After taking a look at the vitamin count of my initial design, it became clear to me that the real problem preventing the more widespread replication of RepRaps is related to cost, and complexity. More vitamins contributes to both problems.
Since we know that printed thermoplastics does not quite make long running, reliable bearings, we have to make sure that only the hardest working axis gets the bearings, and a minimum amount of it for each.
Another little problem I discovered, is if you max out the build area, according to the GADA spec, and print a solid block of PLA of 300x300x100mm, your work-piece will weigh more than 11kg! Try lugging that around at speed…
Thus, there are a couple of new concepts on the charts, with a stationary work-piece and low cost low weight Cartesian bot as the main components…
This past weekend I had the terrible misfortune (and luck) to have to repack my study. At the same time I tried to put together some Gerber files for the making of a RAMPS 1.4, some Pololu drivers and ad hoc diverse items to get the electronics sorted.
Turns out that a PIC16F84A costs about just as much as an Allegro A4988 stepper driver chip, except that shipping is not included, and I just happen to have a hand full of the little PICs in a drawer. Add to that the fact that you could actually control two L298N drivers using only one PIC, and that with micro stepping (of your preferred type as well) included sold the idea to me hook line sinker.
Of course there is a price to pay. The code for it did not exist until earlier this afternoon.
In the next few days I will build a single stepper prototype, and extend it to the dual stepper controller by week’s end. I feel another thingiverse thing brewing in the wings…
Another great advantage is that my Arduino code can keep following the current Sprinter development, and will not have to be patched in order to be compatible to the strange XY axis drive that I wanted to experiment with.
Cost of material is one of the factors that is holding me back at the moment. I found a couple of different suppliers of the linear slider elements I need to get the basic frame movement off the ground, but it seems that what I need will cost more than I can afford at the time.
No Problem – Time to get creative.
I am currently putting my energy towards a Ultra Low cost linear guide system, that will use commonly available (and affordable) components. I have already built and rejected two variants of the idea, and while the last one was exceptionally stable, the movement was a bit on the stiff side – would have made the motors work much too hard. I am still thinking of keeping these for the Z-axis… We will see.
The goal is to have a smooth, easily moved but stable and accurate linear slider system, for less than R50 (about $5 these days) for 50cm length, with the sliding elements included. The prototypes will be fabricated from common materials, but the final design will be fully printable, save one or two low cost vitamins. More later.
When complete I will post the design on Thingiverse for the benefit of the Open hardware community.
Now that the hot-end and extruder is defined, it is time to dust of the heap of pieces from my last attempt, and actually build my x-y axis. The idea is to have a very low weight, and low cost x-y axis complex. Linear sliding surfaces will probably be felt only, although there may be some unique challenges to that idea.
This weekend we’ll just have to wait and see.
In the mean time I am rediscovering QR codes. Makes me think. Each of the systems I build from this first Repstrap (including itself) will have a model page, that highlights each iterations features, advantages and know problems in pictures. Each of these systems will have a QR code on a label attached to it, to give info about system.
Each borrowed part will also have its Thingiverse QR code attached to it in a practical way. This will make documenting the systems much easier and a lot more complete.
This weekend I am building my extruder, even though the hot end is not quite complete yet (no heating or sensor…) Let’s call it a wooden Greg’s like hinged accessible extruder. The body of it has the same mounting footprint as Greg’s extruder, making it possible to print myself an upgrade later.
Note that this would be a temporary extruder to get me printing – something that would enable me to design and print the multi material extruder.
The body of the extruder consists of three pieces of wood:
Foot: Sized and drilled to be drop in replaceable with a Wildseyed compatible extruder.
Body: Closely modeled on Greg’s except for the motor placement – the available tooth gears I have were either too small or too big. Had to go with the too big one (obviously)
Idler: Hinged on a piece of metal plate because of the space behind the large gear is restricted.
The gears in use is a 15T small gear and a 75T large gear. Gives a ratio of 5, slightly higher than Greg’s 4.3, or Wade’s 3.367. I am driving it with a smallish stepper motor, but it should pack plenty of punch with this high gear ratio. Stepper speed will not be a problem, since I plan to use (modified) Sprinter firmware.
The filament is driven by a M8 bolt, hobbed using the M3 tap in a drilling machine method. The bolt has a smooth section below the head that extended too far into the bolt, so I tapped an extra 2.5cm of thread onto the bolt. It now has a tailored fit 😉
It is nice to see the pieces taking form… Pictures to follow.
There are a couple of concerns though, not least of which PTFE creep… apparently Teflon does not have a polymer “memory” and might flow like glass given time and pressure. Will have to do some serious testing – at least there is time for that.
Then there is the issue of the 1mm aperture. The 0.8 refers to the wire that gets fed though it, and not the actual hole size. Found some ancient (2005) reprap hot end literature, and the way they made the hole smaller was quite interesting. They melted some lead-free solder into the hole, while keeping it open with stainless steel wire, chosen for it’s diameter. The solder does not stick to the stainless, and allows the maker to pull the wire out afterwards, leaving a perfect hole.
This is unfortunately not viable with the high speed / high temperature operation people are forcing out of the hot-ends with the latest firmwares, like Sprinter. The solder would just melt out again. Got some ideas from it though… Watch this space.
I have been brainstorming on potential extruder designs for Morgan, keeping in mind the 3+ material design spec I set myself. To make this work it will have to be a Bowden cable setup of sorts with a material selector in the extruder feed mechanism. To keep down costs only one extruder stepper motor should be used.
Looking at current extruder hot-end designs, and at material cost it is very clear why some new designs are shying away from the use of high temperature insulating materials like PTFE and PEEK, although these designs seems to compromise a bit too much. Clearly we need to find a way to successfully implement some lower cost material that will perform the task of keeping the hot end hot, and the cold end cold, while conserving as much energy as possible in order to remain well within the GADA requirement for power use…
Enter (yet another) another new hot-end.
I am going to use a combination of brass, stainless steel and silicone rubber to do the required. My initial checks and calculations makes this look good on paper. Information on the new hot-end should appear here soon, and follow on Thingiverse. It should enable us to place three to four nozzles in relatively close proximity, while still keeping the weight of the print head down to the minimum and keeping heat where it belongs – in the hot end(s).