Dave, Kirk, Jon, thanks!

2011/11/5 Dave <[email protected]>:
>
> What I did to verify that the VFDs were the problem was to run the
> machine without the VFD powered up - pull the fuses or unwire them..
> Then try running parts for an hour or so. Remove the cutters if need
> be.. . If the problem vanishes you likely have an VFD related noise problem.

It is easy to switch the VFD off, but I am having difficulties to come
up with a plan, how to check, if the tool takes correct path. I will
consult with client's specialists - maybe they can suggest something.

2011/11/5 Kirk Wallace <[email protected]>:>> My experience
seems to indicate:>> _Always_ have a filter on VFD power inputs.

I have a question here:
Filters on VFD input lines have been mentioned several time here. Is
this something close to what is meant here:
http://www.newark.com/tyco-electronics-corcom/10vr6/rfi-power-line-filter-10a-700ua/dp/52K3303

2011/11/5 Jon Elson <[email protected]>:
>
> You can run memtest86 from the install CD and let it run for at least
> several
> hours. This program gets the CPU warmed up and exercises memory
> heavily. If it finds even a single error, don't bother with anything else
> until it can run without error overnight.
>
> Also, you can run the latency test for a few hours, and exercise web
> browsers, USB memory stick inserts/removals, glxgears and other
> operations to see if the latency gets bad.

I ran latency test with glxgears and tried to open/close firefox last
time I was there. But I think that I ran the test for 15-20 minutes.
It was pretty fine - max jitter on base-thread was reaching 25000 ns,
but base thread is not used and max jitter on servo thread was around
16000 ns, servo period is standard 1 ms.
I recall that I ran latency tests, when I initially installed EMC2 on
the machine and there were cases, when max jitter reached >800K ns -
almost another servo period. Also that time I ran the test for less
than 15 minutes.
I will try to figure out, how can I these tests for longer time period
without sitting next to the machine - it is several hours to drive
from my place. And I think that client will not be happy about machine
idling for a whole working day, but I will handle that.

Viesturs

------------------------------------------------------------------------------
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to