Readpoints

[CFisher]

I admit it. I'm a newbie.

Would someone give a short primer on how readpoints are used in lieu of a manual alignment - allowing one to jump right into DCC?

All our programs begin with Manual alignments then DCC - I'm wondering if there's a better way...

Thanks...

[bw_bob]

Place your probe like in a diameter or something. Do a readpoint which reads the position of your cmm. Then make that zero,zero,zero. Put in dcc & start programming from there.

[cmmguy]

If you use readpoints it assumes that the part is reasonalby well aligned(skew) as when you programmed it. Also, you should still do the basic alignment (under DCC) as you would have the manual alignment, dont count on a single pass DCC alignment to be sufficient.

[davey c]

Good points from bw_bob and cmmguy. I use readpoints for all programs I write, so I will attempt to go a little more in depth. As bw_bob said, a diameter (center of a hole approx level with surface) is a good place to start. I usually use a Datum B hole since this will be the first feature you measure. As cmmguy said, your part has to be oriented on the table the same way you program it when you run it in the future. I use a simple Excel spreadsheet for a setup instruction. (Example attached) Position the probe at your desired readpoint, use hot-key control G to insert readpoint. Do an alignment to set the readpoint to 0,0,0. (As bw_bob pointed out) From this point it depends if you're using a CAD model or not. I do 99% of the time, so this applies to CAD. Pick three 3-D features. (doesn't matter if they're datums or not: you'll be doing another alignment later) Set the targets to actual machine coordinates and vectors and keep the CAD theoreticals. Insert an Iterative alignment to these features. Go back to right after the readpoint alignment and insert a DCC mode command. You should be able to run what you have so far in DCC mode. Then pick your print datums and do your final alignment. I always do a "measure all always" option so it runs the alignment features twice. After that just measure all the features you want to check. Part of an example of this type of program is attached also. I hope I've given enough info to get you started and if I've missed anything hopefully someone will correct me. There are people on this board far more knowledgeable than myself. I know I learn more reading postings here than any class I've ever taken. I hope this helps you out. Good luck.

Dave

[T Miller]

The only thing I may add is you may want to open up your pre/retract hits for the first dcc alignment if the part is not on a fixture.

[Roberto]

In the past we had a programmer that liked this technique.
Can I ask what's the real advantage of it? I see the initial manual alignment gets eliminated. However, more alignments are done plus "measure all always" in my opinion add up more inspection time.

Thanks.

[cmmguy]

The main reason from my standpoint is reduced operator required steps which reduce the chance of mistakes and therefore possible crashes...
There are no additional alignments beyond what you would do without the Readpoints. The manual alignment steps are replaced by a preliminary DCC alignment which would be faster than an operator's manual alignment.

[AuRules]

The purpose of readpoints for me is only to move the probe above the part wherever it at to begin DCC and to make up for operators who has not read the screen when it say's MOVE PROBE HEAD CLEAR OF PART.

[John Kingston]

In the past we had a programmer that liked this technique.
Can I ask what's the real advantage of it? I see the initial manual alignment gets eliminated. However, more alignments are done plus "measure all always" in my opinion add up more inspection time.

Thanks.

The "measure all always" comes from an iterative alignment that was used. You can use readpoints without it.

[Matthew D. Hoedeman]

The "measure all always" comes from an iterative alignment that was used. You can use readpoints without it.

NOT using the MEASURE ALL ALWAYS and also mess up your results.

See this post for the reason why. It does not talk specifically about the iterative alignment, but the effects are the same, if you are NOT aligned 'correctly' before measuring the finaly iterative alignment features, you can be grossly wrong in your final results.

http://pcdmisforum.com/showthread.php?p=55265#poststop

[davey c]

Hey Matt,
What you're talking about is exactly why I do a "measure all always". I use Rayco stand-offs to locate the part in the same place on the table. However, you always have just a little bit of "jiggle". When an operator follows the setup instructions, they don't even have to do the readpoint alignment. The software knows where the part was last time it was run. They just set it up, position the probe above the setup and choose run one part (or multiple parts) from the menu selections I have set up in the program. When it measures the alignment features the second time, the hits will be in EXACTLY the correct location and vectors. I think that's what Matt was refering to. And I don't think it adds more inspection time. Especially when inspecting multiple samples. Slap the part up there and go. I try to make programs as user friendly as possible. And I get that warm fuzzy feeling that the results are accurate and repeatable.

[John Kingston]

NOT using the MEASURE ALL ALWAYS and also mess up your results.

See this post for the reason why. It does not talk specifically about the iterative alignment, but the effects are the same, if you are NOT aligned 'correctly' before measuring the finaly iterative alignment features, you can be grossly wrong in your final results.

http://pcdmisforum.com/showthread.php?p=55265#poststop

What I meant was you can use readpoints with a standard 1-2-3 alignment. I didn't mean turn off "measure all always" when using an iterative alignment. I agree that not using it can mess up your results.

[Matthew D. Hoedeman]

What I meant was you can use readpoints with a standard 1-2-3 alignment. I didn't mean turn off "measure all always" when using an iterative alignment. I agree that not using it can mess up your results.

Not talking about you, but about this:

However, more alignments are done plus "measure all always" in my opinion add up more inspection time.

Thanks.

Yes, it adds a little more time, but time well spent (IMO)!!! Without the measure all always, if the part is not cubical (in which case, do a 3-2-1 alignment), you will not be aligned correctly and will be reporting bad data. You can be passing bad parts and rejecting good parts. If the part has 3-D vectors for the iterative alignment features, you SHOULD do measure all, always.