SoliForum - 3D Printing Community — A technique for superior prints

Re: A technique for superior prints

2015-01-19T03:46:12Z

Not to get off on too much of a tangent on the brush, but, this version moves it in more to try and take advantage of the side-to-side wipes. Also a bit less material, and, I opened up the counterbores on the bottom a bit. I may use a shorter section of brush and move it in even more.

Back on topic...

Re: A technique for superior prints

2015-01-17T02:49:24Z

You don't have to change any gcode with the brush. Currently the head will not make much contact with the brush with its left to right wipe, but I think you'll find the back and forth cleaner motion with the brush is far superior to the wiper anyway, and that the left to right motion isn't that important.
anyway... you can always change back to the wiper if you want... you won't even void your warranty!

Re: A technique for superior prints

2015-01-17T02:27:13Z

I didn't occur to me to ask...if I change out the wipers for the brushes do I have to have the ability to change a build file or anything like that...or can I just swap them out and replace and let the printer do its thing same as before. Novice user here and not familiar with Gcode or anything like that.
Thanks

Re: A technique for superior prints

2015-01-17T01:54:01Z

Just added the brush and holder. Seems like a great upgrade! Thanks for posting.

Re: A technique for superior prints

2015-01-15T23:53:22Z

I completely removed the wiper box when I added the e3d mod and didn't miss it because it seemed to just smear the head but I like your brush idea. May try to rig a way to put a shorter mount in place of the box now and add your code for cleaning. Thanks.

Re: A technique for superior prints

2015-01-15T23:16:15Z

And for anyone interested here are the STL files for the 3 objects...

Re: A technique for superior prints

2015-01-15T23:11:35Z

boxcarmib wrote:

Here's the .fff file

Thanks. I am going to give it a try tonight.

Re: A technique for superior prints

2015-01-15T23:01:44Z

Here's the .fff file

Re: A technique for superior prints

2015-01-15T21:10:30Z

Nice write up. Could you export the .fff file and post that?

Re: A technique for superior prints

2015-01-15T19:26:39Z

Thanks!

Re: A technique for superior prints

2015-01-15T18:41:50Z

Sorry... I meant to post the file. Here it is. It requires you take a welder's brush (http://www.homedepot.com/p/Lincoln-Elec … /100341127) square up the end and then cut the brush length to about 40~42 mm in length.
The holder is a replacement for the wiper blades which needs to be unscrewed and this holder put in its place. Because of the overhang on screw holes, you'll need to scrape out any filament hanging around in the screw hole or it won't seat properly or be at the right height.
I'm working on a complete drip tray replacement that will allow you to drop in a wire brush without any cutting. The handle is long but it slopes down so it wouldn't be in the way, and who wants to cut wood anyway? If I wanted to cut wood I'd be a woodworker for a hobby, not a 3d printer maniac.

Re: A technique for superior prints

2015-01-15T18:19:37Z

well done, I like the brush scraper idea...did you print the "box" the brush is sitting in? If so care to share the file?

A technique for superior prints

2015-01-15T18:06:59Z

How these tests were performed:
All tests on DaVinci 1.0 running 1.1.J firmware using XYZ natural filament with a bed temp of 80 and extruder temp of 220.

T1 series: XYZware 1.1.1
default configuration except for .2 mm resolution

T2 series: Simplify 3D
default DaVinci 1.0 configuration except for .2mm resolution

T3 series: Simplify 3D
default DaVinci 1.0 configuration except for .2mm resolution
turning off extruder slowdown on short layer times
addition of layer change gcode to send the extruder for a walk and a clean

In spite of all the claims of superior performance using Simplify3D, I just don't see that much of a difference. HOWEVER, S3D's ability to add layer change code allows one to add gCode... that of course is something XYZware doesn't and never will allow you to do.

The most dramatic difference is in the printing of the cone. As many of you have experienced, pointy ends result in existing layers not having enough time to cool before the next layer is extruded leaving a gob at the top of your print where a small point should be. The solution? Send your extruder for a walk. Originally I just added gCode to walk around the perimeter of the bed, but oozing led to some pretty rough edges, so I then elected to move the printer to the cleaning blades and run it through a cleaning cycle. This provided superior results for the cube and the dovetail joint, but left a lot of artifacts on the cone. Granted these threads are easily removed with the scrape of a utility blade, but in an attempt to reduce these artifacts as much as possible I replaced the DaVinci wiper blades with the wire brush head from a welder's brush. While it did not completely remove the artifacts (the ones that remain are the product of the extruder leaving the print, not from returning from the extruder wipe). If I understood retraction better I might have more success in reducing the exit artifacts.

Although it is not as apparent in the pictures, my technique removes the bowing in the first 5 or 10 mm of a print. And for me personally, the flat top of the dovetail joint is great news indeed as my parts now fit together off the bed and reveal impressive accuracy and tolerances.

This technique does slow things down of course, but with S3D's ability to chain processes I can select to only send the extruder for a layer walk for the bottom and the top 10 mm. This seems to provide consistent superior results.

I suspect my results are what would be available if one has a print cooling fan, but I haven't seen a practical solution for adding one to the DaVinci yet. In the meantime, I'm happy with the results I'm getting and I'd far rather have prints with a few artifacts that can be easily scraped off then failed print after failed print.

S3D's attempt to control this problem is to give you the option of slowing down the extruder for layers with short print times, but for prints like the cone, this only makes matters worse as it just keeps the hot extruder pressed against the print and actually getting the existing layers hotter and hotter. If you include my extruder walk code for any portion of your print, make sure you turn off the "slow down extruder for small layers" option.

Finally, the lack of a fan, or more innovative approaches to this problem is a very solvable problem, the physics and the math are well understood. Here's a great link to a description of the whole phenomenon that a slicer could incorporate into its layer calculations.

http://cdn.intechopen.com/pdfs-wm/21961.pdf

Here is the gcode I add to S3D's script for layer change code: (If you are using the DaVinci's wiper blades, then uncomment the left to right wipe lines. Personally I feel the replacement of the wiper blades with the wire brush is a great improvement and I don't use the wiper blades at all anymore.

G1 X-45 Y10 F4000 ; move to a safe place, set speed as desired
G1 Y200 move head to front for manual cleaning if desired
G1 Y10
G1 Y200
G1 Y10 ; move back and forth to scrape
G1 Y-15
G1 Y10
G1 Y-15
G1 Y10
G1 Y-15
G1 Y10
G1 Y-15
G1 Y10
; G1 X-25 F1800 left and right to scrape
; G1 X-45
; G1 X-25
; G1 X-45
; G1 X-25
; G1 X-45
; G1 X-45 Y100 F4000 ; move head to safe position