Advanced Search

I think you need to specify a home position that's some distance from the home/limit switch location

linuxcnc.org/docs/html/config/ini-homing.html#_home

Hi All

Thank you for your feedback, greatly appreciated
The Y gantry completes the four step process now, well ... very nearly.
I think my problem was here
[AXIS_Y]
[JOINT_1]
has negative values
HOME_SEARCH_VEL = -10.000000
HOME_LATCH_VEL = -5.000000
(and joint 2 had the same negative values).
[JOINT_2]
I changed the signs to both positive
HOME_SEARCH_VEL = 10.000000
HOME_LATCH_VEL = 5.000000
After that change, the Y-axis homed in the four step way.
I don't understand why,  maybe gantry motors that face one another need opposite signs, I have no idea.
I have attached three screen photos with halshow
first - just switched on power supplies, and launched Linuxcnc.
second - after clicked estop, and clicked power.
third - end state after all thee axis homing has completed.
Physically the Y gantry does the correct four step movement.
However, Linuxcnc errors out right at the very end, indicating joint2 on limit switch error.
The  y-is-homed is red , while x-is-homed, y2-is-homed and z-is-homed are all yellow.
These are the cheap 2mm proximity sensors , tommylight said earlier that these are best avoided.
I will buy better quality ones.
If anyone knows why that sign change worked please let me know


 

Hello Aciera, Thank you for your replay , The orignal GMOCCAPY  I used on RPI4 is OK, but on RPI5 is not OK

I am new to LinuxCNC and  ProbeBasic, but not to Linux or Python.

I want to modify the "Status" tab to be similar to what  anfänger did in the thread forum.linuxcnc.org/qtpyvcp/39734-show-your-screens#177730 (see below).   I also want to change so the "Main" tab is the default on start-up

Is there a getting-started guide or notes?  I reviewed the ProbeBasic docs kcjengr.github.io/probe_basic.
Thanks!!!

I upgraded via backports to 6.12.22+bpo-rt-amd64, still get a 5ms max delay.

I ran it a few times
rtt min/avg/max/mdev = 0.053/0.116/5.600/0.269 ms
rtt min/avg/max/mdev = 0.054/0.099/4.908/0.216 ms
rtt min/avg/max/mdev = 0.054/0.106/4.672/0.241 ms

I've not had a warning since I ran the IRQ script.

We will see once I start actually cutting

ooh, nice. you think you can help me config my ethercat setup. i have different drives, but im stuck at the "how do i even make ethercat work" part. 

i looked through your files a bit, but of course they dont start cause i got no mesa card in mine.

# net machine-is-on => hm2_7i92.0.stepgen.07.enable

AFAICS, nothing sets hm2_7i92.0.stepgen.07.enable true so it can never run.

try removing the comment (#) :

net machine-is-on => hm2_7i92.0.stepgen.07.enable

The attempts to run the spindle are by using:

M3 S10 $0 <=== that works
M3 S10 $1 <=== does not work

I am trying to configure LinuxCNC for use on a rose engine with 2 spindles + 6 other axes.

Everything works for Spindle0 & the other 6 axes of movement (X, Y, Z, B, U, & V), however I cannot get Spindle1 to work. Any ideas would be greatly welcomed!

I've copied the key items I believe are needed below.  The complete versions of the HAL and INI files are attached.  (The include files for the spindles had to be renamed to be .ini for attaching them, but they are actually .inc.)

hal fileloadrt [KINS]KINEMATICS
loadrt [EMCMOT]EMCMOT servo_period_nsec=[EMCMOT]SERVO_PERIOD num_joints=[KINS]JOINTS num_spindles=[TRAJ]SPINDLES
loadrt hostmot2
loadrt hm2_eth board_ip="192.168.1.121" config="num_encoders=0 num_pwmgens=0 num_stepgens=8"

loadrt pid names=pid.x,pid.y,pid.z,pid.b,pid.u,pid.v,pid.s0,pid.s1

(other sections are also configured for SPINDLE 1, and are setup fundamentally the same as they are for SPINDLE 0)
 ini file[TRAJ]
SPINDLES = 2

# [JOINT_7]
# SPINDLE = 0
# TYPE = ANGULAR

[SPINDLE_0]
(this works as expected, so I've not added the details here)

[JOINT_6]
SPINDLE = 1
TYPE = ANGULAR

[SPINDLE_1]
(the rest is setup just as with SPINDLE_0)
 Side Note: Please don't refer me to RoseKins.  I tried that and it will not meet our needs.

mjoconr wrote:

--- 10.10.10.10 ping statistics ---
60000 packets transmitted, 60000 received, 0% packet loss, time 60401ms
rtt min/avg/max/mdev = 0.056/0.084/7.709/0.139 ms

That is not good, at all.
If the PC has a RealTek network card, you will find a lot of replies here to install the DKMS drivers for it, i would skip that and try a new RT kernel, something like 6.11 or above.

Thanks, it works

besriworld wrote:

At a later stage, I plan to add an additional 1kW servo motor, which will operate in torque mode to act as a dynamic load on the main 7kW motor. This will increase the system's rigidity. Currently, I'm using the encoder from the main motor
 

That sounds interesting.  How are you planning to couple the secondary motor to the spindle?

Are you sure the proximity switches are deactivated (red in halshow) when not at home position?

If these were inverted, the "home" move might actually be LinuxCNC backing off what it thinks is
a closed switch.

Kept looking last night and ran the IRQ script found else where on this forum.
This seemed to improve things, I left LCNC running all night with glxgears and other things running and did not get the messages again.

Ping result
cnc@minimill:~$ sudo chrt 99 ping -i .001 -q -c 60000 10.10.10.10
PING 10.10.10.10 (10.10.10.10) 56(84) bytes of data.

--- 10.10.10.10 ping statistics ---
60000 packets transmitted, 60000 received, 0% packet loss, time 60401ms
rtt min/avg/max/mdev = 0.056/0.084/7.709/0.139 ms

The issue maybe be fixed I'll see how we go actually using it today.

1. make sure both Y spin in the correct direction
2. what sensing distance do the proximity switches have? 1 and 2mm should be avoided. And switches are not mounted correctly, they should be sideways to prevent crashing into stuff.
For 1:
make a normal 3 axis test config, test one Y with a short jog, change drive inputs and test second Y with a short jog.
This can be done on software but more complicated to explain.