Im a Demis Newbie, Im really good at C@!^pso, what should i know......

I DO NOT like the use of “move” or “clearplane” points, in Calypso the probe just moves in and out of the part between features pretty much automatically.

When using autofeatures, there is a setting that allows you have the probe move in/out from the feature.

As you have a rack you can write a program to do all your calibration.
(autocalibration)
Look up parameter sets in the help file for additional information on this.
You dont have to use move or clearplanes if you turn on avoidance moves.
But be careful where you put them.
Most people prefer clearance planes as you can set them so you dont have crashes.
pcdmis is stable enough when installed, maintained and used properly.
Just because some people keep having crashes doesnt mean we all do.
If you have a ph10 head get the boss to invest in an sp25 if you want to do scanning.

Hi,

Many years ago after 9 years as a PC-DMIS programmer I was desperately looking for work (was a victim of downsizing). I found a job but it was programming C@!^pso. I figured it must be similar and just like you found it to be substantially different. I am now back to programming PC-DMIS after convincing the company to get a B&S machine. I know the differences very well. Don't lose hope because PC-DMIS has many, many shining qualities.
The first one to mention is the Edit Window: I always run from Command mode, almost everything is there to edit or no more than one click away (you don't have to dig through dialog box after dialog box to change stuff).
Vectors: You will need to learn vectors because that is key to PC-DMIS. Yes, in Calypso you could see vectors but you couldn't touch them.
Calibration: By no means are calibrations tied to programs as a matter of fact calibrations are extremely powerful. Yes, it is true that you have to be in a program to calibrate a probe but it doesn't have to be the probe used in that program. You only need to calibrate what you need to (Mark Used) and if you need to change to a different probe you just select the probe you want and the A,B angles all, by design, match (no choosing which left or front goes with which right or back) AND you don't have to requalify every probe every day.
Points: Yes, PC-DMIS is all about points. It's about making collections of points that are points, circles, cylinders, planes, scans, etc. Calypso is all about circle functions, cylinder functions, plane functions, etc. With PC-DMIS you are able to precisely position points and precisely position points based off of other features to measure features that you never thought were possible before. I do it everyday.
Patterns: In PC-DMIS you can create a feature, a constructed feature, a dimension, and anything else related. Select and copy all of this. Then create a pattern and everything is patterned into new features. (all just as easily editable as the original, not "locked" up inside of a pattern function).
The CTRL keys: ctrl U (run program from cursor position), ctrl E (run single feature), etc are very big allies.
Alignments: alignments in PC-DMIS are very, very powerful, not limited to only using specific features, and not limited to the amount of constructed features you can use. One thing I did like about Calypso was you could rotate to an X, Y, Z coordinate in "space". In PC-DMIS I do a construction between two separate alignments to accomplish this. In PC-DMIS alignments are "daisy-chained" throughout the program (each builds off of the previous) or recall alignments bring back previous alignments. They are not "tied" to each feature.

All in all I find that I am at least 3 to 4 times more productive and by far can measure parts more comprehensively than my Calypso counterparts. So don't be discouraged, it is a little bit of a learning curve to switch.

I went through training for Calypso, and ran a Zeiss CMM for about six months while we had it in-house for a demo. Ultimately, I chose Brown&Sharpe CMMs for our company as Calypso couldn't support our factory automation goals. Overall, PC-DMIS is a more open-ended system than Calypso. You can measure pretty much anything, though more complex tasks my require you have good command of trigonometry and Euclidean geometry.

All that said, this is probably my best comparison of the two:

Calypso is a more polished product, similar to an Apple computer, but it does tie your hands at times. PC-DMIS, like Windows, is a bit rougher around the edges, but you have a lot more freedom so long as you can figure out how to use it.

I DO sorta like the flexibility of the articulating head (Although I do question the accuracy).

I'm with you there, that's why a majority of the machines I've purchased have fixed heads. We try to use star styli configurations for accessing features as much as possible, even on machines with an articulating head. I've never been able to recognize a distinguishable difference in accuracy, but I do get nervous when I have a large scanning probe and long stylus swinging around on the machine.

I DO like the resistance to breakage due to the probes being held together by magnets and the TP20 range of motion.

If you're not breaking things, you're not trying hard enough!

I DO NOT like the heavy use of individual "Hits" instead of the scanning head on the Vast Xt (On my B&S that’s a different attachment)

Using touch point measurements is generally more accurate than scanning, so long you use proper technique (quantity and distribution of measurement points). Scanning collects more data, which is great for evaluating form, but introduces its own distortion artifacts which reduces accuracy. That said, we use scanning religiously where I work due to form requirements (mostly roundness) on our parts. The more form deviation that is expected on a feature of a part, the more points you will need to evaluate the feature.

I DO NOT like that you can’t just qualify all the probes in the rack at once, you have to open different files and qualify the probes used in that file, then open another file and qualify those probes...... and so on. (To better explain, the qualifications seem to be tied to programs, instead of being a machine function)

You can add a tip qualification as a command into a program. We have "auto calibration" programs we run every shift that qualifies all standard styli builds on our machines, and then checks a ring gage with all of the tips. Just search for "auto calibrate", "auto calibration", "auto qualification", or similar terms here on this forum.

I DO NOT like the use of “move” or “clearplane” points, in Calypso the probe just moves in and out of the part between features pretty much automatically.

This has been a fairly fundamental difference between PC-DMIS and Calypso until very recently. The newest version, PC-DMIS 2012, has added the ability to build a "clearance cube" around your workpiece, which functions almost identically to Calypso. Clearance planes are similar, though they only work in one plane rather than as a cube around the part. That said, I prefer move points, largely due to my back ground as a CNC programmer.

I DO NOT like the Stability of the Demis platform (see all the threads in here about unexplained shutdowns and program crashes).

There's more to this than just PC-DMIS. I have worked closely with my IT department to make our PCs as stable as possible. We set up one new PC, with a clean install of just Windows 7 and PC-DMIS, and ran a CMM with it for a few months. When everything was working good, we made an image of that PC and copied onto new PCs for all our other CMMs. There's no extra junk like an ERP, timeclock, email, or anything else that could cause an application conflict. We get maybe one crash per month at most, which is actually better than I've had with Calypso on some workstations.

That said, we have one manual CMM on our shop floor, which our old IT guy tried to lock down as much as possible. You can't even right click anywhere, save files anywhere, etc. It crashes every few minutes. I believe the majority of its problems are permissions.

PC-DMIS is a big program, with a massive amount of registry entries tied to it, and files stored in lots of places (including places Microsoft says you're not supposed to put things). Like many large applications for niche markets, there's not going to be same resources behind development and testing as there is with say Solidworks, CATIA or Photoshop. I think everyone needs to try CNC programming with Esprit before they're allowed to complain about PC-DMIS.

...and I have all my years of experience in Zeiss Calypso.

That's okay, we forgive you.

Of all the things stated in this thread no one has mentioned the fact that in PCD you can see and edit the code. It is the biggest drawback to Calypso that I can't assign a variable to do some of the math and then construct a feature using the variables.

You can check the accuracy of your PH9/PH10 head very easily, and I do it often (say, every 6 months or so).

Calibrate a variety of angles (I use 0,0, 45,-135, 45,-45, 45,45, 45,135, 90,-180, 90-90, 90,0, & 90,90) and I use a "long" build up (to see the most error) of the PH9, 50mm ext., TP2, 10mm ext, and 4x20 probe. Then make a program that checks a sphere using all those angles. Run the program 10 times, saving the data for statistical anylisis. My head always repeats within MACHINE specs (not machine & head & probe specs, but within MACHINE specs), and you can't get any better than that. Can't see how a fixed head would be any better. I mean, can't be any better than machine specs, right?

Can't see how a fixed head would be any better. I mean, can't be any better than machine specs, right?

It's one less piece to worry about wearing out and gradually losing its accuracy/repeatability. That said, none of my machines have had issue passing first order specs, with or without an indexing probe head.

It's one less piece to worry about wearing out and gradually losing its accuracy/repeatability. That said, none of my machines have had issue passing first order specs, with or without an indexing probe head.

Just like everything else, including the CMM itself, YOU the operator/programmer must keep track of it's ability to do it's job as it should or you are just wasting your time in QC.

Of all the things stated in this thread no one has mentioned the fact that in PCD you can see and edit the code. It is the biggest drawback to Calypso that I can't assign a variable to do some of the math and then construct a feature using the variables.

There are parametric programming methods in Calypso, though it's not as straight forward as using features names in the PC-DMIS edit window. The engineers from Zeiss worked with me on developing parametric programs during the time I had the demo machine.

Everyone has posted some very good and accurate stuff, but I'll throw in a few more ones and zeros:

1) Decisions:
Calypso is setup to be heavy-handed with making decisions for you. This can speed you up if the decisions are correct for your application, or slow you down and cause problems if they are not.
PC-DMIS leaves far more decision-making to you.

2) When you change settings:
PC-DMIS has nearly every setting, whether temporary or global, available with one hot-key and two-clicks, or three-drop-down-menu layers. With practice you can make changes very quickly.
Calypso has most of it's settings buried deeper within 5 to 7 drop-down menus or worse yet sub-sub-sub-windows. However it does have a far more powerful and user-friendly "change many identical things at once" editor. PC-DMIS has one (DATFIELDEDIT Wizard) but it's not as friendly.

3) Reporting:
Both have extensive report-customization options, and both are cumbersome. PC-DMIS does have an edge with more graphical reporting available with fewer clicks.

4) Automation and High-Level-Language:
PC-DMIS's automation has an ease-of-use factor that eclipses Calypso by a factor of over 9000. It becomes part of the code, rather than being buried in hidden menus of mysterious connection like Calypso's PCM, and is very simple to write and execute with logic. PC-DMIS can also be very easily fully automated using Visual Basic, making it a natural fit for robotic and linked-system applications.


Making the software transition is hard (BTDT) but it is a quite valuable experience that will ultimately make you a more powerful programmer in any language. Stay goal-focused rather than methodology-focused, and learn/invent/prove new methods as you hit your goals.

- Josh