Old style start for program?

Iterating alignments in manual mode is possible, but is extremely silly, when using a DCC machine.

I always try to use the analogy of setting a part up on a CNC machine to explain the three alignment steps (level, rotate, origin). When you level, that is like sitting the part on the machine table - it can no longer move up or down and can only slide left or right, forward or back or spin around whilst staying on the table. The next step would be to square the part up by running an indicator along one of it's sides or pushing it up against a straight-edge or some dowels - that is the rotate. You are stopping the part from spinning around on the table. Finally, you might slide the part up against a stop or run an indicator around a bore to determine it's centre before bolting everything down tight - that's the final, origin. Now the machine can move to any point on the part relative to your work-offsets - or alignment in CMM terms.

​Based on Neil posted on other thread. It should be very classic start with plane ( 3 points) line ( cast from circle and half circle) and a point. Anyone could explain to me why using iterate alignment?
If the Tol is out, keep on manual hitting all those points ( I have another program for third party with similar start with tight tol)

I agree with you.

​ ​
That's just CMM cruelty.
WHAT A MONSTER ​​

​Based on Neil posted on other thread. It should be very classic start with plane ( 3 points) line ( cast from circle and half circle) and a point. Anyone could explain to me why using iterate alignment?
If the Tol is out, keep on manual hitting all those points ( I have another program for third party with similar start with tight tol)

I agree. I don't think an iterative alignment is necessary for that part. It works, but it is more complicated than necessary.
I'd probably just use a read point alignment that starts from the CIR1 hole, but I'd have to see the part geometry to say for sure.​

Many times I have needed to use an iterative alignment for the manual alignment, BUT, I do see some issues with it being called an 'old style' start. The only time I ever use an iterative for the manual is when you can NOT (with the features used) do a 3-2-1 alignment. Doesn't happen often, but it does happen.

OLD STYLE wouldn't allow any slot to be used in an iterative alignment, it was a FUBAR waiting to happen, had to (I have to, I use V3.7) construct a point a the slot, then use that point, not the slot.

ALSO, when I had to use an iterative alignment for the manual alignment, I used "easy" features, not necessarily datum features. I also cranked up the TARGET tolerance to 25mm or more so that it would NOT ask to re-measure any of the features (if you can't hit within an inch, get a different job!), and I use MEASURE ALL FEAT = NEVER, they will not be run unless marked. You will not be prompted to position the probe to measure 'feature x' unless your hit(s) aren't within the target value.

Since you are using 3 points & 2 'holes', your FIXTURE tolerance can be MUCH lower since you are using the minimum possible features (3 level, 2 rotate, 1 origin), I use 0.05mm (0.002") fixture tolerance and it isn't an issue, since it will solve with ZERO deviation in all the locating axis for each feature.

ALL that being said, you should NEVER use a manual alignment as the 'final' alignment, it should be followed by a DCC controlled alignment, thus the reason for 'easy' features and huge target tolerance for the manual.

If they do NOT have a DCC alignment after that manual alignment, then they are absolutely clueless about what they are doing and I would suspect many more issues in the program.

If using iterative alignment...

Manual iterative should be MEASURE ALL ONCE with a large target pnt radius & like 2-3 iterations

Should immediately be followed by a DCC alignment (MEASURE ALL ALWAYS) with a tighter target point radius

Above said..this program is very poor quality & agreeing with above I'd suspect there are a lot of issues in it.

if it is set to measure all once, then even if you no longer NEED to measure the manual for more parts because of your setup, it will force you to measure them again anyway, won't it? Even if you unmark them, when you run, it will mark them for you, won't it? Measure all = never and it will only measure them IF you mark them, so if you have a 'good enough' setup, you can unmark-all, the mark-all no manual and you won't have to probe the part again.