Advice for my printer design

hi guys,

I´m designing my next 3d printer, and i´m doing this just because it´s fun an i like it !

But i do have a question. I´m trying to go for a basic setup using dual extrusion (and trying to have an easy hotend replace system) so that i can have 2 nozzle sizes or two colors.

I´ve design it using a bowden extruder, with an adapted wade extruder, and i currently have a 200x200x200 printing area, but i do have a moving printing bed (xz head, y bed).

On my research i got to tho point of wondering if a bowden extruder is a viable choice for this type of machine?
I see huxley works this way, but i also found comments on the web saying bowden is for ultimaker style machines.

Can a bowden extruder use 3mm abs and pla in a viable way or should i only consider 1.75mm?

tanks, and what do you think of the look of the printer?

Having built and modified many extruders, but not the one you speak of, they can do anything, at any size. In particular, most nozzles have enough space to be modified within a range of .3 - 3.0 mm. Don't let the pundits of over-engineering be the enemy of good. In the end, whatever you do or make will work VERY well, it just won't be "perfect" in the eyes of some overly-complex engineering calculation. Take it from me, I have 2 engineering degrees, and none of the classic equations and constraints matter. Really open your eyes to this and you'll see the freedom of design no one has before. Its just a piping hot tube, and plastic flows through it like very thick honey. That's it in a nutshell, minus 1000 less important details.

Though one exception I will say, if the size of the exit hole is larger than the filament, then obviously you could push filament out without mixing properly if it is fed too fast. You want some backpressure.

Also, your system should be designed to take BOTH 3mm and 1.75mm. The prusa plastistruder, while finicky, melts 1.75mm PLA or ABS into a pool in the hot end about 4 mm in diameter. This means it would take 3mm. But you need a stronger stepper gearing I think, to make it work well, or you can turn up the temperature to decrease backpressure, a small sacrifice.

I think the printer look is fairly clean. It's a bit blocky (large, chunky), and has too many exterior screws, but that is my artistic opinion. It looks like a solid design, and has better symmetry than I'm used to seeing.

I think that bowden actually works better with 3mm because it is easier to push through a tube. I've never used one though, so take that with a grain of salt. People say that the main problem with bowden extruders is that reversal is difficult to properly control. That means that strings between 'islands' of printed material are fairly hard to avoid. Several production machines have bowden extruders that seem to work great though, so it's possible.

One thing that you may not have considered with your design (it's tough to tell from the one photo) is that with two extruder you need to have a wider x axis so that both extruders can print all of the way to the edge. The closer the two extruders are to each other, the easier that mod is to make. Also it looks like the hot ends are spaced pretty far from the rods. I would be afraid that the torque generated would bend the rods in both the z and the x dimension. Moving the rod closer to the exterior of the casing and shortening the arm that the extruders are on will probably result in tighter looking prints.

Other than that - it looks good. It looks a lot like the UP! printer.

well the printer current footprint is 400x500x355 (excluding the spools). Its current over the limits i set before and i will still try to reduce the size.
I also think it as to many exterior screws, and thats also an aspect i want to redesign.

The hotends are about 60mm from the x smooth rods, and now that you mention that point i will do a little review and reduce that distance (switch the hotend with the fan and i can move the smooth rods more close to the front acrylic). But the two hotends do cover the entire printing surface- well at least in the 3d model .

I am also trying to leave room so that i can easily switch from the bowden hotends to a direct drive extruder ( but that is still is study).

I'm curious, what program are you using that produces such awesome renderings? where did you get the 3D color model of RAMPs 1.4?

i´m using solidworks 2012. The model of ramps is from grabcad.

I love these computer generated designs.. it always works on the pc. the extruder head needs to be bigger, and allow space for cooling before ptfe tubing, as it is right now the tubing will deform and stretch because the hot region is way too short, unless your doing something new. For example i don't see any air vents or aria near cable insert for cooling. one way you could keep it short is about 4mm before tubing connection have water cooling directly to block. that is about the only thing i think will allow such a short hot end that will stay cool enough to attache a cable. at high temp ptfe is like bubble gum, it will start to stretch and deform, unless it has something to support its shape and it is kept cool when that support is missing.


are you sure you want to use bowden design? would it not be more trouble free for new users to have a dual extruder setup that had small but powerful and light motors?

I agree with James. I think two drive gears near the extruder head is going to be far more reliable and require less (or no) retraction. Instead of attaching light motors there (which is hard) how about translating the mechanical motion over a long distance? Just a thought, but I once had a toy helicopter that was driven with a steel wire inside a plastic tubing. The whole thing stretched but still delivered force at quite a distance.

Does the solidworks have a version that doesn't cost $$$?

FYI, you need VERY surprisingly little space between a hot end and the housing behind it to keep it cool (i.e. 60C or less). I dont know why everyone worries about it so much. A simple piece of 1/4" oak wood or many layers of kapton is plenty to shield the back from getting too hot. Its not like heat will conduct up the filament or anything, just the few mm of airspace is enough to keep the temperatures inside the brass. Stress about the consistency of the filament feeding in instead.

James i don´t think i understand your point. The hotend is a j-head (its not 100% right on the measurements but it is very close). I haven't done the connection of the tube to the hotend, but are you saying that it needs even more space and cooling between the extruder head and the tube? (see the images on my second post).

Well i´m sure i want to try the bowden extruder also, but not necessarily on this printer. My current printer has a wade extruder, and i´ve got to the point of getting good prints (but still very far from being a masters in repraps), but what i like most in the project is experimenting. But i will think about the option you and Simba propose.

I hate to highlight and focus on negatives but it is written here:

"The Bowden cable system has one major drawback: Hysteresis. The plastic filament will compress in any extruder, but putting pressure on such a long length of filament will multiply the effects of this compression, leading to springiness. The flexibility of the PFTE tube exacerbates this problem."
[reprap.org]

Balance that with the obvious compact space/weight saving benefits. Awesome idea anyway.

Hi, let me throw my share of knowledge into this:

I tried boweden on my Prusa. It is a fantastic design and has the biggest advantage of taking out the mass. However comes with a disadvantage of sensitive setting. After some frustrating time spent on it, I have not migrated to Wade. This is not for starters. Please see my post on my extruder set up issues Bowden set up issues
Having said that, I am revert back to it once I maser the 3d Printing for the precise reason mentioned above.

In the mean time I have satarted to think on the lines of transferring torque without motors on the carriage. I had a short discussions in the local group of 3ders (Melboourn CCHS - Extruder torque transfer - is it worth trying. I consider it worth giving a try. The flexible shaft technology has improved a lot. Please check this manufacturer/vendor that has very good specs Suhner. One thing I am not sure is if they work both clockwise and counter clockwise.

Couple of design considerations checks:
1. levelling the print bed?
2. Adjusting any parts that are subject to regular wear and tear - belt tensioning etc or ocallsional spills
3. Is there any way we can get XY bed and static Z - then we dont have to be worried about bowden.

pdesigns Wrote:
-------------------------------------------------------
> i´m using solidworks 2012. The model of ramps is
> from grabcad.

Just checked the price of Solidworks - its a ripper - about $5k

What version are you using?

simba, the design wont work unless the ptfe tubing goes down to the hot end, because no matter what there will be a little bit of back-flow, and pla flows easily into this space between the ptfe end and the inside of the hot end no matter how good the design. you will want this to melt again after extruder heats back up,and allow the rest of feedstock to be pushed past it. that is only possible if the ptfe is down all the way to the hot zone to reach melting temp.


the pla will also flow up the tube if it remains hot. the peek insulation keeps the inside warm and the outside cool. you need a way for the ptfe coming out of the hot end to get cooled down, or it will fail during a print because the pla will back flow and heat up the tubing up past the support of the peek walls. once the temp goes above 200C ptfe becomes week and will change its shape based on physical stress. PLA in a glass state will form a bulge in this section of ptfe, and your hot end will be ruined.

Also the peek keeps heat in better than the nozzle end, and can be hotter than the thermistor reading temp that is near the nozzle without cooling or convection.

either way I'm raising a warning flag it wont work the way it is. bowden design is different than a direct drive extruder.

i will give this a more in depth thinking, and will consider all this input. Tanks for this little discussion.

Solidworks is totally out of my budget, i´m using it at my workplace (currently this is a period with little work in Portugal...). But it is a great software for some uses, for organic 3d and fast 3d concepts it might be better to use a polymodeling based one like 3ds.

If you are thinking of designing a 3D printer just for a fun it's good. Everything that is designed must have dummy put aside it.

@ jamesdaniel, coincidentally I have run into this problem this evening. Pleas read my post and the photo attached http://forums.reprap.org/profile.php?1,15281]PTFE choked[/url]. The peek is retaining the heat towards the extruder end. Surprisingly when I opened the hot end, the PTTE is hollow on nozzle end.

Simba Wrote:
-------------------------------------------------------
> I'm curious, what program are you using that
> produces such awesome renderings? where did you
> get the 3D color model of RAMPs 1.4?


You can find the RAMPS model here: with stepper drivers, without stepper drivers.

jamesdanielv Wrote:
-------------------------------------------------------
> simba, the design wont work unless the ptfe tubing
> goes down to the hot end, because no matter what
> there will be a little bit of back-flow, and pla
> flows easily into this space between the ptfe end
> and the inside of the hot end no matter how good
> the design. you will want this to melt again after
> extruder heats back up,and allow the rest of
> feedstock to be pushed past it. that is only
> possible if the ptfe is down all the way to the
> hot zone to reach melting temp.

I have never heard of the PTFE lining inside the extruder and the PTFE bowden tube being one piece. Do you really mean that? I thought that the usual way was to use a normal extruder and terminate the bowden tube above it with a clamp. That's what I did and it works just fine.

ptfe goes down to the hot end, inside the extruder. for examples look at ultimaker.

I don't understand why they would choose a material that melts so near the ABS Tm.

do you mean Tg, the glass transition temp? Tm, the melting temp is 327c avg for ptfe.


glass transition temp of abs is 105 °C it flows at 200-250c I don't know of its true melting temp as it is amorphous.

glass transition temp of ptfe 130 °C it melts at 327 °C

glass transition means it is in a rubbery flexible state. abs properties allow it to become more liquid like at higher temps.

the material properties of abs allow it to flow while in glass transition temperatures.



abs wont flow in nozzle until it is at a higher temp.

both materials have a huge range where in the glass transition.

[en.wikipedia.org] for abs
[en.wikipedia.org] ptfe

Simba Wrote:
-------------------------------------------------------
> I don't understand why they would choose a
> material that melts so near the ABS Tm.


The short version is PTFE is very slippery, so the incoming plastic doesn't stick to it, even in it's rubbery/molten state.
PTFE does creep over time, the better hotends use larger diameter PTFE tubes and don't use it for anything structural.
I'm not sure there are a lot of other materials that could be selected with similar properties, IME the PEEK usually fails first if the hotend overheats.
There are some newer designs using very thin stainless steel thermal breaks without a PTFE liner, I haven't tried one, nophead did some tests on a stainless extruder some time ago.

pdesigns Wrote:
-------------------------------------------------------
> i´m using solidworks 2012. The model of ramps is
> from grabcad.


Hello,
Could you please refer me to the file? i could not find any ramps model anywhere in there!
Thanks,
Mike

Hi, its been a while, but i went to check and i did not found the ramps 3d model on grabcad neither, so probably i was mixing things up
and downloaded it from the links that archistrong posted above.

Good find! Thanks