early concept of my Be Bot printer (B as in Belt and Bär (bear/baer)). Using 3 independent belts for X/Y/E. The X/Y using double precision (Gun tackle). I am not using a single belt like the H-BOT design, to minimize interaction between axis.
Further All Motors (Stepper or Servo) are outside and fixed, so size and weight doesn’t matter (well inertia of the armature does). The Z- Axis (not shown) will be 3(or 4) Spindles so calibration of the bed (tilting) is possible (maybe also 5-axis printing). I still think about an upside down printing process with a cooled teflon plattform around the nozzle.
Using belts allow to scale this (maybe using Vive Lighthouses for precision at room scale) - as pulleys can be printed and additonal Belts are inexpensive this should keep costs still low. Only the Extruder belt is looped and may be a problem on bigger scales as this would be a non standart part.
When the extruder is being driven you will pull from one side when you retract you will suddenly pull from the other side. There is always slop in the bearings and that’s where the deflection will come from.
As said - the force is importand. It is also the reason not to use the H-Bot design. The slop in bearings can be reduced as the whole system (belts) are under tension. ( further about that topic was discussed here https://plus.google.com/+UlrichBaer/posts/MeCRNAAuHFe )
The Y bridge stage makes sense (like Cartesio etc) but I’m not sure I’m following the stationary X stage. Is that two motors set up like an HBot, but with a full loop? How long is that belt going to be?
@Sven_Eric_Nielsen you seem so sure about this but you are just wrong. I know it is hard to wrap your mind around this and as they are partly on top of each other it is not directly visible. Blue is just a 90° version of green (with different pulley distances and sizes).
So let me explain blue. The blue belt’s ends are fixed at the black pulleys. The stepper at the upper left corner is now turning cw, moving the belt from the left to the right side. The bridge is moving to the left. Hope i could clarify this. Please let me know if you agree or still think this would not work and please elaborate this further so i can learn from your astute mind.
@Ulrich_Baer I think the problem is the angles and overlap in your renders. Would you be willing to show two separate drawings with only the X stage and only the Y stage?
Well honestly, it’s usually not very difficult for me to wrap my mind around something like this. As long as the drawing is clear
And its not really obvious where something is fixed and where things are moving/turning. And it’s also obvious that I’m not the only person with problems to understand the “drawing”.
Now that I know where the belt is fixed and now that I know that it’s in principle a pulley system with a reduction of 2 I have some difficultys to understand the reason.
What do you expect? Yes, in theory (if your ignore all physical facts) you could get 2x resolution. But you have longer belts, higher speeds on motors which are not designed for high RPM, you are probably forced to use a 32bit electronic and so on. I don’t say that I don’t see advantages. In fact I thought about similar thing several times. But always dropped these ideas because I was not able to find enough advantages over a coreXY.
And due to the fact that I worked on a lot of different remote extruder approaches I also have to ask how you plan to convert your rotation around Z axis into a useful rotation for the extruder. Gearbox or something like a sharp wormgear?
Ok, that’s more clear, thanks. Looks like it will work. But is the blue belt routing really better than a traditional XY bridge gantry design with a separate belt loop on either side? That would add 1 motor (or a torsion sync rod or whatever) but eliminate four idlers, shorten the free belt length for better stiffness, help rigidize the gantry square against extruder drive belt forces, and reduce the envelope constraints around the XY carriage.
@Ryan_Carlyle not sure what you mean with the torsion sync and additonal motor? As the XY axis are dedicated there shouldn’t be any more problems about pulleys and length per axis as with the H-Bot or Core XY.
@Sven_Eric_Nielsen
to clarify i have seen that you worked on those systems before. Just a friendly advice: Question if something is not clear - I assume that it didn’t came to your mind that you may missed something. (see https://en.wikipedia.org/wiki/Curse_of_knowledge ) or that i am maybe as clever as you are. And i know that i also overlook something regularly.
So back to the topic. The H-Bot has a severe problem with imbalanced forces, which is why the coreXY was invented. The Core XY has a problem of crossed belts, which is why many using threads instead of timing belts.
As in my description i am not sure to use Stepper - maybe DC (brushless) with hall position sensor (aka servos).
As i want to use a third belt for the extruder i’d like to have X and Y independent to minimize interaction. And additional belts/pulley are not expensive. Also a timing belt can transmit forces better compared to a thread and without friction at the drive pulleys due to overcrossing.
(the friction problem could be solved by a guide like here:
or
but this is more complex or not suited to be scaled up)
Regarding the extruder - i have made a first (unsatisfactory) test with an axial worm drive to feed the filament (see my post https://plus.google.com/+UlrichBaer/posts/PQ4hdQU4jJs )
So the solution will need some more procastination (overthinking), but at the moment i will eventually feed the fillament 90° into the Hotend - so after it is molten it will exit axially to the extruder drive. Maybe the extruder pulley need to be made bigger to lower the belt force- will see.
You know already that crossing belts on CoreXY are quite outdated these days?
I like in general the idea of closed loop systems (servos). But it’s also not the cheapest solution. In fact it’s one of the most expensive. So if you can spend money on those motors you probably also don’t need to print your pulleys
Anyway, if you have a cheap/smart solution for a closed loop system I’m getting really curious!
I’m also curious to see how you solve the filament drive “problem”.
In general I think a remote extruder is a very interesting approach.
Which firmware/electronic do you think of?
but you are just wrong. I know it is hard to wrap your mind around this and as they are partly on top of each other it is not directly visible. Blue is just a 90° version of green (with different pulley distances and sizes).