Thread: Renishaw probe weird behavior

Today on the PFH_4800 I had the strangest thing happen.

I use probing all the time within a program to locate parts that are hard to locate precisely with hard stops. Today I was standing next to the door working on a program and just happened to see that the spot drill didn't spot. Just lucky timing to look in I guess.


The part had probed X and Z and both were wrong. It appears that it just recorded the start points of the probing rather than the result of touching. The optical pickup had some chips on it, and the only thing I can think of was that it couldn't correctly see the probe and rather than Alarm out it accepted the start point as having tripped. The location wasn't even close, so no chips were involved.

Very strange, I sat there for the next few cycles and made sure that it updated the WPC each time it probed the part. No further problams.

The idea of having the wrong value entered in the WPC was very scary as we all know how bad that can be.

I have a .4" start point for the probing for safety and that was close to the error in location.

Has this happened to anyone else?

Thanks in advance.

Techguy

I've never had that happen before, but it's good practice to set a tolerance for the location and or size of the feature you're probing. If you know your fixture repeats within .005, just set the tolerance to .006. If this happens again, you'll get an alarm.

Check your tool change motion, pull stud and chips in the spindle or on the tool taper.

Chips usually cause a "closer" error and not a mile off.

Tool change motion: I've seen it where the ATC timing was a bit off and once in awhile, a tool gets clamped when it's not all the way in the hole yet. BAsically, the drawbar collet is clamping on the major diamet or your pull stud. If your probe ran like that, you could be a mile off in X Y and/or Z.

Pullstud: A loose pullstud can have the same effect here where running tool length is now longer than the Data length. Which equals bad probing and it can be intermittent...

Other than that... human error... Did someone start that particular part run and skipped the probing? Or perhaps had to restart the machine but re-read G10 lines (if you use them) accidently (and therefore the probing update got negated)?

Anyway, as Joe mentions, you can put in tolerance to check for this while probing. Even if you don't rely on 'hard stops'... I still use them for closer loading reference points to be able to tolerance reasonably...

I've seen some really weird behavior when the battery is starting to die.

Thanks for the input guys, I will check the pullstud, etc.

Joe, I am looking at the parameter book and I do not see a parameter to use for a tollerance limit for x-face and z-face probing. I see one for manual measurement and skew, but not for face. Any help would be great.

I was running the parts at the time and I am sure I didn't interupt the program and restart.

The battery is new.

I was just hoping someone had seen this happen when the probe sensor had chips on it. I have had alarms before if the sensor signal was low because of chips.

Thanks guys.

Ahhh, I've never used the Mazatrol probing routines. I've only used the Renishaw Inspection Plus macros.

Psychomill will be able to tell you how to add a tolerance to the Mazatrol routine (if it's possible).

Joe, always nice to get your input.

You have never used the Mazatrol probing routines, you are missing out on one of lifes greatest things. Just kidding, I do about 80% of my programming in Mazatrol and use probing all the time inside programs.

Of course that could be because I suck at Mastercam and hate EIA programs. Ignorance is no excuse for me though. I guess when you are doing everything from programmer, parts loader, and delivery boy, my learning curve takes longer. I do have a parts loading jocky now so I don't have to talk to myself all day. I still do talk to myself while programming though, and hand gesture a lot.

I run all of my EIA programs within Mazatrol and use the Mazatrol probing to adjust WPCs for G offsets. I feel safer that way.

Thanks again for the input guys.

Techguy

One time I was at the shop wasting time on a Saturday, and I said to myself, "Self, you should explore Mazatrol a little bit."

I went to the program screen, typed in a new work #, and for the first time ever, hit the Mazatrol button. I just stared at that black Mazatrol screen for about 30 seconds, and then walked off to do something else. That's the extent of my Mazatrol experience.

Mazatrol probing will get you by for many things and for something as simple as WPC updates.... it's simple and fast. However, you lack true control and in fact, are missing out on many many things compared to EIA probe programming. But, that's another story...

Mazatrol probing doesn't have a tolerence feature that I know of. There are things you can do with the probe data based on measured features but that's not what you're looking for. The only way I've done reference tolerancing in Mazatrol is to write in a macro to compare a reference offset to the measured offset. If that value exceeds your requirement, you get a macro alarm.