Profile Scan Closed problem:

[Qinsp]

I'm trying to reverse engineer a gasket by using the part it goes in. It's a flat surface with a gasket receiver groove. I'm using v4.1 .

I do the part alignment, then take two points with joystick in the "PART" coordinate system (part is at a 15° angle to the table).

I start the trace, and it slowly climbs in Z out of the groove. The start points have a different Z value, so I set them identical. (they should have been already, but were off .001") It still climbs out. So I have to manually set the vectors, and as I change them, they keep changing back a little. After I do this a couple of times, they stay put and the trace comes out fine.

Is there a way to just tell it I want an XY trace?

[JamesMannes]

I'm trying to reverse engineer a gasket by using the part it goes in. It's a flat surface with a gasket receiver groove. I'm using v4.1 .

I do the part alignment, then take two points with joystick in the "PART" coordinate system (part is at a 15° angle to the table).

I start the trace, and it slowly climbs in Z out of the groove. The start points have a different Z value, so I set them identical. (they should have been already, but were off .001") It still climbs out. So I have to manually set the vectors, and as I change them, they keep changing back a little. After I do this a couple of times, they stay put and the trace comes out fine.

Is there a way to just tell it I want an XY trace?

What you need to do is adjust the vector in the cutvector box. You should see 3 choices in a box with IJK values. You'll notice after taking the 2 points on your surface that the cutvector of your 1st point is going to be relatively close to either -1 or +1 in the "K" vector, that coincides with the Z axis. Double click where it says cutvector and a window will pop up allowing you to adjust the vector. Make the I and J "0", and make K a 1 or -1 depending on what value is already in the window(should be something like -.999123 or .999123).

As a side note, Did I read post of yours recently that stated you just got your CMM installed? If so don't expect miracles from yourself or anyone else running the software for quite a while. If you have taken the introductory class, the basic stuff is relatively easy, but anything alese the instructor may have shown you(or a sales demo) looks easy because they have been doing it for a while. Keep posting to the forum and the guys/gals that frequent it will help you best that they can.

And welcome!!!

[Qinsp]

Thanks!

I don't get a choice to wait for the training. We ordered a used Mitutoyo BHN1015 and it was severely damaged during shipping. So I rushed out and bought a B&S Global Performance 12*15*10? new so I could make our commitments. First job arrives next week, but first training class is 3 weeks away. So the "boss" has to figure it out pronto or tell my customer to pound sand.

I did get it to work by editing the vectors, but the data was not useable. I had probe comp turned on, and it was a mistake. Huge errors with a 2mm ball on even gentle transitions. They need to rethink this feature. Even the 20 yr old Mitu software does it better.

So far we can collect data with it, but the programs never execute. They error out with a cryptic Can't Get There From Here machine control error when you try to execute. Somehow I'm getting machine coordinate system mixed up with the Part Coordinate System, and it thinks I want to go up 41" in Z every time I start a program. There is nothing in the program text that indicates a problem.

Everything works fine while learning, or going in Manual mode, but as soon as you execute, the first DCC command always loses track of the part coordinate system.

The software has some exceptional "Bling Bling" features but doing very basic inspection is clumsy and slow IMO. But I've been around long enough to realize that EVERYONE thinks this when they change software.

With what we do (contract inspection), the name of game is Dimensions-Per-Hour on a brand new first article. This is the lion's share of the business. Programmed parts are perhaps 10% of the work volume. From the time a job is signed in, to the time the customer receives the data determines the value of a software package, not machine drive speed, or software feature list.

I decided on PCDMIS for several reasons:

It is to CMM's what EXCEL is to spreadsheets. It's now the industry standard.

It's ability to be installed on different platforms makes it a possibility to have the same software on all our machines.

Automatic feature recognition means you can shoot multiple characteristics without pausing to tell the computer you want to change entities.

I think it does have the potential for good Dims/hour I just need to figure out how to optimize it, and minimize time spent playing with a mouse and hidden menus.

[bob mappes]

Pat

I feel your pain.
We had our Starrett upgraded a year ago. While they did the install of a new controller and PCDmis I was in training. Came back to difficult parts and only basic skills to check em with. Was that ever fun. NOT

Wish I was closer I'ld come over and try to help you.
You are not that far from Hexagon in Irvine, try droping in on thr guys there. I bet they will help you some. Eric generally works late there anyways, at least he used to when I was there last fall for some advanced training.

and as was said before....keep coming here (forum) with questions. you can usually get answers (and some smack) pretty quickly.

take care
bob