- June 17, 2017 at 1:40 am #36204
I have built and configured the MPCNC as per this site instructions. I have been trying to get my first full CNC job completed (I am building a 737 flight simulator and busy making the 737 panels and the TQ). I am using Estlcam 9 for generating the GCode and Repetier for interacting with the MPCNC machine. What I have found is that if I do multiple Z steps (e.g. I go 2mm at a time to cut through 6mm MDF), the first step round is perfect, however the moment the 2nd step round starts, it is not following exactly in the 1st step round cut-out. I have re-checked all the screws to make sure that the MPCNC is as rigid as possible. It seems as with the pressure exerted by the router bit on the MDF, the MPCNC slowly gets out of synchronisation/alignment? Using 24000RMP, 60mm/s feedrate and 20mm/s plunge rate.
Can anybody help me to resolve the issue?
Attachments:June 17, 2017 at 2:14 am #36209
Have you tried reducing your feed rate significantly?
60mm/s seems very high to me.June 17, 2017 at 8:54 am #36211
You should update your estlcam, or at least make sure it is above version 9.37. For these reasons.
Did you do a test drawing, using my gcode, how did it look?
Mike is correct 60mm/s is too fast, far far too fast. Watch some of my videos and looks at the settings I use.
And 20mm/s on the Z axis will crash your controller or just skip steps. The max is 8.4mm/s and if you look through the first link I recommend around 4mm/s for the z axis.
I just did a bunch of 4mm per pass cuts at 15mm/s in pine and am confident I can go much deeper per pass but am running these test to verify that.. To speed up you cut you do not want to move the machine faster, but actually slower and deeper per pass, the less passes the faster it finishes by a significant margin.June 17, 2017 at 10:34 am #36212
That sounds like a fun project. The others are right that you should reduce the speed to start with. Once you know what it looks like when it works, you can increase the depth per pass out the speed. I personally like to go deeper, because it the bits wear more evenly, instead of only filling the first 1mm.
Another silly question, the board didn’t move, right? It was well clamped? I am only suspicious of the skipped steps theory because the errors look so smooth. I would have expected it to skip a bunch of steps at once, and none for a while. But it is possible.
You must be logged in to reply to this topic.