#linuxcnc-devel Logs

May 08 2017

10:39 AM skunkworks: jepler, it really is amazing. I had pretty low expectations going into this. It works quite well - delightful. I don't think I have had a failed print yet.
10:40 AM skunkworks: (that wasn't my fault)
10:40 AM skunkworks: I just found out my cousin has 3 of these things. 2 v1 and 1 v2
10:40 AM skunkworks: she loves them
10:41 AM skunkworks: Telling me all the mods that need to be done. (cables under the table need to be dressed up better)
11:03 AM jepler: 3 of them? wow.
11:04 AM jepler: I have seen other people talking about problems with the cables under the table. I haven't looked
11:04 AM jepler: what have you been printing so far?
11:12 AM skunkworks: toys ;)
11:12 AM jepler: skunkworks: did this cousin have any advice about the problem you were seeing when you homed Z?
11:13 AM skunkworks: Have not had a chance to ask her. But I have not needed to jog it since the first day ;)
11:16 AM skunkworks: I have been scarilly hitting print and walking away.
11:17 AM skunkworks: scarily
11:18 AM skunkworks: tops, elephants, robots, so on.. (thingverse files)
11:22 AM jepler: fun
11:22 AM skunkworks: you are running cura on linux?
11:23 AM jepler: yes cura 2.3 on debian jessie amd6
11:23 AM jepler: 4
11:24 AM skunkworks: cool - I didn't know it had a linux version. my first attemt to run it in a 7 virtual machine faild and I ended up booting the laptop to the windows partition.
11:25 AM skunkworks: is it as easy as apt-get installing it?
11:25 AM jepler: apparently I'm behind the times, their download page is showing cura 2.5
11:26 AM jepler: I would have told you I downloaded a .deb file from ultimaker.com but now I'm not sure
11:26 AM jepler: their download page is giving me a ".AppImage" file today, which I don't even know what that is
11:28 AM jepler: yeah somewhere between 2.3.1 (the version I have) and 2.5, they switched from a .deb to a .AppImage
11:28 AM jepler: https://ultimaker.com/en/products/cura-software/list has old versions
11:28 AM jepler: direct download link of 2.3.1: http://software.ultimaker.com/current/Cura-2.3.1-Linux.deb
11:30 AM jepler: direct download link of 2.5.0: https://software.ultimaker.com/current/Cura-2.5.0.AppImage
11:30 AM jepler: "Works with most reasonably recent desktop Linux distributions. Well, almost." proclaims appimage.org
11:33 AM jepler: the appimage seems to work on my debian jessie. it didn't import my settings from 2.3, though, which is a drag.
11:33 AM jepler: even worse, it seems to have eaten my 2.3 settings. argh.
11:34 AM skunkworks: oops.. Sorry for starting that..
11:34 AM jepler: it's not your fault
11:35 AM skunkworks: Do you use the 'magic numbers' for layer hight?
11:36 AM jepler: sometimes I remember to
11:36 AM jepler: I have done some comparison prints and to my surprise it does seem to make a difference
11:39 AM jepler: I think the microstepping is pretty inaccurate/ineffectual, maybe particularly on Z
11:44 AM jepler: anyway, I was convinced after running a .200 print and a .21875 print, some pretty severe layer-to-layer variations in the .200 print were fixed in the .21875 print
11:51 AM jepler: if you ignore microstepping, there's a 7.15% variation between the two closest Z increments which would repeat irregularly over the full height of the object
11:52 AM jepler: if you print something in "vase" mode (Z increases slowly and continuously instead of all at once) and open up so you can see the Z axis nut, you can watch the individual full steps occur even though the gcode is commanding much smaller increments than that
11:55 AM jepler: I sure wasn't seeing 16 discrete movements for each facet of the hex nut, so I must have been seeing 8. Even the half steps must be ineffectual/inaccurate.
06:39 PM JT-Shop: pcw_home: are you out of stock on the 7i42TA?
07:45 PM pcw_mesa: Yeah for a week or so