GearHeads Forum

James:

  Ensure you save the project with your gear in it, then load that project into GT ( Gearotic Thoughts ). Rigth click on that gear and select STL export. The STL you get is orders of magnitude better than GM's STL's. It will be closed manifold and have no holes in it, GM's stl's are full of holes and open manifold. It makes it hard to import them for further work in programs that accept STL's.
    One of the many reasons GT is being written is the poor STL quality from GM. I just didnt know at the start how important the 3d model outputs would be so they werent designed properly. GMs replacement will also have greater CNC capability for cutting stl's. If you have the development version installed GT should be an icon on your desktop..

Thx
Art

Hi, I see my thread has run off on its own..

Yes, once you get comfortable with metric it really is easier, both the maths and visualization, than inches.  I still have trouble at the macro, meters and decimeters, but for small scale stuff, millimeters and down to microns, metric is the way to go, IMO.

What I am futzing around with now is trying to get a crown gear. I have not seen where GM supports them since they are essentially a form of bezel gear.  I have been trying to produce a model of one in GM, which it does pretty well. And then import the model into another CAM program, CAMBAM in this case, to take the model and generate the g-code for it.  The couple of attempts I have made at it haven't been successful, little things like the model not coming in and rendering right, and i haven't had the time to go edit it to the host application's liking.  Tomorrow I am going to take another whack at it.  Unless a fix in GM is in the works? 

Kirk:

Interesting..

Welcome to the dark side..personally, I think once you get used to metric you never go back.
By the way, take note that a knuckle angle should be around 15 degrees, more than that and youll get very thin and sharp teeth.. other than that they seem a great type of gear for extruders..

Art

Thanks, Art!  I ordered a small photo tent kit so maybe my photos will improve, also forgot to angle them a bit to show the spokes.  I guess the spokes were a bonus

I know stl's don't have units.  This 3D printer stuff is forcing me into the metric dark side, almost can think in metric now.

The thickness difference might be due to something that got changed in the many months since I made the first knuckle, but it's still just for fun.  I'll be out of town for the kid's graduation, so I'll try a pair of gears with the new version when I come back.  Knuckle extruder gears could be interesting.

I've been experimenting with different size nozzles, and not paying a lot of attention to tuning settings it for print quality - that print was pretty rough, and it's done a lot better, so don't take it as an example of PB print quality.  There's a how small can you go treefrog printing mini-competition going on, the leaders are printing them 2mm high with 0.1mm nozzles.  I'm not even in the running, but haven't gone any smaller than 0.3 yet.  Just did some test prints with a 1.2mm nozzle to look into somebody else's problem, way different.  The extruder becomes the limiting speed factor, which the original poster didn't expect.

If you even accidentally brush against a 0.2mm bit it'll snap, and maybe even heavy breathing will break one.  I'm probably not up to anything smaller, yet.

Kirk

Hi Kirk:

Just printed a couple of knuckles here. They mesh very well together, as if their a locked set you cannot pull them apart side to side, its an interestig type of gear really..quite unique. Didnt find any trouble doing them and they look pretty symetrical.
HErringbone you should be able to do by end of week if I can finish a few items this week.

Art

Hi Kirk:

  The Helicals, ( knuckles are a type of helical ) never had spokes, BUT the information
of the spokes selected was stored in the data. GT isnt really loading the project the same way
GM does. GT, once it loads a GM project basically has an outside contour and spoke information as well as thickness and such, from there it does a re-extrusion based on its
own code, not the extrusion library I used to use.

  Because of this some things will vary, ( though thickness isnt one of them , they should be the proper size. The next version of GM will allow you to turn those spokes off or on, or change the shaft size and such in real time on the screen as a 3d object. I had to do it in this order so that extrusions and such work, then Ill be adding the code to actually recreate
the gears including spoking and such.

    STL's have no setting of mm or inch, they are in the native units used when GM created the gear. A gear of 1" or 1mm is the same STL in size. This is another thing that will take care of itself when GT gets a bit further along and can do creation as well as extrusion.

Ill run some tests here on thickness and such, and Ill print a knuckle so
we can compare. Ill let you know what that looks like on mine.

  I wasnt aware that people were making .1mm nozzles, thats interesting...
sounds rough, I can understand why so many broke bits.

  Your printer seems somewhat rougher than mine, but I suspect thats more
  software than anything else. Slicr seems to have no trouble with the
  STL's Ive been running, but then I use the UP software so it may be that.
  I do run Slicr as well though, and Asimp as well to test all of them. GT
  has been moved quite a bit the last few weeks though, so I think its close to
  time for me to release a new version for testing.It wont be too long.
 
 
  Thx
  Art
 

Right after I got my Printrbot I printed a knuckle gear after fixing the .stl file.  Tonight I took the same project file and tried it with GT, and came up with some questions.

GT made the gear with triangle spokes, and a thinner spokes and rim than I would have liked.  Is there a way to vary the spoke parameters and type of spokes?  Tried changing them in GM and resaving, but it didn't seem to make a difference.

I had to rescale the stl file from inch to metric (or print a very tiny gear).  When I did the printed gear diameter was the same (very slight variation, but I've gotten much better at calibration).  However, the gear thickness was 2/3 of the one I printed back last August.  Not sure what this means; my Z calibration hasn't changed since it's allthread and I used the calculated steps/mm all along.

The original stl file crashed slic3r 0.9.9, both unscaled and scaled in slic3r.  Netfabb said it was fine.  I used meshlab to scale and remesh it, went from 294284  to 147140 faces, this one sliced OK and is the one I printed.

The first time I exported the stl I did it across the network to the notebook connected to the printer, and it took about half an hour.  When exported to the local hard drive it took well under a minute, and only a few seconds to copy with windows explorer.

Not GT's problem, but the teeth look more arrow shaped than I'd like.  It's probably a no support/overhang/cooling issue.

Don't laugh too loud; here's the pix.  The white gear is the one from last year, the red one's tonights, the lighting's not very good so I hope they turn out.  It's printed with 0.10mm layers, 0.3mm homemade nozzle, and in PLA.

The black gear is one that shipped with my printer.  I'm waiting until GT can make one with a hex insert and maybe reduced center.  Don't care about the herringbone...

Kirk

ArtF wrote: But, it appears to me those problems are primarily of extruder size,
which is now down to .3mm or so..

I've made nozzles as small as 0.2mm, and others are down to 0.1mm.  The hot end on my Printrbot uses a 1/4-20 external thread so brass acorn nuts are a good starting point.

It's been good for carbide drill bit sales, broke several in my 0.2 to 0.5 set.  Fortunately it had two of each size, still have one of each remaining.

Kirk

Looking forward to seeing them when yiu are able.


Thanks

Ken:

  Ill try to post some more pics of the things Ive printed tomorrow.
From what Ive observed so far, the problem isnt the process, I too was not
impressed by many of the web examples I ve seen...and its why I held off
so long. But, it appears to me those problems are primarily of extruder size,
which is now down to .3mm or so.. and the methods the software uses for things
like supports and such. TO make the support easily removable I suspect took awhile to
develop. Also doing hollow itesm took some thought..but the software is pretty seamless
at that as well. They claim of course its due to their special ABS.. I disagree, in fact I lowered
the temperature of mine with a resistor in the sensing circuit, and am now using normal ABS
from repraps..same quality really. So its mostly software which woudl differentiate between a
really good 3d printer and an OK one. Luckily the one I picked seems to have pretty good software.
  Ive heard of much tinkering that has to be done, but I havent found that as yet. All but one print
has turned out perfect and its totally seamless for the mmost part. I print from my desktop,
the unit beeps and anywhere from 20 minutes to 3 hours later it beeps again and you open
the door to your piece. Removing supports has been as easy as using a fingernail for the most
part. I printed a bearing yesterday...loose balls, but it rotates freely. ( Bearings will be in GT..)

  Anyway, Ill see if I can post more photos of what I print as I go..

Thx
Art