> From: Viesturs Lacis [mailto:viesturs.la...@gmail.com] > Hello! > > I have a question about a machine that I converted to LinuxCNC 5 or 6 > years ago. It is a 4 axis mill that is used for making jewelry - > milling wax parts that are then turned into nice things with lost-wax > technique. > > The issue is that Z axis is losing height. That is best observed, when > the milling is done with rotary head. Here is an example - the height > difference of first passes vs last passes is approximately 1 mm: > https://www.picpasteplus.com/v.php?i=507b0d6be4 > > The thing is that the client told me about this problem few years ago, > but then it somehow disappeared or at least the position error was not > that huge and it was possible to fix the wax part before proceeding. > > Machine is running 2.7.4 version. Here are the INI and HAL files: > https://pastebin.com/N7tk8nnW > https://pastebin.com/epyUVbpp > > I am convinced that all the possible mechanical causes have been > checked. For the final check, we brought Z axis to particular > position, marked motor shaft placement regarding motor case, then did > the milling and afterwards brought it back to the same position - > marks on motor shaft _did not_ match the case. > > I changed the feedrate and acceleration values 2x lower just to make > sure that machine was not missing steps. Or is it still missing steps? > > So I would like to ask for some ideas, what else should I try and where to > look. > > Viesturs >
It's possible that your couplers are slipping on the shaft. I see in the INI file you don't have home operations enabled on Z? Without that absolute physical reference you can't know for sure that you have stepper motor skipping or a slight rotation of motor to transmission. That a return to the same co-ordinate is different mechanically doesn't rule out mechanical slippage. For example. If you had a home switch and homed again after the missed Z position and then found it was back to the same target position but the mark relative to the motor case is now different then you have physical slippage. If after homing and movement to the target position and the marks once again line up then you are losing steps. If that’s the case determine if it's a full step (1/200) of a motor turn or is it a fraction of that. It's possible, as Gene is so fond of saying, that your optical isolators have aged and are slowing down and a step doesn't quite make it through although at 10uS per step that's likely not an issue. Something to look at... John Dammeyer _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
