Advanced Search

tommylight wrote:

Can you power off the machine and leave only Mesa powered?

I turned everything off except the PC and it no longer disconnects on its own and it doesnt show my mesa connection. But when I turn the mesa board and the machine on it starts disconnecting again every 45 seconds. So now I think you guys have a good point and maybe this could be hardware related.

Also make sure all power management is disabled on the host PC
Similar issues can come from Ethernet (EEE) or PCIE power management

I had a problem with a Mesa board disconnecting a couple years ago.

Turned out it was a bad/weak DC-DC power supply.  Voltage would drop intermittently, causing the Mesa to reset and I'd get the 'error finishing read' nightmare.  Looked like a network issue, but @PCW helped me diagnose it as a power problem.

Maybe double-check whatever is powering the Mesa and see if it's dodgy.

Can you power off the machine and leave only Mesa powered?

I did not try another LAN cable.
The machine and PC are grounded properly.
I dont have a DVM at the moment but as soon as I get one I will measure those.
I believe this is a software issue because even with qtplasmac not running I get the disconnected error every 45 seconds
 

Did you try another LAN cable?
Is the machine and the PC grounded properly?
Use a DVM and measure between Mesa 24V GND and machine case, in AC mode, while the machine is ON, post result.

 

Sounds like host or 7I76E issues

Whats does

ip a

report?
 

 

What does:

sudo chrt 99 ping -i .001 -q -c 60000 [card_ip_address]

report? (it will run 1 minute then print statistics)

[card_ip_address] is either 10.10.10.10 or 192.168.1.121 depending on 7I76e jumpering

I don’t think it is Plasma/EMI related. My plasma cutter was not connected or turned on.

Is this possibly Plasma/EMI related?

That is, do you still get communication loss when the Plasma source is off?

  I ran pnconf wizard and made a configuration for my CNC plasma table. I am able to home and jog my machine for about a minute until I get this error. 
It seems like the ethernet connection is disconnecting on its own somehow. Any ideas on how I could get the ethernet to stay connected so I can use my machine.

Here is the full file

loadrt mux2 names=mux2_x,mux2_y,mux2_z
addf mux2_x servo-thread
addf mux2_y servo-thread
addf mux2_z servo-thread

loadrt wcomp names=plus_x_trigger,minus_x_trigger,plus_y_trigger,minus_y_trigger,plus_z_trigger,minus_z_trigger
addf plus_x_trigger servo-thread
addf minus_x_trigger servo-thread
addf plus_y_trigger servo-thread
addf minus_y_trigger servo-thread
addf plus_z_trigger servo-thread
addf minus_z_trigger servo-thread

# this is the deadzone for each axis to stop drift, also can make bigger to stop x/y axis moves easier without one triggering the other
setp plus_x_trigger.max 0.15
setp minus_x_trigger.min -0.15
setp plus_y_trigger.max 0.15
setp minus_y_trigger.min -0.15
setp plus_z_trigger.max 0.15
setp minus_z_trigger.min -0.15

loadrt lut5 names=jog_x_enable,jog_y_enable,jog_z_enable
addf jog_x_enable servo-thread
setp jog_x_enable.function 0x80
addf jog_y_enable servo-thread
setp jog_y_enable.function 0x80
addf jog_z_enable servo-thread
setp jog_z_enable.function 0x80

loadrt or2 names=or-x-jog,or-y-jog,or-z-jog
addf or-x-jog servo-thread
addf or-y-jog servo-thread
addf or-z-jog servo-thread

loadrt scale names=scale.x,scale.y,scale.z
addf scale.x servo-thread
addf scale.y servo-thread
addf scale.z servo-thread

# mine were 0-4.8v. Assuming a range of -1 to +1. 2/4.8=.416 z was way off balance and cannot see how to adjust so do it in sw
setp scale.x.offset -1.00
setp scale.x.gain .416
setp scale.y.offset -1.00
setp scale.y.gain .416
setp scale.z.offset -0.80
setp scale.z.gain .416

net x-scale <= hm2_7i76e.0.7i76.0.0.analogin0 => scale.x.in
net y-scale <= hm2_7i76e.0.7i76.0.0.analogin1 => scale.y.in
net z-scale <= hm2_7i76e.0.7i76.0.0.analogin2 => scale.z.in

# the mux is used to select 0 or the current value and is implementing the deadband. Cannot let the analog value through otherwise in interferes with jogging from ui
net gate-x-scale <= scale.x.out => plus_x_trigger.in minus_x_trigger.in mux2_x.in1
net gate-y-scale <= scale.y.out => plus_y_trigger.in minus_y_trigger.in mux2_y.in1
net gate-z-scale <= scale.z.out => plus_z_trigger.in minus_z_trigger.in mux2_z.in1

# may not be needed anymore
net x-is-homed <= halui.joint.0.is-homed => jog_x_enable.in-1
net y-is-homed <= halui.joint.1.is-homed => jog_y_enable.in-1
net z-is-homed <= halui.joint.2.is-homed => jog_z_enable.in-1

# only allow jogging if in the jog menu linuxcnc/gmoccapy seemed to complain in any other state
net machine-is-injog <= halui.mode.is-manual => jog_x_enable.in-0 jog_y_enable.in-0 jog_z_enable.in-0

# fixed for all axis multiplied from .15 to 1 from scale above. Change scale is difference needed for each axis
setp halui.axis.jog-speed 600
setp halui.joint.jog-speed 600

# wcomp is used here as deadband. Controls relevant jog plus or minus
net enable-x-jog-plus <= plus_x_trigger.over => halui.axis.x.plus or-x-jog.in0
net enable-x-jog-minus <= minus_x_trigger.under => halui.axis.x.minus or-x-jog.in1
net enable-y-jog-plus <= plus_y_trigger.over => halui.axis.y.plus or-y-jog.in0
net enable-y-jog-minus <= minus_y_trigger.under => halui.axis.y.minus or-y-jog.in1
net enable-z-jog-plus <= plus_z_trigger.over => halui.axis.z.plus or-z-jog.in0
net enable-z-jog-minus <= minus_z_trigger.under => halui.axis.z.minus or-z-jog.in1

# if all 3 conditions are met i.e. not in deadband, is homed and is manual then select mux and route analog value to axis
net jog-x-either <= or-x-jog.out => jog_x_enable.in-2
net enable-x-mux <= jog_x_enable.out => mux2_x.sel
net jog-x-speed <= mux2_x.out => halui.axis.x.analog

net jog-y-either <= or-y-jog.out => jog_y_enable.in-2
net enable-y-mux <= jog_y_enable.out => mux2_y.sel
net jog-y-speed <= mux2_y.out => halui.axis.y.analog

net jog-z-either <= or-z-jog.out => jog_z_enable.in-2
net enable-z-mux <= jog_z_enable.out => mux2_z.sel
net jog-z-speed <= mux2_z.out => halui.axis.z.analog