All Activity

Hello, My add-on needs to run a blocking GOM command while a Modbus polling loop keeps running for a two-station workflow. Since GOM commands can only run on the main thread, I tried using multiprocessing.Process for the blocking scan. This works in the edit mode, but after the add-on is locked and packaged, the child process fails with ModuleNotFoundError when importing the add-on module. Using a temporary .py file with subprocess is not an option because it would expose the source code. I also tried threading, but it sometimes causes unexpected crashes in software, so I would prefer not to use it. Is there an official way to use multiprocessing or a separate Python process in a locked add-on without exposing the source code? If not, what is the recommended way to handle this kind of two-station parallel workflow?

I'm not for sure but aren't those surface roughness callouts to be measured with a profilometer? Rv (maximum valley depth) is one of the options on our formtracer Re is not but a google search shows it is not a standard primary roughness parameter but may be in reference to Rc or Rp. I have not used any of these parameters, so I have nothing else

I usually use getActual().inTolerance and from what I have seen it works great.

Can you provide more background what exactly is meant with Undercut. Undercut could be in general a root relief or Undercut evaluation characteristic which is not defined in standards. I know that e.g. Klinglenberg has such an charactersistic but you have to clarify what exatly has to be evaluated. Same for the other charateristic. Maybe you can share the drawing and the .act via PM?

Download Free CAD Models, STEP Files, and Test Results | NIST PMI Information and sample CAD files.

You're welcome. I apologize for not giving better information to start with.

From the great @Tom Oakes: You can create what I have designated as "Ghost Holders". Let's say you have one physical 3 position XXT holder. I add a new stylus system holder. Then, define the location of the new holder using the original #1 holder. I put my standard configured probe in the #1 holder and my RDS-CAA configured probe in the #2 holder. When I want to switch from standard to CAA, it stores the standard version and then picks it back up as the CAA version. Obviously, this can cause some confusion so everyone has to be on the same page.

Anything higher than 0.0005 is unusual for us. (That is including using RDS+CAA+0.6mm tipped L-probes with extensions)

You can set up and qualify a probe as a CAA probe. Then set up and qualify the same physical probe as a list-probe. Calypso will let you assign as many probes as you like to an adapter plate/chipID. The caveat is that you cannot have both probes installed into the probe rack at the same time (since they would need to use the same holder). To make this work we have programs that are either "only CAA", or "no CAA" for that probe and we reassign the correct probe to the holder before running it.

I'm presuming that you visually inspected the stylus for damage (under magnification)?

I wonder - have you tryied "Polygonize large data volumes"? It should handle such situations.

would be difficult, pretty sure gear pro does not have this, even if it did, you would need the proper ball to shaft ratio probe tip to not shank. Looks like you do have it measured - which is a start. Looks like you may be able to add a root relief and try that, or possibly add a K-band in this area as well. Worst case, try reading on chart, counting dots - not ideal, however.

there is also an option to 'align to feature' which should do it for you.

When you load the same mesh as CAD it should show no deviations. Working as intended. Thats why i mentioned a reference sample or master part. To my knowledge, you can`t compare freeform without a CAD. The CAD is the information for the comparision.

When i load it as a CAD model there is ZERO deviation. How would i show Actual mesh data having low and high spots without using CAD model?

We are currently creating a new template for ISO GPS-related dimensions. Design intends: Datum A is the mid plane of the 2 drafted planes A1 and A2 Datum B is the mid plane of the 2 drafted planes B1 and B2 Datum C is a plane normal to the axiy. we have noticed that the actual fitted plane and the displayed coordinate are not perfectly aligned. There appears to be a slight angular difference as well as a small positional offset between them. computation mode ISO 1101 or geometry shown similar deviation but exactly the same. Any idea how to fix it?

It all depends on what you do with it how many stages and how complex/large your parts are. what you work with: i got several offers for different configurations and the Pc for the Scanbox they proposed is significantly more powerful than for other setups. How clean your workflow is. Done perfectly the need for stronger hardware decreases. I for myself hate waiting and sometimes my workflow could be a bit cleaner so i advocate for as powerful as possible with "normal" desktop hardware. Something like this (from 05.2025) or more recent equivalents. The Ada CPU are extremely expensive. I can not tell if a consumer CPU would work aswell. I also can not tell you how good any more cores for the CPU would scale the effective power level.

Glad you are having better luck than we are. We use NX at my company, and it is like pulling teeth to get data to transfer. I was told a JT file gives the best import results to Calypso, but I found that an NX converter add-on in Calypso is giving "better" results. These results are still not enough, as they only generate linear dimensions for me so far. From what I can tell Calypso still needs a lot of improvement in this department. I hope you have better luck on your end. I appreciate you sharing any additional insight from your findings.

Martin, I've thought about upgrading my laptop but the fans have just started to sound bad and it's nearing it's planned replacement. Jochen, how much more powerful would you specc the stationary? I've checked on Dell and their T2 tower look like it could be specced to be too powerfull.

Do you have PCM? If so: getRTOffset() Rotary Table Offset value getRTPosition() Latest position of Rotary table getRTPositionCMM() Latest position of Rotary table in machine coordinate-system rtPositionOffset() Define offset for RT (angle in radians) rotateAbsolute() rotateReference() goHome(vector) Rotates RT to specified vector rtIsActive() Boolean value, Is rotary table active? true/false rotateWithDriveFree() Boolean value, Is rotary table active? true/false

It works correctly in the result element if said result element is below the characteristic it is evaluating. Unfortunately, this won't work for what I need. I ended up just grabbing the actual, upper limit, and lower limit and writing an If statement. With this method, I couldn't just use read a PCM file, because some characteristics behave differently with grabbing actuals. So, I had to copy and paste because those certain characteristics (caliper distance, angles, etc) needed modified code. I'll probably look back into it later in the future, but as of now I found an okay way around it.

You have to use getNominal instead of getActual: getNominal("MyProbe").toleranceState

Thank you @Richard Stanich

Hello, Could someone explain me how to evaluate undercut and max edge finish? This is how I made it: I have arguing our engineers of this topic, because they say the tolerance should be from -20 to 90 microns.

The same for me.