{"id":15673029,"url":"https://github.com/jonnor/tapecore","last_synced_at":"2025-05-06T22:12:54.027Z","repository":{"id":138841669,"uuid":"65357870","full_name":"jonnor/tapecore","owner":"jonnor","description":"Digitally fabricatable machines using low-friction tape for sliding surfaces","archived":false,"fork":false,"pushed_at":"2018-04-30T19:14:39.000Z","size":24212,"stargazers_count":16,"open_issues_count":0,"forks_count":3,"subscribers_count":3,"default_branch":"master","last_synced_at":"2025-05-06T22:12:46.226Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/jonnor.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2016-08-10T06:50:52.000Z","updated_at":"2025-01-08T15:11:53.000Z","dependencies_parsed_at":"2023-04-08T06:36:59.852Z","dependency_job_id":null,"html_url":"https://github.com/jonnor/tapecore","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jonnor%2Ftapecore","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jonnor%2Ftapecore/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jonnor%2Ftapecore/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jonnor%2Ftapecore/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/jonnor","download_url":"https://codeload.github.com/jonnor/tapecore/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":252776600,"owners_count":21802469,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-10-03T15:35:33.217Z","updated_at":"2025-05-06T22:12:53.563Z","avatar_url":"https://github.com/jonnor.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# TapeCore\n\nConstruction techniques for digital fabrication machines and linear axes that aim\nto be extremely easy-to-replicate. Key ingredient is using tape surface for plain-bearings\non parts made with commonly available 3d-printers and/or laser-cutters.\n\n# Handy\n\nOpen-bottom frame, meant to be positioned over workpiece/table.\n\n## Handy A4\n\nSmall version, with A4 workarea (22 x 30 cm). \n\n![TapeCore Handy A4](./doc/handy-a4-concept.png)\n\n[FreeCAD project](./handy-a4.fcstd)\n\n[Bill-of-materials](https://docs.google.com/spreadsheets/d/1o2K1h2c_w2d49ZTSx7vD1f-7HGYju6wTmgCQS48S1lg/edit?usp=sharing)\n\n### Status\nMay 2018: **Needs revision**\n\nVersion 1 with split rails has a lot of backlash in the Y axis,\nprimarily due to the gantry twisting. This is planned fixed in a future version.\n\n### Handy A4 penplotter\n\nFirmware: [GRBL with servo](https://github.com/jonnor/grbl/tree/handy-a4-plotter).\n\nCAM/MachineControl: Recommend [LaserWeb4](https://github.com/LaserWeb/LaserWeb4). Set gcode options, Tool On: `M3 S40` and Tool Off to `M3 S0`. \nCan also use Inkscape gcodetools + [custom postprocessor](./zservo_post.py) + Universal Gcode Sender\n\n### TODO\n\nv1.0 = penplotter\n\n* Fix Y axis backlash. Use rail on each side?\n* Improve diagonal stiffness. Wedge connecting rail with side?\n* Add endstops\n* Improve wire guides for motors\n* Fix parametrics to respect material size+clearance\n* Make pen pressure more even. Use a spring/rubberband in downward direction\n* Publish files with part layout ready for lasercutting and 3d-printing\n* Document assembly steps\n\nv1.1\n\n* Gantry: Avoid needing glue to assemble\n* Gantry: Move adjustment of slider pressure to bottom. Wider sliders on top \n* Gantry: Move all wire attachment to front. Requires idlers/wires to be placed vertically\n* Add wire guides to all idlers. Can be disassembled without loosing wires\n* Add carrying holes to sides\n* Add holes for attaching to table/wall\n\nv2.0 = laserengraver\n\n* Test running plotter vertically\n* Test with laserhead, primarily engraving\n* Add laser guard around diode (Rishalaser style)\n* Add laser reducing light on sides+top, using window tinting film.\n* Microswitch safety turns off laser if lifted from table\n\nv3.0 = reprap?\n\n* Test maximum cut depth in wood/plastic and 2.5w\n* Test splitting gantry into 2/3 pieces\n\n## Learnings\n\n\n### Silicone good lubricant with Kapton\nPure silicone spray, like also used as mold release, is the best.\nWorks well with Kapton, gives good glide and seems to stay around for a while.\nMechanical lubricant spray based on silicone also works (tested manufacturer: WD40).\nHowever is a bit too fluid, smells badly, and seems to dissolve glue of the Kapton tape.\n\n### Mixing directions important with 3d-printed sliders\n\nIf parts are printed with sliding geometry in same direction they have same periodic pattern.\nThis becomes slightly 'interlocking' when sliding against eachother.\nIf one part is printed 90 degrees this locking effect does not happen.\nEx: If A is printed with sliding in X direction, print B with sliding in Z direction.\n\n\n```\nBAD\nA:  | | |\nB:   | | |\n\nGood:\nA: | | |\nB: _ _ _\n```\n\n### Ideas\n\n* 3d-print the 'dragchain' that guides wires etc out to toolhead.\nSet of interlinking tubular pieces, possibly print-in-place.\nPossibly held together by fishing line.\n\n\n## Handy A2\n\nMedium-sized machine, A2 workarea (60x42cm)\n\n* Scaled up version of Handy A4 design\n* Aim for 60x120 footprint or smaller. Can still fit on a typical table\n* Splits gantry, sidewalls and rails into 2/3 pieces.\n\n# Axi\n\nCantilevered CoreXY, inspired by Axidraw V3.\nA4 workarea, for use with pen or laserhead.\n\n![TapeCore Axi concept drawing](./doc/axi-concept.png)\n\n[FreeCAD sources](./axi.fcstd)\n\n## Status\nJust a concept sketch, not actively developed.\n\n\n# Doverail Slim\n\nFirst 3d-printed linear axis. Geometry inspired by dovetail slides found in lathes and mills,\nand the narrow rail format inspired by Hiwin MDG type linear actuators.\n\n[VIDEO: First tests](https://www.youtube.com/watch?v=5IGngfO671M)\n\n![Doverail Slim testing](./doc/doverail-slim-test.jpg)\n\n![Doverail Slim model](./doc/doverail-slim-model.png)\n\nFlexing fingers on the slider gives some extra tolerance for variation along rail.\nM3 screwholes for attaching to the support of the rail.\nMultiple pieces can easily be joined together to build longer axes.\n\n## Limitations\n\nSide-loading capacity is limited due to springyness of slider fingers.\nRotational force will easily twist the slider.\nThe short slider length and narrow rail contributes, but also primarily due to flex of slider fingers.\n\nSo for a sliding Y bed (Prusa i3) would need dual rails.\n\nNot suitable for cantilevered designs (Y axis on a PrintrBot Simple or Z-axis on Ultimaker).\nIt may be possible to compensate a little bit by using multiple widely-spaced sliders, as one would with a LMU6/8/10 based design.\n\n## Learnings\n\n### Tape adheres poorly to printed PLA\nPoor adhesion of tape to printed PLA is a challenge.\nOffice-tape, Kapton and PET all stick pretty bad to the rough surfaces.\nEspecially for internal curvature. Applying tape to the rail is tricky due to its length, and while mounted together.\nHaving a precisely cut tape piece ahead of time would help, as would a mounting jig which still allows easy access to underside.\n\nEdges to be taped should be: flat, tolerate slight misaligment and excess, ideally allow wrapping tape to adhere to itself.\n\n### Vegetable oil gives higher static friction\n\nWhen adding a couple of drops of olive oil to the PLA rail with PETP taped slider, the slider became much more 'sticky'.\nPossible that sliding friction went down, but for plain bearing usually static friction is the biggest problem.\nWiping it off with a wet cloth fixed the problem.\nNot known if a less viscous oil and/or hydrophobic/mineral oil would not have this problem.\n\n### Sliding non-coated printed parts works\n\nSince taping the rail is a bit tedious, we're primarily testing without any surface treatment on rail.\nAgainst a PETP taped slider, friction is OK and there is no noticable wear after hundreds of repetitions.\nLong-term wear is still unknown.\n\nNot tested, but it looks like PLA against PLA may actually perform OK for very simple tasks with low load, low precision requirements.\n\nThis should be even better with a filament like PolyPropylene or Nylon,\nwhich have a lower coefficient of friction and has self-lubricating properties.\n\n## TODO\n\n* Test the design on a Delta 3d-printer, or a CoreXY pen-plotter\n* Make the FreeCAD properly parametric, with variables for length/height/width and tolerance \n* Change to 4x mounting holes on the slider, ideally compatible with MGN9\n* Make an adjustable slider design that works\n\nContributions welcome!\n\n# Doverail Wide\n\nA redesign of the Doverail Slim to have a much wider rail (40-100 mm),\nand using a slider which is fixed on one side and screw-ajustable tensioner on other side. \n\n![Model of wide Doverail](./doc/doverail-wide-model.png)\n\nThis should significantly improve side-loading ability, and make suitable for more kinematic systems.\nExamples would be an XYZ gantry-based system (as on many CNC-mills), and sliding-bed Y (like Prusa i3) using a single rail.\nEven for a CoreXY laser or 3d-printer, the ability to have the X/Y rails vertical (like SmartRap Core)\nsimplifies how axes and end-effectors are attached.\n\nDisadvantage is the higher material use and printing time.\n\n## TODO\n\n* Finish first model, print and test an axis\n* Test with Kapton tape on slider, nothing on rail.\n* Test with UHWM PE tape.\n* Test on a vertical-rail X gantry, with end-effector\n* Test on a sliding-Y bed\n\n## Ideas\n\n* Create sanding jigs, by adhering sandpaper to inner/outer shape, then slide one across the other.\n* Use a steel rod as the idler instead of (608) bearing. For instance the head of a hex bolt?\n\n## Tribology of common materials\n\nNeeded is some proper test procedure and data for coefficient-of-friction tests for commonly available materials and their combinations.\n\nExamples:\n\n* Candle wax. Polyurethane coating, epoxy coating.\n* Lubricants: vegetable oil, sewing-machine-oil, mineral oil, silicone oil.\n* Surface treatment: Sanding, heating, buffing\n\n\n# TapeXY\n\nExperiment in super low-cost XY stage, for uses in digital fabrication techniques\nwhich have minimal weight and forces on head (laser engraving, 3d-printing).\nInspired by the [RishaLaser](http://rishalaser.org) project.\n\n[Youtube VIDEO](https://www.youtube.com/watch?v=Eynk2ZyVWqM)\n\n![TapeXY fully functioning prototype](./doc/tapexy-second.jpg)\n\nKey features:\n\n* Using low-friction tape as basis for gliding surfaces,\ninspired by [a design](http://www.thingiverse.com/thing:3554) by Peter Jansen\n* Using braided Nylon/Polyamid wire (\"Spectra line\") instead of timing belts.\nLike on Tantillus and some Delta printers.\n* Reproduction with primarily lasercutter (or CNC mill), in wood/acrylics.\n* [CoreXY](http://corexy.com) kinematics\n\n### Status\nNot actively developed. Use [Handy](./README.md#Handy) instead\n\n### Parts\n\nFabricated\n\n* Frame [FreeCAD](./tapexy-frame.fcstd)\n* Gantry [FreeCAD](./tapexy-gantry.fcstd)\n* Head [FreeCAD](./tapexy-head.fcstd)\n* Pulleys. NinjaFlex/SemiFlex for friction. [FreeCAD](./pulley-ninjaflex.fcstd)\n* Idlers. 2x [sideguides](http://www.thingiverse.com/thing:31216) for 608 bearing\n\nAssembly: [FreeCAD](./tapexy.fcstd)\n\nMain vitamins\n\n* MDF/Plywood/Acrylic 3-6 mm.\n* 2x NEMA17 stepper motors.\n* 1x Motion driver board. RAMPS or similar\n* Cables for motors and endstops\n* Low-friction tape, width at least 20mm. Recommened:\n[UHMWPE 1in 3mil](https://www.amazon.com/JVCC-UHMW-PE-3-UHMW-Polyethylene-Film/dp/B00WUU61A) |\nAlternative: Kapton tape, as used for Reprap 3d-printers.\n* Strong line. Recommended: Braided fishing line (Spectra or similar).\nAlternative: Nylon, Polyester or linen.\n* 2x Endstops.\n\nMisc vitamins\n\n* 8x 608 bearings, for idlers.\n* 8x M8x40 bolts, for idlers.\n* 8x M3x10 bolts, for stepper.\n* 4x M4x25 bolts, for head/gantry.\n\nEffector\n\n* 1x Laser diode module including driver.\n\n\n## Done\n\n* First working prototype of XY stage, driven by RAMPS/Cura\n* [Laser diode driver](../currentsource). Linear constant-current source based on op-amp+MOSFET, for 5volt supply.\n* Fully operational prototype, driven by RAMPS w/Marlin\n\n### Limitations\n\n* Non-standard lasermodule and attachment\n* Gantry stiffness is too low\n* Small working area, approx 15x10cm\n\nMissing\n\n* Holes in backwall for motor cables\n* Mounting holes for endstops\n* A belt/line attachment and tensioning system. Just use strip-ties??\n* Holes for attaching RAMPS/controlboard\n* Cables management for cables going to head/effector\n\n## I-beam\n\n![First I-beam prototype](./doc/ibeam-first.jpg)\n\nDesigned to be a basic construction module, from which relatively large multi-axis machines can be made.\n\n[v1 on Thingiverse](http://www.thingiverse.com/thing:1850612).\n[v1 demo video](https://www.youtube.com/watch?v=oqJivOp4JyQ)\n\nFeatures\n\n* Rigid I-beam structure, with double-layer core for rotational stiffness\n* Trapped nuts allows attaching from all sides\n* Use as linear guide for slider by wrapping low-friction tape\n\nSource files for FreeCAD\n\n* [Beam](./ibeam2t.fcstd)\n* [Slider](./ibeam2t-slider.fcstd)\n* [Motor connection](./ibeam2t-motor.fcstd)\n\nVitamins needed:\n\n* Low-friction tape. UHMW PE or Teflon ideal. Kapton also works OK\n* NEMA17 stepper motor, incl M3 screws\n* Braided line (fishing line or similar)\n* 4 pieces M4x20+\n* 4 pieces M4x15+\n\n### TODO\n\nBeam\n\n* Improve trapped nut pattern in sides. Top-left and bottom-right (diagonal) should face one side\n* Remove redundant (blind) holes in top/bottom layer\n* Make tabs wider, reduce number by approx half\n* Split up cut parts into multiple, with top/bottom/sides being a staggered/lapped\n* Maybe reduce height/width a little bit. 30x40 mm?\n\nTests\n\n* Test making axis in multiple parts, with total length 1.5-2x than machine working area.\n* Test in 3-4 mm plywood. Faster to make and possibly self-replicatable. Is it rigid enough?\n* Design \u0026 test a CoreXY using the I-beam as basic building block\n\nOther designs for lasercut beams (untested)\n\n* I-beam. [FreeCAD source](./ibeam-20.fcstd)\n* II-beam. [FreeCAD source](./ibeam-30.fcstd)\n\nTools for calculating stiffness of beams\n\n* http://www.had2know.com/technology/I-beam-calculator-moments-engineering.html\n* http://www.engineersedge.com/section_properties_menu.shtml\n* http://www.amesweb.info/SectionalPropertiesTabs/SectionalPropertiesIbeam.aspx\n* https://en.m.wikipedia.org/wiki/Deflection_(engineering)\n* https://en.m.wikipedia.org/wiki/List_of_second_moments_of_area\n\n## CNC Z-axis\n\n[Servo-driven Z-axis with ER11 shank driven by brushless motor](./cnchead.fcstd)\n\nTODO\n\n* Add rack\u0026pinion\n* Finish servo attachment\n* Add holes for tightening screws for pen holding\n* Test with pen\n* Add pulleys\n* Test with CNC spindle\n\nServo options. Should at least have 5 kg-cm\n\n* [JX Servo PS-4806HB](https://www.banggood.com/JX-Servo-PS-4806HB-6KG-High-Torque-Standard-48_5g-Servo-for-RC-Models-p-1013373.html), 6kg-cm\n* [TowerPro MG996R](https://www.banggood.com/TowerPro-MG996R-Metal-Gear-Digital-High-Torque-Servo-55g-p-982287.html), 10 kg-cm\n\n\n## Pen Z-axis\n\n![Servo driven Z axis for a pen holder](./doc/tapepen-v1.jpg)\n\n3d-printed files [on Thingiverse](http://www.thingiverse.com/thing:1850720).\n[FreeCAD source project](./tapepen.fcstd).\n\nUsing a printed V-rail geometry, covered with low-friction tape.\nHas adjustment for the tension, was very practical to dial in the correct amount.\n\nFits pens 8-15 mm in diameter.\nLifts straight up, perpendicular to surface. Roughly 10 mm travel.\nUnlike many others servo-driven pen axes which rotate when lifting pen. \n\nUsed in the penplotter which was the award given to the winners of Oslo Innovation Award 2016.\nSuccessfully printed the diploma live on stage.\n\nVitamins needed:\n\n* Microservo. SG92R or other with same size head, https://www.adafruit.com/product/169\n* Low-friction tape. UHMW PE or Teflon ideal. Kapton can be used as replacement\n* 2x M3x30 + M3 hex nuts for attaching pen to slider\n* 2x M3x30 for adjusting tension on the slider\n\nTODO/improvements\n\n* Use a slightly bigger/stronger servomotor\n* Use as basis for a lens-moving Z axis for lasers.\nProbably use 3 v-rails instead of 2, keep tensioning principle.\n\nExisting designs by others\n\n* Servo+fishingline, spring in one direction. http://www.thingiverse.com/thing:749118\n* Servo with 2-bar linkage, lasercut. http://www.thingiverse.com/thing:4185\n* Servo with PTFE bushing, spring in down direction. http://www.thingiverse.com/thing:13407\n* Servo with rotating slot linkage, lasercut+3dprint. http://fab.cba.mit.edu/classes/863.14/people/nathan_melenbrink/Week_15.html\n* Servo with steel bars and POM bushing parts. CNCed https://www.tindie.com/products/ijinstruments/servo-actuated-pen-slide/\n* XY pan+tilt mover, with Z motion. http://www.thingiverse.com/thing:31463\n\n## Wallmounted 3d-printer\n\nConcept based on TapeXY mechanics, adding a wire-driven cantilevered Z-axis.\n\n![Wall-mounted 3d-printer](./doc/wallframed-3dp-concept.png)\n[FreeCAD project](./wallframed2.fcstd)\n\nDesigned for mounting on a wall, to keep precious desktop space untouched.\nInspired by machines such as the SmartRap Core.\n\n## Reproducability\n\n### Power needed to reproduce itself.\n\nMr.Beam II says that with 5W they can cut 4 mm plywood in two passes (no focus adjustment).\nIt seems to be a blue diode laser with no air assist. This probably sets 5W as the a lower limit.\n\n* Adding air-assist might improve cutting efficiency a bit.\n* Adjusting the focus into the material for each pass might help\n* Using PWM to overdrive the laser periodically may also help\n\nIf one could cut 4 mm acrylic instead, or 5-6 mm plywood, that would make it simpler to make a more rigid machine.\nThis _might_ be doable with 8-10 watts IR diode. If more is needed, may need to assemble 15-20 watt using multiple diodes.\n\n[8W diode laser cutting 4mm plywood single pass](https://www.youtube.com/watch?v=gWxxxrpdF10)\n[5.5W laser cutting 4mm plywood single pass](https://www.youtube.com/watch?v=fT6_FLB4vyM)\n[3.5W laser cutting 3mm acrylic single pass](https://www.youtube.com/watch?v=rqom-Ns8LqA)\n[3.5W laser cutting 3mm plywood 5 passes 12mm/s](https://www.youtube.com/watch?v=iQ14XSbGCJk)\n[5.5W laser cutting 3mm plywood 2 passes](https://www.youtube.com/watch?v=j0H8-iMtKs4)\n\n### Construction tricks\n\nTo make more reproducible\n\n* Use I/H beam or similar structures to achieve stiffness\n* Use 2x 2-3mm material instead of 1x 4-6mm, sandwiched using interlocking geometry.\n\nGeneral best practices\n\n* Use same bearings and screws types everywhere\n\nSelf-reproducability.\n\n* Need to create structure bigger than own working area.\nLet I/II-beams consist of multiple sections, with staggered joints?\nDo testing of a long beam (50 cm+). Flatness, stiffness, maximum load.\nKeep compatible with alu profiles? T-slot/Makerbeam/Openbeam\n\n## Laser diodes\n\n* Bangood [1.6W](http://www.banggood.com/Wholesale-Laser-Equipment-c-3491.html),\n[3.5W](https://www.banggood.com/450nm-3500mW-3_5W-Blue-Laser-Module-With-TTL-Modulation-for-DIY-Laser-Cutter-Engraver-p-1103261.html),\n[5.5W](http://www.banggood.com/445nm-5_5W-5500mW-Blue-Laser-Module-With-Heatsink-For-DIY-Laser-Cutter-Engraver-p-999283.html)\n[7W](https://www.banggood.com/EleksMaker-LA03-7000-445nm-7000mW-Blue-Laser-Module-With-Heatsink-For-DIY-Laser-Engraver-Machine-p-1127310.html),\n[15W](https://www.banggood.com/445-450nm-15W-Blue-Laser-Module-Mark-On-Metal-for-DIY-Laser-Engraver-Machine-p-1137779.html)\n\n* Osram PL TB450B [DigiKey](http://www.digikey.com/product-detail/en/osram-opto-semiconductors-inc/PL%20TB450B/PL%20TB450B-ND/5719266)\n* [DTR's Laser Shop](https://sites.google.com/site/dtrlpf/home/diodes)\n* [Laserdirect@Ebay](http://www.ebay.com/sch/laserdirect/m.html?_nkw\u0026_armrs=1\u0026_ipg\u0026_from\u0026rt=nc\u0026_mPrRngCbx=1)\n\nFibre guided\n\n* [Ebay: 8 watt NIR 980nm](http://www.ebay.com/itm/980nm-8W-Fiber-Coupled-Laser-Semiconductor-Fiber-Coupled-laser-8000mW-IR-Laser-/141524214275?hash=item20f3802203:g:jsMAAOSwMpZUolDB)\n* Ebay: [3W](http://www.ebay.com/itm/Laser-Diode-3-1-Watt-923nm-Fiber-Coupled-100um-JDSU-6380-L2-3-1W-NEW-63-00123-/111598682682?hash=item19fbccc23a:g:mhMAAOSwPhdU3sXA)\nand [5W](http://www.ebay.com/itm/Laser-Diode-5-Watt-915nm-Fiber-Coupled-100um-JDSU-6390-L3-5W-NEW-63-00192-/111598671030?hash=item19fbcc94b6:g:0M8AAOSwqu9U3r8Y)\n\nShould be possible to [combine](https://www.rp-photonics.com/beam_combining.html) multiple such fibre laser diodes with one focusing lens (maybe with colliminator).\nTo achieve powers of N times the individual laser source. Though with simple combinators, the brightness will not increase so much, due to widening of beam size.\n Also the heatsink no longer needs to be on the head, reducing weight of moving parts.\n\n## CO2 lasers\n\nFor cheap lasercutters, CO2 lasers are used.\n\nThere are several aspects which are challenging.\n\n* Laser tube itself\n* Cooling the gas\n* The optics\n* Power supply\n\nPrinciple\n\nUsing gas mixture of helium, nitrogen, and carbon dioxide gas.\nA common standard mixture is 4.5 percent CO2, 13.5 percent N2, and 82 percent He. Penn suggests using a mixture of 14:14:72 in narrow-bore CO2 laser.\n\nAactive cooling to keep the temperature of the discharge tube below around 30 degrees C or so.\n\nFlowing-gas lasers\n\n\"the flowing-gas CO2 laser requires a vacuum pump to achieve the low pressures (10-30 torr or so)\"\n\n* [35 Watt Flowing Gas CO2 Laser Tube Kit](http://repairfaq.cis.upenn.edu/sam/rconway/35wtkit.pdf), assembly instructions.\n* [Flowing gas CO2 laser conversion from sealed gas](https://www.photonlexicon.com/forums/showthread.php/17719-Flowing-gas-CO2-laser-conversion-from-sealed-gas)\n* [First Homemade CO2 Laser Built From Scratch](http://jarrodkinsey.org/co2laser/co2laser.html)\n\nReferences\n\n* [Understanding CO2 laser](http://www.laserk.com/newsletters/whiteCO.html)\n* [RP Photonics Encyclopedia: CO2 lasers](https://www.rp-photonics.com/co2_lasers.html), describes different types. Sealed tube versus\n* [Homebrew CO2 Laser Design and Construction Notes](http://www.timefracture.org/laserdocs/laser_design_notes.html)\n* [RepairFAQ: Home-Built Carbon Dioxide (CO2) Laser](http://www.repairfaq.org/sam/lasercc2.htm). Looots of information\n* [RepairFAQ: Carbon Dioxide Lasers](http://www.repairfaq.org/sam/laserco2.htm). Loots of information\n\n## Related projects...\n\nOf mine\n\n* [Clothing](../clothing), various techniques rely on laser\n* [Rishalaser](../rishalaser), build of existing open-source low-powered laserengraver\n\nExisting low-cost laser diode engravers\n\n* [smartDIYs](http://www.thingiverse.com/thing:1026345) open source kit. Lasercut acrylic, steelrods+timingbelts, motor on gantry.\n* [Mr. Beam](https://www.mr-beam.org/) open source kit. Kickstarter 2014. 3d-printed + wood-frame. [Octoprint-based](https://github.com/mrbeam/OctoPrint) software\n* [Emblaser](https://darklylabs.com/emblaser-overview) [2](https://sites.google.com/site/3dprinterlist/lasercutters/darklylabs-a3-diode-laser)\n* [Fabool Laser Mini](http://www.smartdiys.cc/fabool-laser-mini/) claims to be openn source. Also has a CO2 laser available.\n\nOther inspirational machines\n\n* [MPlus One](http://www.thingiverse.com/thing:1104249).\nCantilevered, moving bed like PrinterBot Simple and Smartrap, with lasered/milled structure.\nMaybe has potential in combination with tape principle, possibly making even cheaper than a CoreXY?\n* Axidraw V3\n* Cloth of the Bot\n\nPossible addons\n\n* [Lasercut vacuum table](http://hackaday.com/2016/03/19/diy-vacuum-table-makes-lasering-even-easier)\n* [Air assist for diode laser](http://www.thingiverse.com/thing:1688209), 3d-printed using a radial fan\n* [Air assist for compressor](http://www.thingiverse.com/thing:72691).\nSeems a small oil-free compressor (airbrush etc) is good enough. Recommened ratings are 30-50 PSI at 1-2.5 CFM.\n\n## TapeZ\n\nMoving-bed Z-axis stage for laserengraver/3d-printer to go with TapeXY.\nAlso uses Kapton tape and braided Nylon wire.\n\nExisting wire/belt Z-axis:\n\n* [SmartRapCore alternative Z](http://www.thingiverse.com/thing:896556)\n* [Sli3dr](http://richrap.blogspot.co.uk/2014/07/3d-printers-big-printers-small-printers.html)\n* [Igentis](https://www.youmagine.com/designs/ingentis-a-tantillus-variant)\n* [Printxel](http://printxel.blogspot.ca/2012/12/printxel-community.html)\n* http://forums.reprap.org/read.php?177,310249\n* http://forums.reprap.org/read.php?160,170024\n* http://forums.reprap.org/read.php?177,192178\n* http://forums.reprap.org/read.php?14,319321\n* http://forums.reprap.org/read.php?279,188149\n\nA challenge is that Z-resolution/precision might be a bit low without gearing.\nThe hold-torque of a NEMA17 stepper might also not be enough to keep bed up.\nReduction gearing of between 1/4 and 1/9 is desirable.\n\n## TapeCrane\n\nMy dad wants to build a cheap, large-scale 3d-printer (workspace with sides of 50-100cm)\nwhich could be stowed away when not in used. For this he had two ideas for the kinematics.\nOne being a robot arm with two joints (elbow and wrist), working in the XY plane. Like SCARA.\nThe arm up/down or the bed would move to form Z axis.\nThe effector is cantilevered, making it tricky to keep rigid, and there are multiple\nsolutions for the inverse kinematics and singularity points.\n\nThe other idea was more like a crane, a [cylindrical robot](https://www.google.no/search?q=cylindrical+robot\u0026tbm=isch).\nIt has the advantages of much simpler inverse kinematics (because no articulated joints),\nand that the structure can be made more rigid in various ways:\nUsing tensioned wires (like a guy-wire), by supporting the far end, and/or using a (moving) counterbalance.\n\nThe most common configuration, has a rotating base. Alternatively, one can rotate just the (elevated) arm.\nEither the whole arm can move back \u0026 forth to move end-effector, or the effector can travel along a static arm.\n\nUnlike a cartesian or delta robot, several cylindrial robots could work into the same workarea.\nThey could potentially collaborate on a single (large) part, either for speed or multi-filament.\nOr indepentently making different parts, or multiple instances of the same part.\n\nExample of multi-arm collaborations\n\n* Flexible + rigid printing\n* 3d-printing binder + laying out fiber threads\n* Rough 3d-printing + precision lasering/milling\n* Conductive + non-conductive 3d-printing\n\nChallenges:\n\n* Having enough resolution in rotating joint. Probably will need gearing, challenge then becomes low/zero backlash in gear\n* Mimizing angular momentum of rotating joint, for maximum speed / minimum force. Keeping weight down, and it close to center of axis\n* Cylindrical sectioned build area, does not fit typical objects well. Can partition the circle into 3/4/6/8 rectangular build areas\n* Decreasing precision at long-reach. Can also be seen as feature, better precision on small objects\n\nTODO:\n\n* Sketch a rough concept\n* Test an arm beam, with tensioned wires on top+bottom for ridigity\n* Implementing the IK for cylindrical in firmware\n\n# Experiments\n\n## PlateXY\nExperiment in extremely simple plotter/laserengraver.\n\nFeatures\n\n* Head is wire-suspended between fixed rails (no gantry),\n* Glideblocks tensioned by the drive wires\n* Custom rails fixed to standard plate\n\n### Status\n**Aborted**. In favor of [Handy](./README.md#Handy)\n\nThe gantry-free design was found not good enough for penplotting use.\nThe lack of locking between the two sliders meant they would move independently, causing a twist in the head.\nAt A4 size it might have been barely acceptable, but not better than simpler kinematics like [Axi](./README.md#Axi).\n\nHowever this kinematics can be useful when precision is not needed, and it is desirable to hide the mechanics of the machine.\nFor instance with the rails mounted to floor/ceiling, and using transparent fishing line, one can make pieces move around in XZ 'invisibly'.\n\n## Kapton 608 bearing\n\n.. Wednesday 14 Oct 2015, Oslo ..\n\nAttempted to create a 608-sized bearing/bushing. 3d-printed, using Kapton tape added afterwards.\nUsing a V-profile to make the bearing support some axial load.\nInitially 120 deg, but had to increase to 135-degrees to make it possible to insert inner part into outer ring.\nTolerances when 3d-printed (on a Ultimaker Original) seemed generally OK.\n\nHowever, major issues were found when trying to assemble.\nThe Kapton tape was not particularly sticky to the print,\nand would not easily fold around the V-profiles, especially the outer ring (with inward V profile).\nWhen pressing the inner part of the bearing in, would easily\nmove the carefully aligned tape... Avoiding folds/creases was quite tricky.\n\nI was able to succesfully assemble one bearing, with tape only on the inner part.\nThis rotated quite nicely, with not too much play.\nConsidering the application of belt pulley, where some excentricity is quite OK.\nIndicates that if one can figure out a practical assembly, the idea has some merit.\n\nIt could be that in larger scale, and with a more adhesive tape and, or with simpler geometry,\nthat the idea would have some merit. But with v-style 608 the process was too fiddly to be practical.\n\nFor TapeXY it could be possible to design an alternative idler/pulley solution.\nFor instance one could integrate the belt bottom/top guides, and use top/bottom of\nrotating part for supporting the (tiny) axial load.\n\nIt is however desirable to be compatible with a standard solution,\nor at least also allow a standard solution - since a hacked thing will (probably?)\nnever be as good as a proper bearing.\n\nTODO\n\n* Test cutting low-friction tape to revolved V-shape using laser.\nIf necessary, try to use additional glue. A jig for pressing the tape against when adhering could also help?\n* Try an inverted design: inward V on inner part, outward V on outer ring.\nThis should be easier to adhere tape to, as the ring has shape towards the top/bottom, not hidden inside.\n* Test using UHMW PE / PTFE tape.\n* Design an integrated bearing+belt idler for CoreXY usage, compatible with 608-based\n\n## References\n\nWork by others\n\n* [3d-printed bushing](http://www.thingiverse.com/thing:1196801), replaces speciality linear bearings\n* [3d-printed beam surface](http://www.thingiverse.com/thing:9080/#comments)\n\nGeneral machine building\n\n* [Methods for minimizing gear backlash](http://machinedesign.com/motion-control/methods-minimize-gear-backlash),\nspring-loaded distance, spring-loaded split gears, backloading/dual-gear-train\n* [3d=printed anti-backlash gear](http://www.thingiverse.com/thing:98393), dual layer with printed integrated springs.\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjonnor%2Ftapecore","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjonnor%2Ftapecore","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjonnor%2Ftapecore/lists"}