x 5mm id bearings
4 - precision ground 5mm shaft.
5 - 1 precision direct drive gear ( Pick your flavor, I opted for a 8 dollar SS built piece with steel set screw)
6 - one appropriate closed loop belt.
7 - one rollerskate bearing
8 - misc hardware.

When designing the extruder make the motor mounts slotted to account for some adjustability and belt stretch.

All of the above parts can be bought for not allot of money from hobby shops and or online stores. There will be very little if any slop, and the extruder should function properly for hundreds and hundreds of hours. Coarse gears will beat themselves to death and or warp when the stepper heats up.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

If your gears are beating themselves to death, you're doing something very, very wrong.

The amount of slop in a planetary gearset is ridiculously small, and even after 9 months of heavy use hasn't changed a bit. I would be really surprised if your slop was significantly lower (considering the number of parts, and how crappy many most Gt2 pulleys are), but even if there is, it's not enough to make one bit of difference in print quality. I've had geared extruders with lots of slop, I've had direct drive extruders with zero, in the end there was no difference in print quality due to slop.

And actually you can go as far as you want with software correction in this case, since it only occurs on direction change, I.E. retraction, which doesn't require absolute positioning.

I'm not saying your extruder is bad, or even some of your reasoning, but you are spreading F.U.D.

I was talking about 3d printed gears. I was not talking about gearboxes, those have their own slew of problems as well. Like some ridiculous gear ratios in the 5:1 to 40:1 range which in my book is way to high. One thing about planetary gearboxes is that they are durable as heck but not really very willing to change direction. As for gt2 pulleys being crappy. They do not matter much as the belt tension compensates for it fairly well.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

I've run all types of extruder drive systems (other than belt drive) and a lot of extruders, and like I said, there was no difference in print quality.

And change of direction being tough... You're kidding right? Those little gears do not even come close to the mount of torque required to re-direct a delta effector or cartesian carriage, particularly at high speed. Extruders run at such a slow speed compared to the printer that it would never get enough momentum to cause this difficult change in direction that you speak of. If the motor couldn't even handle a low speed direction change, it wouldn't have anywhere near the torque needed to extrude filament.

Again, I agree with you to some extent on the ratios (and printed gear wear), but the rest is just ridiculous.

Personally I would not use a gearbox when two pulleys can do the same job. That is my point on the matter. There is more than one way to skin a cat.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11