[Alexxmf]

Hi all,

I have written a macro to put an oil groove on a part using the G32 cycle. It works OK, but I am having a slight issue. The macro is as follows:

(OIL GROOVE MACRO)
(SET USER INPUTS BELOW)
(REQUIRES PROVE)

(USER INPUTS)
#100=1.(ABSOLUTE Z START POSITION)
#101=-26.(ABSOLUTE Z END POSITION)
#102=25.(SPINDLE SPEED)
#103=48.0(START DIA)
#104=49.25(END DIA)
#105=0.025(DEPTH OF CUT - ON RADIUS)

(CALCULATIONS)
#105=ABS[#105]*2(DEPTH OF CUT ALWAYS POSITIVE AND ON DIA)
#106=ABS[#100]+ABS[#101](TOTAL CUT LENGTH PER PASS)
#107=#106*4(CALCULATE FEED/REV)
#108=ABS[#103-#104](CALULATE AMOUNT TO CUT)


T0707M08(CALL TOOL AND OFFSET)
G00G97G99S#102M03(R/MIN - FEED/REV - SET SPINDLE SPD C/W)
X#103Z[#100+20.](RAPID TO INITIAL POSITION)
Z#100
WHILE[#108GT#105]DO1(SET LOOP)
G0U#105(FIRST DEPTH OF CUT)
G32Z#101F#107(Z PASS 1 - IN)
Z#100(Z PASS 2 - OUT)
Z#101(Z PASS 3 - IN)
Z#100(Z PASS 4 - OUT)
#108=#108-ABS[#105](UPDATE COUNTER)
END1

(REMAINDER)
N1111G0U#108(LAST PASS)
G32Z#101F#107(Z PASS 1 - IN)
Z#100(Z PASS 2 - OUT)
Z#101(Z PASS 3 - IN)
Z#100(Z PASS 4 - OUT)
G0Z20.
X250.Z200.(SAFE POS)
T0700(CANCEL TOOL OFFSETS)
M05
M01

So basically for every revolution of the spindle, the tool moves in and out twice. The problem i'm having is on the final pass (the REMAINDER section of my macro). I don't know if it is a control issue (Fanuc series O-T), but when #108 becomes equal to #105, it should skip to the REMAINDER section of my macro (as #108 is not greater than #105). However, it does one more loop of the WHILE loop, then goes to the remainder. (It may be worth adding that after this 'extra' pass on the WHILE loop, when the counter is updated, the value returned on the control is '*********', not zero??).

I hope this is explained well enough as this is incredibly difficult to put into words!

Thanks in advance,

Alex.

[sinha_nsit]

Such things may happen due to floating point errors in calculations.
Replace #105 by [#105+0.001] in the WHILE loop, and check if it helps.

[norbert.barnard]

Hi,
the system counts in binary, so you always have a chance to have a 0.000000001 discrepancy either way.

1. it is better to use GE, not GT, there is always a chance the can be equal.
2. resolve your rounding issues with #108, INCLUDE #108=[RND[#108*1000]/1000] after the calculation.
3. #108 duty is only useful in the while loop
4. in the (remainder), to be sure it is better to use N1111 G0X#104 as the final pass.

have fun