I'm quoting a message from an EMC user to see if anyone else has seen 
any signs of this problem.

richard harris wrote:
> Jon,
>
> i got a chance to run halscope today and on coarse threads <13 tpi it 
> will occasionally not reset the encoder count, count is numerically 
> low when it does reset 25k.  I do not see this when i run fine threads 
> >18 tpi.  When the count does not reset the threading cycle will rapid 
> the tool.
> I verified that I am running emc2 2.2.6, UPC SN 33 rev 2.3, US digital 
> 500 line encoder in a shielded cable with one side of the cable grounded.
> What should I look for in halscope to further diagnose the problem?
>
I just ran several pieces at 12 TPI, using a program derived from the 
EMC sample program threading.ngc

One interesting quirk I did see was that the finish end of the thread 
was compressed to a higher TPI.  I re-ran it at lower spindle speed and 
it was fine.  I'm not sure how that particular program figures out how 
soon to pull out at the end.  It shows up on the coarser threads due to 
the Z motion beginning to decelerate earlier, I'm guessing?

Anyway, I don't know how many passes the sample program performs, it 
looks like about 20.  So, I have run it for at least 40 passes here 
without ANY blips at all.  How "occasionally" is it?  Once per thousand 
or once in ten passes?  I am NOT running 2.2.6 here, the directory says 
2.2.7, but I'm sure this was compiled from CVS source on 11/29/2008.  I 
am also using a much finer resolution encoder, it has 1728 lines, or
6912 counts/rev.  That also "shouldn't" make any difference.  What is 
the date code either scribed or on a printed label on the EPROM chip 
(U4) on the UPC?  The final version for those older boards with the 5V 
FPGA chip was 5/2/07.  But, that is the same version I am using on my 
minimill to test this problem.

Finally, after thinking about this some, I find NO WAY the UPC board's 
encoder counter can POSSIBLY know what TPI thread you want to cut!  That 
simply doesn't make sense.  All it could possibly be affected by is 
spindle RPM,  and the possibility that your encoder index pulse is 
somehow different at different RPM.  I have certainly seen the rapid 
effect when the encoder fails to start from zero when the driver 
believes that the encoder DID see the index pulse.  That is "not 
supposed to happen" with the latest UPC firmware.  One last thing, 
coarser threads require more spindle HP, and probably more Z-axis power, 
too.  Any possibility that more load on these motors could increase 
system noise?

Have you run the diagnostic program in the "commtest" mode?  If the 
diags show communication errors between the PC and the UPC, all bets are 
off!  You might need to run the tests with the spindle drive on.

Jon
_________  end of quote and my reply

Anyway, I am running a different version of EMC here, compiled from CVS 
on 11/29/2008.  Richard is apparently running an older version, 2.2.6 
says it was released on 10-Aug-08, but I don't know when the code was 
actually locked up for that.  But, it should still be recent enough to 
have all the changes that were important, as of about July 2007.  So, 
does anyone else have experience with lathe threading or rigid tapping 
at 12 or less TPI?  I don't see why that should matter to the encoder 
counter reset on index function, but that is what Mr. Harris reports.


Jon

------------------------------------------------------------------------------
This SF.net email is sponsored by:
High Quality Requirements in a Collaborative Environment.
Download a free trial of Rational Requirements Composer Now!
http://p.sf.net/sfu/www-ibm-com
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to