Haas g-code -- inneresting.....
 

On Friday, May 3, 2013 11:18:18 PM UTC-5, Existential Angst wrote:
Awl --



Mostly I've been bitching about the clumsiness/complexity of the Haas

control, esp. it's miserable file-handling protocols. Hooray, Fadal....



But alladat complexity is finally yielding some fruit, as today's

experimentation revealed.



I got the blast-air option with the GR510, which looks almost like coolant,

except it's an extendable/retractable piston that shoots air very close to a

spinning tool, controlled by M83. For chip clean out, pre-empting bird's

nests, etc.



So you can make a move, cut, and blast some air on the tool.



But what about doing this WITHIN a canned drill cycle?? Really, quite a

different scenario, g-code-wise.



Now, G-coders know about canned drill cycles, and calling subroutines, which

when used with drill cycles, are simply lists of coordinates.

Suppose these hole coordinates (for any number of holes) are in a subroutine

L1000. So for a typical spot/drill/tap procedure, you'd issue a G81 for the

spot, and call L1000. Then G83 for the drill, call L1000. Finally, G84 for

the tap, call L1000. Ackshooly "L1000" is Fadal-speak, but no matter....

lol The point being, subroutines *greatly reduce* the possibilities of

error, typos, etc. Nothing worse than tapping a hole that is, well,....

not there!!!!



L1000 might look like this:



X1. Y1.

X1. Y2.

X1. Y3.



for a column of holes.



AMAZINGLY, you can do sumpn like this in the subroutine:



X!. Y1. M83 P1000, where M83 is an m-code for "blast air".

X1. Y2. M83 P1000

X1 Y3. M83 P1000



And then call G83

L1000



and the blast air will execute along with the holes!!!!



So when the tool finishes a hole, the Z axis pauses for 1 second (1000 ms),

and the blast air blows **** off the tool.



**Right within the subroutine/canned drill cycle**!!!! Effing amazing!!!!!



Cupla side problems, tho....

The drill cycle *also* interprets the P1000 as a dwell!!!! LOL!!!!

So the tool does two things. It dwells at the bottom of the hole for 1,000

ms, and then when it comes up, pauses for 1,000 ms whle the air blasts.....



Nothing is for free, I suppose. Speaking of which, dat effing blast air

uses a lot of air!!!! holy ****....



But at any rate, I thought cnc-ers, g-coders out there would get a kick out

of m-codes working WITHIN a canned cycle.



Hooray Haas!!





I also un-mounted the Renishaw probe.

First, I don't know how the fukn thing works, it's so goddammed complicated.

Second, it's a crash -- a VERY EXPENSIVE crash -- just waiting to happen.

So I just took dat **** off, whilst I'm drilling, fixturing the table.

What an effing RELIEF!!!!

--

EA

Once you learn how to use dat effin Renishaw probe....
You will DEMAND one on every machine you run/own after that.

Haas g-code -- inneresting.....
 

"Cross-Slide" wrote in message
...
On Friday, May 3, 2013 11:18:18 PM UTC-5, Existential Angst wrote:
Awl --



Mostly I've been bitching about the clumsiness/complexity of the Haas

control, esp. it's miserable file-handling protocols. Hooray, Fadal....



But alladat complexity is finally yielding some fruit, as today's

experimentation revealed.



I got the blast-air option with the GR510, which looks almost like
coolant,

except it's an extendable/retractable piston that shoots air very close
to a

spinning tool, controlled by M83. For chip clean out, pre-empting bird's

nests, etc.



So you can make a move, cut, and blast some air on the tool.



But what about doing this WITHIN a canned drill cycle?? Really, quite a

different scenario, g-code-wise.



Now, G-coders know about canned drill cycles, and calling subroutines,
which

when used with drill cycles, are simply lists of coordinates.

Suppose these hole coordinates (for any number of holes) are in a
subroutine

L1000. So for a typical spot/drill/tap procedure, you'd issue a G81 for
the

spot, and call L1000. Then G83 for the drill, call L1000. Finally, G84
for

the tap, call L1000. Ackshooly "L1000" is Fadal-speak, but no matter....

lol The point being, subroutines *greatly reduce* the possibilities of

error, typos, etc. Nothing worse than tapping a hole that is, well,....

not there!!!!



L1000 might look like this:



X1. Y1.

X1. Y2.

X1. Y3.



for a column of holes.



AMAZINGLY, you can do sumpn like this in the subroutine:



X!. Y1. M83 P1000, where M83 is an m-code for "blast air".

X1. Y2. M83 P1000

X1 Y3. M83 P1000



And then call G83

L1000



and the blast air will execute along with the holes!!!!



So when the tool finishes a hole, the Z axis pauses for 1 second (1000
ms),

and the blast air blows **** off the tool.



**Right within the subroutine/canned drill cycle**!!!! Effing
amazing!!!!!



Cupla side problems, tho....

The drill cycle *also* interprets the P1000 as a dwell!!!! LOL!!!!

So the tool does two things. It dwells at the bottom of the hole for
1,000

ms, and then when it comes up, pauses for 1,000 ms whle the air
blasts.....



Nothing is for free, I suppose. Speaking of which, dat effing blast air

uses a lot of air!!!! holy ****....



But at any rate, I thought cnc-ers, g-coders out there would get a kick
out

of m-codes working WITHIN a canned cycle.



Hooray Haas!!





I also un-mounted the Renishaw probe.

First, I don't know how the fukn thing works, it's so goddammed
complicated.

Second, it's a crash -- a VERY EXPENSIVE crash -- just waiting to happen.

So I just took dat **** off, whilst I'm drilling, fixturing the table.

What an effing RELIEF!!!!

--

EA

Once you learn how to use dat effin Renishaw probe....
You will DEMAND one on every machine you run/own after that.

Well, I got the option based on testimony from here, PM, etc.
So I'm not griping about it, just saying it's a little premature for me
right now, and it IS in the way.
I'm just learning the machine, doing basic setup/drilling of the table, etc.
I'll basically operate the gr510 with basic g-code, and figger out the
renishaw stuff later.

I watched one of the techs setting it up for me -- I was in a daze -- and I
still haven't recovered -- I couldn't believe it wasn't going to crash...
lol
--
EA