Thursday, July 09, 2009
Another Repstrap crawls into existance...
What I believe is Australia's first working Rep(st)ap.... I've built a *JunkStrap*.
it is defined loosely as: "built from "whatever I had laying around".
more images here: http://davidbuzz.googlepages.com/buzzs_repstrap
for those interested:
* top-left is a 2nd-hand torioidal transformer and diodes/capacitors for 45 supply ( drives steppers )
* black boxes at left are of-the-shelf stepper drivers locally sourced, using 45 supply.
* bank of 10W 18ohm white resistors near left is to prevent the 45V from frying my steppers. These resistors get HOT, so my steppers don't have to. :-)
* rails are extracted from 3x Ancient Epson A3 dot matrix printers, complete with big knarly steppers and belts. ( two of the rail sets are used for Y, and one sits on top and is used for X ( which runs on Y rails below it.) - one large stepper is visible middle-right.
* Z-axis is my least favourite, but does the job. It's a Aluminium framework with 2x U extrusion elements standing vertical. The timber between is NOT attached to the aluminium, and slides up-n-down between them, using them as guid rails.
* Picture above shows a hand-made ( hacksaw, plastic, bolts ) classic darwin threaded-rod extruder , which has since been replaced with a much lighter and more reliable pinch-wheel extruder. ( see other pics )
* circuit boards are Gen 2 electronics purchased from the rrrf.org shop, except for the controler which is a Arduino Mega, cause my firmware is a bit bigger than usual and 14K was just not enough. 128K is way overkill, but having an additional 4 hardware interrupts ( only on the mega) is very handly. I used them for 2x opto-end stops, and 2x encoder-wheels ( for my extruder, and my Z axis) .
* Y-axis was ( until a couple of weeks ago ) dual-threaded-rod driven ( as the above pick shows to the keen observer) , but is now dual-belt-drive ( see later pics in the link provided ) .
* Extruder "hot zone" is made with high-temp kneadable 2-part putty from local hardware store "Bunnings", it is labeled as "Farenheit 500" ( Manufactured by PC-Products Protective Coating Co. in Allentown , PA, USA) , nichrome wire ( from local electronics store "Jaycar" ), brass threaded rod drilled through, and dome-nut carefully machined and drilled to 0.8ish. ( could do to be smaller, if I had the drill bit )
* Extruder "optical-encoder" is from a HP printer, part number Q-9502. it's got a laser diode and lense enabling it to read quadrature signals from an encoder wheel with about 800 slots. ( no,not 8.... 800! ) - overkill, but it was free. ( not in photo above)
* ABS filament was sourced from New-Zealand, from the same people that supply Vik. :-) Import Duties were bearable, and I have about a 100 years worth now. :-)
* I'm running on very-customised GCode_Interpreter at the moment, and almost finished porting my code to run on the FiveD_Firmware that Adrian released.... it's non-trivial for me, as I use 2x DC motors ( extruder, and Z axis) , not all-steppers as the software is designed for... and my stepper-drivers are not the "step and direction" type. :-)
* Yes, it prints, quite well now. the java software says it travelled 9.4 meters, and printed 5.6m, to make a shot-glass, and took about 15 minutes.
Sorry for the long post, let me lnow if interested in more details and/or more specific pictures.
I am president of "Hackerspace Brisbane" - the First Hackerspace group in Australia! If you are local, get in contact with us!
Excellent machine! I really like the use of the printer rails as linear guides.
How much force do the axes generate? Do you think they'd support milling activities? We're about to start construction of a McWire style CNC machine for our space, and I've got some old printers in the garage I've been saving for a rainy day.
How do you find the DC servo approach on the Z-axis? Would you say the servo is superior to stepper for this?
I don't know about your claim to be the _first_ hacker space in Australia, though. Judging from our google groops it looks like a photo finish:
Let me know if you're ever in the Sydney area.
Gavin. (aka MechatronicsGuy)
VP -- Board of Hackers
Robots and Dinosaurs hackspace in Sydney
In the current configuration, my X(left right) axis is not mounted securely to the 2xY axis(front back) rails, so it would suffer too much "play" to be useful for Milling purposes. It's constrained within the rails, but can be easily lifed out from between them. Gravity is all that holds it down, presently, so it can wobble around. It's a easy fix ( put restraints underneath, like above), but for the extruder, I've not needed to do it. :-)
With regard to the Z-axis, I'm not really using a "servo", just a plain old geared DC motor. It's not even got an encoder on it. I just calculated my layer thinkness as the number of milli-seconds it takes my motor+gears+threaded rod to raise the bed the right amount. it works well for me, and it really simple, because Z only moves one way, and is hardly ever changed.
---a bit off topic---
.. but the founding dates listed here ( http://hackerspaces.org/wiki/List_of_Hacker_Spaces ) show Brisbane group founded 30th March, and first post on the Brisbane ggl group was the 2nd April, and the Sydney group wasn't till 23/28th April , nearly a month later. I also know that when I put a pin in the map at the above URL, it was the *only one* in Australia, as that was behind my motivation for doing it... I couldn't believe Australia was so far behind the rest of the world with this!. :-)
Best of luck in Sydney... I hear there's a near new premises being worked on!
Links to this post: