Advanced Search

That looks like the standard realtek 8168 driver rather than the DKMS driver
so it sounds like the DKMS driver install did not work for some reason

If I understand this correctly, it is trying to read the data starting at address 0305h?
If thats true, I think that address does not exist in my controller.

So what I am taking away from your wonderful description is this :
Comms are happening in both directions, and are understood by each side.
MAYBE the actrual issue here is the register to say enable spindle is the incorrect register in my drive, so the response is illegal data address?

Not sure about the coil observation yet.

Is it odd that there is only 6 bytes received then a time out occurs? Could I have some type of issue with the comms and data is getting corrupt? Again I want to point out this is the same setup I had before on a different VFD and worked fine. I literally disconnected the wires from 1 and connected them to the other.

I want to switch from normal HD to ssd. There is no other change in Hardware, neither the PC nor the CNC machinery.
It works over parallel port and M542 driver as for years. The mill itself is a DIY 3 axis machine.

I downloaded the current Debian Amd64 2.9.3 PREEMPT-RT Hybrid ISO and installed it.
Also I did create a new config for Gmoccapy/machine, my current config is just to old to get converted (got a lot of errors). Anyway, the installation went OK and the machine is working as expected.

Now I did an upgrade
which in general also went OK. Paranoid reboot and start LinuxCNC. Now I got the issue all but Y-axis moving. The motor gives some noise but not really moving. Whatever I tried, changing/relaxing timings and so on, it did not start to work correctly. X- and Z- still working.

I did search but did not found similar issue messages.
The question is, may someone had an easy fix or should I just reinstall and avoid upgrade for some time?

Thanks for reading,
DeJe

i added the same as the google doc "r8168.aspm=0 r8168.eee_enable=0 pcie_aspm=off loglevel=3"
i changed and add " pcie.aspm=0 r8168.eee_enable=0 pcie_aspm=off idle=poll" but still the same "hm2 error finishing read" (I restarted the PC)

 

Not sure how that gave you the '>' prompt but:

I would think that you need to replace this with the path to your local clone of the git tree from this earlier command:

If you intend to drive a small DC motor directly from the 7I96S outputs,
then yes, you need a flyback diode across the motor.
The motor also should not have more than 2A of stall current.

This tells you the following:

1. The component is sending a request to the vfd starting with '[01] [01] ..'
ie it's calling device with address '01' and the function code '01' (ie read coil status) ....

2. In modbus protocol a response starts with echoing the device address ie '01' followed by the functionn code, ie also '01' in your case.

3. Note that the response does not start with [01],[01] but [01], [81]. The '81' tells you that the device has sent you an exception response to your 01 query. The next byte in the response, ie [02], is the exception code (which would indicate that you requested and illegal data address ).
www.simplymodbus.ca/exceptions.htm

Not sure but it seems a bit odd to me to read coil status [01] from address [03] [05] but then have the total number of coils [00] [00]
www.simplymodbus.ca/FC01.htm