Calibration Results

[pguillory]

Hey guys... seems like I am missing something with Probe Calibration. Maybe a brain fart, lol.

Anyways, could someone please explain in the most simplest way, what is the end goal of probe calibration results? Here's one of mine:


Probe file=1X21MM Date=6/5/2020 Time=9:53:39 AM

Check Collision Disabled

T1A60B-90 THEO X 5.9728 Y 0.0000 Z 6.0722 D 0.0394
T1A60B-90 MEAS X 5.9235 Y -0.0602 Z 6.0388 D 0.0394 PrbRdv-0.0000 StdDev 0.0000

What is wrong with this? I thought it was about the StdDev = 0.0000 but my XYZ don't match. This has been driving me crazy and I believe it is preventing me from getting correct measurements with my models. For example when I create a scan and create the path, my probe wants to 'go inside' the part. Sucks.

If anyone could assist me with this that would be great!!

[Matthew D. Hoedeman]

that is because your build/head/machine isn't perfect. If they were, you wouldn't need to calibrate.

[pguillory]

Not sure if this is a "nothing is perfect" response or if you're actually telling me that something is wrong with my machine, lol. Sorry, just trying to get this situated and make somewhat accurate measurements/scans. Seems like I'm looking for one thing while getting another.

[Matthew D. Hoedeman]

Calibration results show you how "imperfect" your stuff is. There is ONLY ONE probe that will show zero deviations for XYZIJK and that is your master probe.

[A_Shomshor]

Squaring your probe head will help some of that deviation.

[pguillory]

Cool, how would I go about that if you don't mind?

[A_Shomshor]

Depends on your head. Some are made to be adjustable. Others, like mine, are trial by fire. Use the cal. sphere, for example, measure at a A90B90 and A90B-90, compare cal. results and adjust.

[Mike Ruff]

As Matthew D. Hoedeman said, it is just showing the "error" of your probe build. As long as you calibrated correctly, those numbers will not affect your results, and don't really mean anything to you. PC-DMIS uses these numbers to relate different tips/ angles to each other

[A_Shomshor]

Having the head square gives me the confidence I'm not shanking out. Helps with process of elimination when your using a small ruby.

[Douglas]

I agree and try to keep everything as close to ideal as possible, a well squared head and fixture will significantly reduce shanking errors with smaller tip diameters

[JEFMAN]

You don't use enough decimal places to check the std dev...
Here, it just says that you have less than 1.2µm, which is a bad calibration !
However, if you use a master probe, then calibrate this tip, it doesn't explain why the scan is "inside the part" !
I would check other parameters (vectors, alignment.....)

[pguillory]

Ahh, got ya! So how should I be set up to check std dev correctly? And if you don't mind, do you have any best practices for your calibration process? JEFMAN

As far as the scan going inside the part, I did realize that the vectors on the surfaces of the CAD are actually going "into" the part. So if I use "CAD Comp" during the scan, it follows the vectors of the CAD which = Probe wants in (I'm assuming).

So creating and measuring the scan without CAD Comp give the 2D cut plane like normal.