Need help with unexpected/random stops on basic Parport/5 axis BOB setup

Hello,
I've been running a rather large gantry type CNC mill, with a 'temporary' basic setup (Axis + parport + basic Bob) for some time. I've experienced random stops since the beginning, with a variable frequency, sometimes none for a few days, but now it's more like a couple per hour.
The system behaves exactly as though I had hit the Toggle Machine Power on Axis. The machine stops and LinuxCNC stops processing the .ngc file. All I have to do is to click Toggle Machine Power and resume working.

My .hal file is set to monitor the Pin10 on the parport to check for E-stop requests, and set iocontrol.0.emc-enable-in to False when the loop betwwen Pin10 and Bob Gnd is open.
I suspected a faulty wiring and/or component in the controller cabinet, and tried a few things :
- moved the e-stop monitoring to another input pin on the Bob
- added a diode+capacitor to act as a reserve power on the Bob 24v power supply, in case I had micro failures
- simplified the connection between PC and Bob
- made sure that the Bob GND was connected to the cabinet GND
- checked all sort of connections
- replaced the Bob with a new, identical one
- tried to monitor two pins (10 & 11) to filter wrong signals. 11 is shunted to GND, and 10 is the e-stop watchdog. If 10 only signals -> e-stop. If both 10 and 11 signal, ignore this 'false alarm'
- I even shunted the Pin10 to the GND on the Bob itself, to eliminate all possible problem downstream
=> no luck
I haven't tried to use another PC, for lack of available suitable machine. I also did not try another parport cable. (although I could have)

The only way I have found to eliminate these uncontrolled stops is to reconfigure my .hal to stop monitoring the external e-stop. This is not satisfactory, as LinuxCNC would not see any possible legitimate e-stop requests from the machine.

Any idea or suggestion of where I should be looking ?

Thank you for reading me, and for your help

Laurent

Hi,

Thank you for your reply.

I should have explained in my first post that I'm far from being proficient at electronics, I have built the control cabinet after a lot of head-scratching, and with the help of a friend.

>Make sure you have a 3 pronged grounded cable for the PC

At the moment, the PC is powered from the control cabinet, where I have 3 phases 400v, neutral and GND arriving (Im in France). From there, we have separated a phase and neutral to get 220v power. So the PC is powered from that, with a 3-pronged power cable, its GND connected to the cabinet GND.

All GND connections are done through a common connection, I don't think there is any daisy-chaining.

I'll try another keyboard/mouse, but can you explain how these could affect the parport ?

Sorry for being such a noob, and thanks again

it may sound Idiotic but change your Mouse
we did get same behaviior on Mill and Plasma as the Mouse USB did get emi like behavior and stoped even the PC completly
try to get the pc to a other powercycle in the house if you got 3phase in house you can use it for devideing the powerlines

Hello,

Thank you for your reply.

I will try :
- another keyboard and mouse
- to power the PC from outside the cabinet
and report here.

Ok, so I'm at the shop now, trying to move forward on this.

Keyboard/mouse : couldn't find a spare USB combo at the moment, and the PC (Dell Optiplex 760) has no PS/2 connections. Tried the next best thing I could think of and unplugged the USB keyboard and mouse after I launched LinuxCNC and got the spindle running idle. Machine stopped after +/- 10 minutes.

Stopped the PC and moved its power supply to a wall socket outside the CNC cabinet, replugged USB keyboard and mouse, restarted everything. I'm now waiting to see if/when it stops running...

One thought : I have added a WLAN card the Optiplex to help set it up, as I had no wired Ethernet available in the office. It's still in the machine. Can this thing create EMI ? Should I deactivate it ?

Thank you for reading me, and for your help
Laurent

OK,
So, I also checked the neutral/phase being possibly reversed : it seems OK.

The machine has been running for a while now, with the PC powered from a wall socket outside the CNC cabinet. But again, this is random, so could stop in a few minutes, or in one hour

My friend suggested that the PC should be powered from inside the CNC cabinet, in particular because the Bob gets its 5v power from the PC USB.
I did not understand exactly why, but I trusted his judgment. What would you say ?