Problem at homing

At short time, the x-Axis and only the X-Axis from my machine don't run permanatly at the homing switch, she stop and starts sporadic.
The X-Axis display value show the same, also stop and start parallel the axis, also no defective contact.
It show's the prozess ist crazy. After homing, there are no problems, the machine run's quickly.

Thx for your support.

Best regards.
Frank

Hey Guys!
Unfortunately, my problem-homing is not yet solved.
I have update my System from Ubuntu 10.04 to 12.04 and LinuxCNC from 2.3 to 2.6.
But this has not yet solved my problem.
The hardware was'nt touched.

I have made a little Video from home run .

!!!! Importen you must hear it and see. !!!!

You can see, on the way to the X-home switch the axis stop randomize.

Is this a cycle problem from the rtail Kernel ? or from the pulsgenerator ?
Can I trace it with the halscope ???

What do you think ???

Thanks form all fuming brain's?

Here the Link to the video.
www.dropbox.com/s/z3y6yg0prfk0ky1/20140918200533.flv?dl=0

That's pretty strange

One guess is noise in the limit switch circuit
(not sure what linuxcnc homing does if confronted with short
apparent limit switch actuations caused by electrical or contact noise)

I would halscope the limit switch signal and search for glitches

are you using any enable connections from your drivers?

i notice it appears to move in reverse fine , so i'd check for noise on your switches as mentioned
you dont mention which breakout your using ? , perhaps if not all ready change your switches to work from 5v rather than ground.
screened cable works wonders for switches .

I see it differently .
In the video you can see that the display stops .
After homing there are no problems in both directions, only at homing.

For me this is a software problem.
I think this is jitter , timing , cycle or similar .
At the moment I can't say details of the computer system .
I have long made ​​no latency test , I will make time today .

I will update you.
THX

It could be a timing issue but if you listen on the video you can hear that the stepgen is running
well and accelerating and decelerating in a controlled manner so high speed timing seems OK
(no thumps or bangs as you would expect with real time issues)

I still think noise in the home switch signal could cause weird homing behavior though
I would expect it to fault in some way.

Are you doing something special for gantry homing?
can you post your hal and ini files, that might help

if this was a timing issue i would expect it regardless of what ever the machine was doing
but this is not the case ,

again i ask are you using the enable to your stepper drives ? , which bob are you using and which stepper drives
and how are you connecting etc , this all has a bearing on your problem

it could of course be the pc , or your configuration , so please archive your config and attach
this information will help us all to rectify your problem quickly

Thanks a lot, but please excuse me if I disagree.
Here are the facts.
The driver is based on the L297 + L298 on the parallel port

0) The trouble comes from day to next day (electrical error possible)
1) both UIbuntu 10.4 /EMC2.3 and at 12:04 / EMC 2.6 (electrical error possible)
1) only the x-axis. (electrical error possible)
2) only at homing (search) (electrical error possible)
3) the software stopted (look at the display) (definitely no electrical fault)
4) The axis accelerates and brakes. (Definitely no electrical fault)
5) after homing there are no problems, the machine runs well. (definitely no electrical fault)

Tomorrow I will install a separate parallel port card. I hope to isolate the problem with that.
If the problem furthermore the I will post the hal an ini files.

Thanks a Lot for all the ideas, and the time that you offer.