Circles in manual alignment are off

I've noticed for the past few parts I've programmed that used a circle in the manual alignment that the center of the circle is off after the manual alignment causing the DCC measurement of the same circle to hit unexpectedly. It does not matter if I use CAD to find nominals or enter them by hand. Once the circle is measured in DCC mode it will be off some amount in some direction. I have been using Find Hole on the DCC circle to get around the problem. I don't run production so the extra few seconds doesn't really bother me, sloppy as it is.

Since I've seen it asked before, I use auto circles for my manual features. If CAD is used it is transformed to match the datum alignment from the drawing. I verify vectors and origins are what they are supposed to be before picking features to measure. Manual alignment is usually the minimum hits on datum features. First DCC alignment is the manual alignment in DCC mode. I follow that with a datum alignment with more hits appropriate to feature size. There are no problems once the first DCC alignment has been executed. Nominals for both the manual circle and DCC circle are the same.

Below is the latest program this has happened on. Also workplane doesn't seem to matter; I've had circles in Y- and and Z+ cause the same problem.

            COMMENT/OPER,NO,FULL SCREEN=YES,AUTO-CONTINUE=NO,OVC=NO,
            TAKE THREE HITS ON TOP OF PART.
PNT1       =FEAT/CONTACT/VECTOR POINT/DEFAULT,CARTESIAN
            THEO/<0.3548,0.0661,0.125>,<0,0,1>
            ACTL/<12.8826,3.7971,-20.6611>,<0,0.000466,0.9999999>
            TARG/<0.3548,0.0661,0.125>,<0,0,1>
            REPORT VEC=<0,0,1>,UPDATE VEC=<-0.0001,-1,0.0001>
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
PNT2       =FEAT/CONTACT/VECTOR POINT/DEFAULT,CARTESIAN
            THEO/<4.1455,2.1597,0.125>,<0,0,1>
            ACTL/<16.5489,6.0403,-20.6591>,<0,-0.0004873,0.9999999>
            TARG/<4.1455,2.1597,0.125>,<0,0,1>
            REPORT VEC=<0,0,1>,UPDATE VEC=<-0.0001,-1,0.0001>
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
PNT3       =FEAT/CONTACT/VECTOR POINT/DEFAULT,CARTESIAN
            THEO/<3.0553,6.6807,0.125>,<0,0,1>
            ACTL/<15.7474,10.5304,-20.6414>,<0,0.0014591,0.9999989>
            TARG/<3.0553,6.6807,0.125>,<0,0,1>
            REPORT VEC=<0,0,1>,UPDATE VEC=<-0.0001,-1,0.0001>
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
PLN1       =FEAT/PLANE,CARTESIAN,TRIANGLE,NO,LEAST_SQR
            THEO/<2.5185,2.9688,0.125>,<0,0,1>
            ACTL/<15.0596,6.7893,-20.6539>,<0.0016946,-0.0036462,0.9999919>
            CONSTR/PLANE,BF,PNT1,PNT2,PNT3,,
            OUTLIER_REMOVAL/OFF,3
            FILTER/OFF,WAVELENGTH=0
            COMMENT/OPER,NO,FULL SCREEN=YES,AUTO-CONTINUE=NO,OVC=NO,
            TAKE TWO HITS ON RIGHT EDGE.
            SURFACE IS MARKED RED ON MODEL.
PNT4       =FEAT/CONTACT/VECTOR POINT/DEFAULT,CARTESIAN
            THEO/<4.6362,2.5,0.0625>,<1,0,0>
            ACTL/<17.231,7.3416,-20.6792>,<1,0,0>
            TARG/<4.6362,2.5,0.0625>,<1,0,0>
            REPORT VEC=<1,0,0>,UPDATE VEC=<-0.0001,-1,0.0001>
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
PNT5       =FEAT/CONTACT/VECTOR POINT/DEFAULT,CARTESIAN
            THEO/<4.6362,5.5,0.0625>,<1,0,0>
            ACTL/<17.2174,9.3276,-20.6792>,<0.9999999,0,-0.0004904>
            TARG/<4.6362,5.5,0.0625>,<1,0,0>
            REPORT VEC=<1,0,0>,UPDATE VEC=<-0.0001,-1,0.0001>
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
LIN1       =FEAT/LINE,CARTESIAN,UNBOUNDED,NO
            THEO/<4.6362,2.5,0.0625>,<0,1,0>
            ACTL/<17.231,7.3416,-20.6792>,<-0.0068087,0.9999768,-0.0000063>
            CONSTR/LINE,BF,3D,PNT4,PNT5,,
            OUTLIER_REMOVAL/OFF,3
            FILTER/OFF,WAVELENGTH=0
            COMMENT/OPER,NO,FULL SCREEN=YES,AUTO-CONTINUE=NO,OVC=NO,
            TAKE FOUR HITS ON DATUM B CIRCLE.
CIR1       =FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,IN,LEAST_SQR
            THEO/<0,0,0.0472>,<0,0,1>,0.1945
            ACTL/<12.6235,3.8274,0.0472>,<0,0,1>,0.195
            TARG/<0,0,0.0472>,<0,0,1>
            START ANG=0,END ANG=360
            ANGLE VEC=<1,0,0>
            DIRECTION=CCW
            USE_PIN=NO
            REPORT VEC=<0,0,1>
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
MANUAL_ALIGN=ALIGNMENT/START,RECALL:STARTUP,LIST=YES
              ALIGNMENT/LEVEL,ZPLUS,PLN1
              ALIGNMENT/TRANS,ZAXIS,PLN1
              ALIGNMENT/TRANS_OFFSET,ZAXIS,-0.125
              ALIGNMENT/ROTATE,YPLUS,TO,LIN1,ABOUT,ZPLUS
              ALIGNMENT/TRANS,XAXIS,CIR1
              ALIGNMENT/TRANS,YAXIS,CIR1
            ALIGNMENT/END
            MODE/DCC
            COMMENT/OPER,NO,FULL SCREEN=YES,AUTO-CONTINUE=NO,OVC=NO,
            BEGIN DCC MODE.
            CLEARP/ZPLUS,1,ZPLUS,1,OFF
            MOVE/CLEARPLANE
CIR2       =FEAT/CONTACT/CIRCLE/DEFAULT,CARTESIAN,IN,LEAST_SQR
            THEO/<0,0,0.0472>,<0,0,1>,0.1945
            ACTL/<0,0,0.0472>,<0,0,1>,0.1945
            TARG/<0,0,0.0472>,<0,0,1>
            START ANG=0,END ANG=360
            ANGLE VEC=<1,0,0>
            DIRECTION=CCW
            USE_PIN=NO
            REPORT VEC=<0,0,1>
            SHOW FEATURE PARAMETERS=NO
            SHOW CONTACT PARAMETERS=NO
            MOVE/CLEARPLANE

BIGWIG7 over 1 year ago

have you ever tried using a readpoint for your manual alignment? It could save you so much headache and time.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
mrollings53 over 1 year ago in reply to BIGWIG7

Lots. I'm only measuring one of these. A manual alignment makes the setup easier because I don't have to square the part up to the table. The part is .125" sheet metal and only uses one probe angle (A0B0).
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
mrollings53 over 1 year ago

Bit of observation I just performed. I leveled and origin Z+ to the plane in my manual alignment before I manually measured the circle. There was no, or very little, error in the position of the circle. I was noticing that the XYZ position of the circle would have the correct measured X and Y coordinates relative to the machine axis (startup align), but the measured Z coordinate would still be the nominal dimension. This makes me wonder if the Z axis tram is off.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
NinjaBadger over 1 year ago

When you import cad it's initially at the machine home position. Because you define your auto circle before you do any alignment step, it's defined at it's cad value in z (0.0427 in your example). Auto circles are always projected to their defined location. So whilst you're manual plane is at -20 in z, your circle is projected to 0.0427 (way above your part). If your part is relatively level to your machine, it doesn't have to much impact other than looking weird. But if your part is at a bit of an angle you will see the issues you are getting.

Several ways round it.
1. Use program mode circles instead of auto features in your manual alignment 2. Take sample hits (one will do) on circles in your manual alignment.
3. Select your plane as the relative feature when creating your circle
4. Level and origin in your plane before creating your circle.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
MattR over 1 year ago

1st level & z-origin on the plane before going to next features. Otherwise the original alignment is still the machine alignment.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
BIGWIG7 over 1 year ago in reply to BIGWIG7

hell that's when i use the rp the most!
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
Donald M over 1 year ago

1st level & z-origin on the plane before going to next features. Otherwise the original alignment is still the machine alignment.

I do the same and never had a problem, so far.

I noticed this in your alignment
ALIGNMENT/TRANS_OFFSET,ZAXIS,-0.125

why are you offsetting the feature you measured for your alignment?
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
mrollings53 over 1 year ago

The datum is the bottom of the part, and I'm probing the top. I like my nominal values to always match the datum structure so I shift the alignment down the thickness of the part. I'm not concerned with reporting accuracy because this is just a manual alignment.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel
MattR over 1 year ago

Offset the plane after doing final dcc alignment. Otherwise if you offset things around before your final alignment it increases chances of something going off wrong especially in the manual mode.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Cancel