Page 1 of 1

Test drawing with my DIY CNC using UC300ETH and UCCNC

PostPosted: Sat Nov 12, 2016 11:50 am
by A_Camera
Not made much "things" so far using UCCNC and UC300ETH, only some tests, but I am pretty pleased with the results.

Here is the latest video, 6 minutes long and a test drawing with a lot of details.

[youtube]https://youtu.be/_cJf9ZBYqo8[/youtube]

Re: Test drawing with my DIY CNC using UC300ETH and UCCNC

PostPosted: Sat Nov 12, 2016 12:45 pm
by cncdrive
Hi Josef,

In youtube right click on the video and select "Copy embed code" and then copy that in between the youtube tags.
Like this:


Re: Test drawing with my DIY CNC using UC300ETH and UCCNC

PostPosted: Sat Nov 12, 2016 1:42 pm
by A_Camera
cncdrive wrote:Hi Josef,

In youtube right click on the video and select "Copy embed code" and then copy that in between the youtube tags.


OK, thank you Balázs. I though it was just the link, not the HTML code which needed to be embedded.

Re: Test drawing with my DIY CNC using UC300ETH and UCCNC

PostPosted: Mon Nov 14, 2016 10:15 am
by A_Camera
Tried with the same speed parameters on Mach3. This is what happened:



...apart from the fact that the letters are in fact uglier when Mach3 is drawing the same test drawing. Look at especially the E, M, a, D, N... well, almost all the letters look pretty ugly. But I'll make a separate video about the differences in the results.

Re: Test drawing with my DIY CNC using UC300ETH and UCCNC

PostPosted: Mon Nov 14, 2016 2:35 pm
by cncdrive
The letters are more "ugly", because Mach3 has problems with it's trajectory planner.
The lower your acceleration will be the worse this will get and the larger the difference of acceleration is between your XY axis (because those axis plays in this code interpolation) again the worse it will get.
Also you don't know how much error the machine will make on the path in CV mode, there is no parameter to define it in Mach.

With the UCCNC there is no such problem, the trajectory planner has no defects like this, the CV controller has path deviation parameters,
so you can exactly define how much the software can go off the path in order to make the job as quick as the CV parameters allow.
The larger you set that the faster the software will finish, but the worse the tolerance will be and the lower you set the CV deviation the slower it will finish,
but the better the tolerance will be. You can define the deviation in the working units (mm, inches, etc.), so you can exactly set the wanted tolerance for different jobs depending on if a job needs accuracy or speed or in the midway. :)

Re: Test drawing with my DIY CNC using UC300ETH and UCCNC

PostPosted: Tue Nov 15, 2016 9:53 am
by A_Camera
Thank you Balázs. Yes, I noticed this the first time I used UCCNC with the same G-code. When it comes to acceleration, with my machine Mach3 can't handle faster acceleration, it causes to crash on X. When it crash, it is always on X deceleration phase. Maybe I reached the limit of what Mach3 can handle with my machine in terms of speed and acceleration. My machine uses direct driven 1605 ball screws and there is an upper limit in possible RPM and acceleration, but since UCCNC can manage higher values, the limitations of the hardware is not the issue with Mach3. Maybe one day I hook up an oscilloscope and compare the differences, but for now I don't have the time. Never the less, I don't think Mach3 will ever be able to track as nicely as UCCNC because of the better trajectory planning, so regardless what I do, I think UCCNC will always provide better results, at least if CV is used, like in my test drawings.