Ethercat HAL driver

sdfzz wrote: And line 380 of linuxcnc/src/emc/motion/homing.c is commented out,

Oh. No wonder I didn't guess!

The issue about the homing in the repos github.com/aschiffler/linuxcnc is fixed

The intention to modify the homing source code was about the following cases:
- absolut / multi turn encoder --> HOME_SEARCH_VEL = 0.0 HOME_LATCH_VEL = 6.6; axis is homed at actual position feedback value; no movement
- drive driven homing sequence and switches connected directly to drives --> HOME_SEARCH_VEL = 0.0 HOME_LATCH_VEL = 9.9; NC waits for home-sw-in and assumes movement by the drive it self; axis is homed to position HOME_OFFSET (home-sw-in HAL pin is connected via EtherCAT to corresponding bit in status word of the drive)
- (Standard) NC driven homing sequence --> HOME_SEARCH_VEL != 0.0 HOME_LATCH_VEL != 0.0; NC commands the joint; axis is homed to position HOME_OFFSET
- (Standard) No homing needed / simulation --> HOME_SEARCH_VEL = 0.0 HOME_LATCH_VEL = 0.0

roschi wrote: - absolut / multi turn encoder --> HOME_SEARCH_VEL = 0.0 HOME_LATCH_VEL = 6.6; axis is homed at actual position feedback value; no movement
- drive driven homing sequence and switches connected directly to drives --> HOME_SEARCH_VEL = 0.0 HOME_LATCH_VEL = 9.9

Magic numbers? Not normally recommended, because somebody might choose those numbers deliberately.

(I suppose that 99999999 and 66666666 would be OK, as nobody will ever home that fast.)

Roschi,

Thank you for your reply. I was aware of 6.6 and 9.9 settings.... I saw it from your previous comments. I guess it is a way to select the desired homing mode, although without a proper documentation, it may cause confusions.

Also, thank you for the fix.

Regards,

Steve

Sascha,

I tried your new project. Installation was much easier, and now I can jog and home my axis.

Thank you!

the first error :Mailbox corrupt received

I've encountered this issue two month ago.
you need to contact the drive manufacturor and update your driver version to 2.0.0