> > I'm making stuff that LCNC will eventually run on a mill or lathe, or > that the Stellarium program can run to manage the precise aiming of a > star chasing telescope, but all the SW running the 3d printer is the > freebie stuff yu can download to work with the printer. > > The src code for this, in the form of .stl files came from a thingyverse > project, and TBT was not at all optimized to make a good harmonic drive. > Schloppy cad work IMO. So much of what I've been doing is tweaking the > sizes for closer performance, and finding a suitable material to make > each piece out of. PLA is extremely brittle so flexgears last 1 to 10 > minutes. TPU has way too much flex, so while it can be made to go thru > the motions, the motions are, to be kind, inaccurate like there's a > batch of rubber bands running things. Then Hubert donated a spool of > PETG, which is fairly rigid, but bendable without sounding like a bag of > popcorn in the microwave if held up to your ear and the sides flexed > inward a half inch on each side. So thats the flexgear material, but > its not a high speed material because it needs 50C hotter nozzles and > beds than PLA, and thats a problem for a $400 printer. So the plastic > flows slower. > > But I'm going to make new end-caps where the nema 17 motor that runs > these mounts, which as you know, can get hot, too hot for PLA because it > will cold flow at those motor temps. This heat also telegraphs up the > motor shaft into the bearing carrier that runs inside the flexgear and > forces the flexgear into the eliptical shape that does all the magic, > destroying its ability to keep a good grip on the motor shaft, so they > are being made of PETG because of its higher heat tolerance and may be > superglued to the shaft in the final assembly. The rest of the assembly > isn't subjected to the motors heat, so can be made of PLA. > > When making stuff out of plastic in a printer, its laid up a layer at a > time. In this case the layers are .12mm thick. Thats quite a bunch of > layers if the part is 40mm tall. > > So the cad design's .stl is fed to a slicer, which makes gcode the > printer understands, telling the printer where to move to and how fast > as it draws a line of hot plastic by ejecting it thru typically a .4mm > nozzle tip as this tip travels over the part flying .12mm above the > previous layer. Yeah, its slow, watching paint dry slow. But it also, > if all the stars are properly aligned, works. This flexgear.gcode is > aboout 40 megabytes. > > There are several slicers, commercial and freebies, I'm useing > UltiMaker's cura-4.7.0, which works fairly well and is a free download > from their site as a self contained appimage. > > The printer runs a gcode interpreter called Marlin, usually on an arduino > but it is no comparison to LCNC. Very simple, it doesn't do curves, so > the slicer has to break the curves up into short straight lines that if > you don't look too close, blend into curves. > > I am of course a new bee at this, so I am both learning and asking a > boatload of questions. > > Part to build bed adhesion can be a problem. You heard glue or hair spray > used and I can testify that when its stuck, it can be stuck tight enough > it will take a piece of glass with it. I finally found a small eyebrow > of glass missing from one of my two borosilicate plates, stuck to the > bottom of a bearing carrier that spent the night in the freezer before > it finally came loose. Call it glue, but what it really is, is a film > that prevents that solid bonding but which still holds the part as it > being built. A kids purple glue stick, smeared around on the glass, > then scrubbed into a just noticable haze by a wet paper towel or a > toothbrush works well. > > Probably more than you wanted to learn in one gulp. :) > > Stay safe and well, Leonardo.
Thanks for the news on your work Gene, it's always good to know you're doing cool stuff. I'm tempted to get into 3d printing because sometimes I need to do things like keyboards, mounting frames, and so on for the machines I'm retrofitting but at the moment I'll send those jobs to a third party 3d printer. Sadly I can't get involved into that yet because there's too much work to do (which is not a bad thing at all). I'm retrofitting a new lathe. This was an hydro-copying german lathe (Heyligenstaedt brand) that was fully automatic but off course now it's a LCNC patient :). The thing is, the lathe had a fixed angled X joint that was casted and machined in that way so I'm now machining that structure to fit some roller guideways and make the X and Z axis perpendicular to work like a regular CNC lathe. I could have got away with that using kinematics but the problem is there was not enough room to fit the servo motor and ballscrew support so this was the more practical solution. On the other hand I'm in the middle of the process of building a CNC router for another project (woodworking). Fortunately, I don't have too much machining left to do in our shop, so now I have to send some of the frame parts to machine so I can finish that little beast. I'm happy to know you're full of work to do. Keep up with that please, it's always a pleasure to read your progress and learn from your experience. It would be awesome if all us the LCNC people could live a little closer so we can share a beer or something from time to time :) Best wishes from here! El lun., 7 sept. 2020 a las 20:58, Gene Heskett (<[email protected]>) escribió: > On Monday 07 September 2020 16:26:56 Leonardo Marsaglia wrote: > > > Gene, > > > > Is this printer controlled by LinuxCNC? I know you've been doing a lot > > of work making LCNC to work with the Pi but I don't remember if you > > were controlling your 3d printer or another machine with that. > > > I'm making stuff that LCNC will eventually run on a mill or lathe, or > that the Stellarium program can run to manage the precise aiming of a > star chasing telescope, but all the SW running the 3d printer is the > freebie stuff yu can download to work with the printer. > > The src code for this, in the form of .stl files came from a thingyverse > project, and TBT was not at all optimized to make a good harmonic drive. > Schloppy cad work IMO. So much of what I've been doing is tweaking the > sizes for closer performance, and finding a suitable material to make > each piece out of. PLA is extremely brittle so flexgears last 1 to 10 > minutes. TPU has way too much flex, so while it can be made to go thru > the motions, the motions are, to be kind, inaccurate like there's a > batch of rubber bands running things. Then Hubert donated a spool of > PETG, which is fairly rigid, but bendable without sounding like a bag of > popcorn in the microwave if held up to your ear and the sides flexed > inward a half inch on each side. So thats the flexgear material, but > its not a high speed material because it needs 50C hotter nozzles and > beds than PLA, and thats a problem for a $400 printer. So the plastic > flows slower. > > But I'm going to make new end-caps where the nema 17 motor that runs > these mounts, which as you know, can get hot, too hot for PLA because it > will cold flow at those motor temps. This heat also telegraphs up the > motor shaft into the bearing carrier that runs inside the flexgear and > forces the flexgear into the eliptical shape that does all the magic, > destroying its ability to keep a good grip on the motor shaft, so they > are being made of PETG because of its higher heat tolerance and may be > superglued to the shaft in the final assembly. The rest of the assembly > isn't subjected to the motors heat, so can be made of PLA. > > When making stuff out of plastic in a printer, its laid up a layer at a > time. In this case the layers are .12mm thick. Thats quite a bunch of > layers if the part is 40mm tall. > > So the cad design's .stl is fed to a slicer, which makes gcode the > printer understands, telling the printer where to move to and how fast > as it draws a line of hot plastic by ejecting it thru typically a .4mm > nozzle tip as this tip travels over the part flying .12mm above the > previous layer. Yeah, its slow, watching paint dry slow. But it also, > if all the stars are properly aligned, works. This flexgear.gcode is > aboout 40 megabytes. > > There are several slicers, commercial and freebies, I'm useing > UltiMaker's cura-4.7.0, which works fairly well and is a free download > from their site as a self contained appimage. > > The printer runs a gcode interpreter called Marlin, usually on an arduino > but it is no comparison to LCNC. Very simple, it doesn't do curves, so > the slicer has to break the curves up into short straight lines that if > you don't look too close, blend into curves. > > I am of course a new bee at this, so I am both learning and asking a > boatload of questions. > > Part to build bed adhesion can be a problem. You heard glue or hair spray > used and I can testify that when its stuck, it can be stuck tight enough > it will take a piece of glass with it. I finally found a small eyebrow > of glass missing from one of my two borosilicate plates, stuck to the > bottom of a bearing carrier that spent the night in the freezer before > it finally came loose. Call it glue, but what it really is, is a film > that prevents that solid bonding but which still holds the part as it > being built. A kids purple glue stick, smeared around on the glass, > then scrubbed into a just noticable haze by a wet paper towel or a > toothbrush works well. > > Probably more than you wanted to learn in one gulp. :) > > Stay safe and well, Leonardo. > > > El lun., 7 sept. 2020 a las 17:24, Gene Heskett > > (<[email protected]>) > > > > escribió: > > > On Monday 07 September 2020 16:01:05 Gene Heskett wrote: > > > > On Monday 07 September 2020 14:57:13 Gene Heskett wrote: > > > > > I just noted that when it starts building the splines on the > > > > > flexgear, that the color is getting air and hair contaminated, > > > > > turning frosty. So since I've got the extra snout installed, I'm > > > > > wondering if its cooling too fast, so I did a tune->fan speed > > > > > down to 52% to see if the slower cooling gives it time to flow > > > > > smooth and restore the color. Report later after its done a few > > > > > layers. > > > > > > > > > > Cheers, Gene Heskett > > > > > > > > It occured to me that less cooling aimed at the nozzle would also > > > > help its ability to supply flow, so I am upping that about 1% a > > > > layer, up to 104% so far, with no skip clicks from the extruder > > > > drive. Too bad I can't control both at the layer count in cura, I > > > > think it might be helpfull. > > > > > > Another PS, I am up to 106% for flow, and the solid candy red color > > > has returned to the splines, 23% for fan speed, I'd guess its > > > turning 400 rpms, so no great cooling effect. This is a definite > > > improvement in laying solid plastic. > > > > > > So put this in your trivia basket. > > > > > > This piece is about done. > > > > > > Cheers, Gene Heskett > > > -- > > > "There are four boxes to be used in defense of liberty: > > > soap, ballot, jury, and ammo. Please use in that order." > > > -Ed Howdershelt (Author) > > > If we desire respect for the law, we must first make the law > > > respectable. - Louis D. Brandeis > > > Genes Web page <http://geneslinuxbox.net:6309/gene> > > > > > > > > > _______________________________________________ > > > Emc-users mailing list > > > [email protected] > > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > > _______________________________________________ > > Emc-users mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > Cheers, Gene Heskett > -- > "There are four boxes to be used in defense of liberty: > soap, ballot, jury, and ammo. Please use in that order." > -Ed Howdershelt (Author) > If we desire respect for the law, we must first make the law respectable. > - Louis D. Brandeis > Genes Web page <http://geneslinuxbox.net:6309/gene> > > > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users > _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
