Solution fo "hm2 error finishing read" with no good PC

lspci | grep -i ether sudo nano /etc/network/interfaces

auto eno1
    iface eno1 inet static
    address 10.10.10.11
    netmask 255.255.255.0
    hardware-irq-coalesce-rx-usecs 0  <<<< try adding this
 

I added the line

I restarted the PC

I still have the same fault

I set my configuration for the tool change because this is where the fault comes at 100% if there is something wrong.

Do all inductive loads switched by the 7I97T have flyback diodes?

Did you verify that IRQ coalescing is disabled?
verify with:

ethtool -c eno1

A a temporary test you can try pinning IRQs with

 

download,

chmod +x pinirq.txt

sudo ./pinirq.txt en01

and then run LinuxCNC
 

I think I know where the problem is.
for a start, i hadn't fitted a free-wheeling diode.
i've just put diodes on all my control contactors.



i've tried again and i have the impression that it's got worse because when i launch linuxcnc and deactivate the e-stop (F1), the message appears more regularly.

I try to control my chip conveyor, the message appears.
i try to drive my tailstock, the message doesn't appear ...

I realize that the commands that cause error messages are those driven by power contactors.

However, all my power contactors are controlled by command contactors.

sinking output:


drive with power contactor
e-stop / fwd turret / rev turret / convoyor fwd / convoyor rev

pilot with control contactor
tailstock output / tailstock input / enable drive / spindle fwd / spindle rev

I've never had a problem with controls driven by control contactors alone.

the video shows when i manually activate (with the control contactor) my turret. only the card is activated, the computer is not powered so linuxcnc doesn't work.
on the video i activate the contactor three times and 2/3 error and 1/3 no error



how can my power contactors disrupt the 7i97T card?