+13
Completed

The option to extract point data from auto features.

Dave Drew 8 years ago in Metrology Software / PC-DMIS updated by Kyle Brummans 4 years ago 15

The option to extract individual point data from auto features, so that I can report xyz values for a single point.

GOOD, I'M SATISFIED
Satisfaction mark by Dave Drew 4 years ago
Under review

I think there's certainly room for improvement in this flow, although I believe it is possible currently by

: 1. create generic point

2. type in nominals

3. type in actuals (ex. cir1.hit1.x, cir1.hit1.y…)

4. create dimension.


A simpler way perhaps could be to offer an option when you hover over hits in the graphics window when editing a feature, for creating dimension on that hit/point

There is a script "To Points" on the PC-DMIS form that will do this already to vector points. So, use that to help develop a feature similar to the "To Points" option used with the Scan features. Then have a choice added for the type of points to be created. Example (Constructed Cast point, Vector point, Generic point). Also, when creating a dimension add an option for the selected feature in the dimension list box to view a list of the individual hit ID's for that feature. Then be able to dimension the individual hits for that selected feature.

All iterations of the "To points" script in the forum are based on my source code. If that code were to move into a commercial product, I have an armada of hungry lawyers just waiting to make sure my interests are secured... ;P

This would HUGE within scanning. I will add the seperate topic as requested.

Construct/Point/Cast from PLN1.HIT[n] (for example) correctly calculates all the nominals and actuals (last time I used it)

ASSIGN/HITN = MAXINDEX(SCN6.HIT[1..SCN6.NUMHITS].Z)
ASSIGN/MAXVAL = SCN6.HIT[HITN].Z

F1 =GENERIC/POINT,DEPENDENT,RECT,$

NOM/XYZ,SCN6.HIT[HITN].TX,SCN6.HIT[HITN].TY,SCN6.HIT[HITN].TZ,$
MEAS/XYZ,SCN6.HIT[HITN].X,SCN6.HIT[HITN].Y,SCN6.HIT[HITN].Z,$

NOM/IJK,0,0,1,$

MEAS/IJK,0,0,1

This is some code I used for pulling high points in my scans. I put up another suggestion to be able to pull points from all features because a scan is kind of treated like an auto feature. This code is very very helpful. If there is a quicker and easier way to select points out of any set of points that would be a big step for data/point manipulation and output.

To ensure we help address the underlying problem rather than treat the symptom, we would be interested in learning more about what you typically need to access point data for. Converting Scans and Features to points can make somewhat unwieldy Measurement Routines. If instead there is a limited set of things that this is required to achieve, it may be preferable to implement nice simple solutions to address those specific things instead.


Ones we're aware of are :-


1 - To be able to easily dimension/report the underlying Point/Hit xyz

2 - To have more control over constructing features from a specified subset of points/hits (i.e. the other highly voted request that we're also looking to implement)

3 - To be able to create alignments based on underlying points/Hits


If all of the above scenarios were supported, are there other benefits in converting Scans and Features to points? If so, please share...


Thanks!



Hi Neil,

My apologies for the delayed response.


In my most recent situation this code was crucial for me to be able to measure a coplanarity spec for a semiconductor chip with hundreds of little solder bump radii. The sensor would scan the complete tolerance zone of the cylinder every 1µm by 1µm. I would then use this code to extract the highest point of each solder bump and construct a best fit plane. See JEDEC Standard JESD22b-108a.

This is useful for extracting high and low points of data. As WLS scanning becomes more sophisticated hopefully we will be able to measure surface finish in the future. I have tried many times to achieve this, it is close but just not capable yet(at least to my knowledge). This code would be essential for things like Rt and Ra specifications and ANSI surface finish standards. As of now I can use this code to extract high and low points for a surface at a normal measurement resolution to check flatness, profile ect.

Breaking scans into individual segments to perform the same procedure would be beneficial as well.

There are many more applications I have thought of but I will have to add them as they come to mind.

From: UserEcho
Author: Dave Drew

The option to extract individual point data from auto features, so that I can report xyz values for a single point.

----------------

I think there's certainly room for improvement in this flow, although I believe it is possible currently by
: 1. create generic point
2. type in nominals
3. type in actuals (ex. cir1.hit1.x, cir1.hit1.y…)
4. create dimension.

A simpler way perhaps could be to offer an option when you hover over hits in the graphics window when editing a feature, for creating dimension on that hit/point

---------------

There is a script "To Points" on the PC-DMIS form that will do this already to vector points. So, use that to help develop a feature similar to the "To Points" option used with the Scan features. Then have a choice added for the type of points to be created. Example (Constructed Cast point, Vector point, Generic point). Also, when creating a dimension add an option for the selected feature in the dimension list box to view a list of the individual hit ID's for that feature. Then be able to dimension the individual hits for that selected feature.

-------------
All iterations of the "To points" script in the forum are based on my source code. If that code were to move into a commercial product, I have an armada of hungry lawyers just waiting to make sure my interests are secured... ;P

This would HUGE within scanning. I will add the seperate topic as requested.

-------
Construct/Point/Cast from PLN1.HIT[n] (for example) correctly calculates all the nominals and actuals (last time I used it)

ASSIGN/HITN = MAXINDEX(SCN6.HIT[1..SCN6.NUMHITS].Z)
ASSIGN/MAXVAL = SCN6.HIT[HITN].Z
F1 =GENERIC/POINT,DEPENDENT,RECT,$
NOM/XYZ,SCN6.HIT[HITN].TX,SCN6.HIT[HITN].TY,SCN6.HIT[HITN].TZ,$
MEAS/XYZ,SCN6.HIT[HITN].X,SCN6.HIT[HITN].Y,SCN6.HIT[HITN].Z,$
NOM/IJK,0,0,1,$
MEAS/IJK,0,0,1

This is some code I used for pulling high points in my scans. I put up another suggestion to be able to pull points from all features because a scan is kind of treated like an auto feature. This code is very very helpful. If there is a quicker and easier way to select points out of any set of points that would be a big step for data/point manipulation and output.

---------------

Targettng 2018 R1 with this one.

I added another thread from the idea center to the description. This will also be handled by the proposed solution here.

Neil is there an article on this anywhere?  I know it's done but it always eludes me when I go to use it.