On 2010-09-02, Karl Townsend wrote:

Karl, what about this routine (with copious coolant): drill a small
increment, then dwell the drill for a small fraction of a second to
break chips, then continue. That would break chips, right? I could
literally stop the Z axis movement to dwell for several times per
second. Would that be enough? Withdrawing to SafeZ every time is very
time consuming.

i

Standard drill cycle G82. I don't use it much, but it is faster. When
you get more than say 4D deep, the chips often quit feeding. The
result is a broken drill JAMMED in the hole. Happens most often on
nearly finished parts. DAMHIKT

So G83 is still safer? Right?

I don't try for top speed on one of parts. Start machine and do
something else. I like to see how many machines I can run at once -
four is the norm.

Or I use run time to surf RCM or even clean up the shop.

I now browse the web, etc from the CNC mill. I configured the kernel
to dedicate one CPU to realtime tasks only, so no matter what I do on
the other CPU, does not affect the realtime task.


i


' drill cycle with dwell
' format is G82 X_ Y_ Z_ R_ F_ D_ Q_
' X is coordinate
' Y is coordinate
' Z is bottom of hole Z value
' F is feed
' R is start point at top of hole, can be ommitted, then start point
is current Z readout
' D is dwell time
' Q is dwell peck distance, if 0 only dwell at bottom of hole