[ezzywave]

We are using an Emuge thread mill that cuts 1 thread per pass.

Fanuc 31i.


(Calling program has this line)

N260 M98<TMMACRO2>L25(repeat 25 times)




%
<TMMACRO2>(MACRO FOR THREAD MILL)
N10 G91 G02 X0 Y0 I1 J0 W.125;
M9
%

This is a horizontal machine, The helical int. starts at the 9 0'clock position. Every time it completes a pass and gets back to 9 o'clock it pauses for a half second before starting the next one.

We are just dry running this now, but I'm pretty sure these threads aren't going to look very good.

Any ideas?

[samu]

it is not a macro call but a sub program call, may be your delay is due to exit en re-enter in the sub program.
try to loop inside the macro instead of repeat the sub 25 time. Pass the number of loop in argument inside the macro

example:
G65 A25. P9212 (MACRO CALL WITH 25 PASSED IN #1)

%
O9212(THREAD MILL MACRO)
#2=0 (SET COUNTER TO 0)
WHILE[#2 LT #1] DO1
N10 G91 G02 X0 Y0 I1 J0 W.125;
#2=#2+1 (INCREMENT COUNTER)
END1
M9 (STOP COOLANT IN YOUR SUB, YOU STOP COOLANT AFTER ONE TURN, NOT SURE IT'S WATH YOU WANT)

[Hardym1]

ezzywave

The reason that it pauses is because of where your EOB is.

Try this:
N10 G91 G02 X0 Y0 I1 J0 W.125 M99;

[sinha_nsit]

I believe Samu's method would work.
Another (crude) way could be to simply copy/paste (repeat) the G02 line 25 times. It will have no overhead of any type.

[ezzywave]

ezzywave

The reason that it pauses is because of where your EOB is.

Try this:
N10 G91 G02 X0 Y0 I1 J0 W.125 M99;

Tried this. Still has the pause every circle. Thank you for the suggestion though.

[ezzywave]

it is not a macro call but a sub program call, may be your delay is due to exit en re-enter in the sub program.
try to loop inside the macro instead of repeat the sub 25 time. Pass the number of loop in argument inside the macro

example:
G65 A25. P9212 (MACRO CALL WITH 25 PASSED IN #1)

%
O9212(THREAD MILL MACRO)
#2=0 (SET COUNTER TO 0)
WHILE[#2 LT #1] DO1
N10 G91 G02 X0 Y0 I1 J0 W.125;
#2=#2+1 (INCREMENT COUNTER)
END1
M9 (STOP COOLANT IN YOUR SUB, YOU STOP COOLANT AFTER ONE TURN, NOT SURE IT'S WATH YOU WANT)

This worked!!! I had tried a similar sub (it had the same pause problem) with an "Increment Counter" like the yours, but I didn't use the DO WHILE function. BTW, the M9 should of been a M99, I did a sloppy cut and paste.

I found another forum post that said said the pause is caused by acceleration/deceleration between circles that was caused by using logic.

Now for my next question- How did you know this would solve my problem??? I was ready to write a sub with 25 lines of G2 moves!!! Not very convenient for different sized holes... I need to learn more about Fanuc "logic".

[samu]

level change i.e. pass from main to sub, sub to main or sub to another sub take a signifiant amount of time, also the use of "goto" statement is to avoid if you don't want to have a pause. The goto function has to scan the program from the current block until the destination is found. Could take a lot of time if you jump a large amount of block. The use of the while loop solve that issue.

[Hardym1]

ezzywave,

Glad samu idea worked out but I just remembered that it should look like this:

M98P0001(whatever your sub program #)L25

Sub program is:
O0001G91 G02 X0 Y0 I1 J0 W.125 M99;

NO EOB after your program number.

Just ran it on my machine with no problem

[sinha_nsit]

Hardym1,
Had you observed the pause problem when M99 was placed in a separate block (which got corrected when M99 was shifted to the previous block) ?

[Hardym1]

sinha_nsit,
No there was no pause when M99 placed in a seperate block, only when EOB was after the program #. I had the same issue years ago and remembered that it had something to do with where the EOB was. It's just that I have not done thread milling that way in many years.

[sinha_nsit]

I do not understand why placing or not placing EOB after O-word should make any difference, as regards execution.