Okuma issue

[Leigh @ Kodiak]

I just had the strangest thing occur on our Okuma Mill. We have been running the same program for two shifts with no issues. The operator this morning ran two complete cycles just fine. Third cycle, the first tool is going fine, and just suddenly took off in X and Y while in the middle of a sub program! The sub program had already been ran on 4 of the parts in the cycle, and the random move happened in the 5th part! Does any bod have a guess at WTF might be going on here?

[RaiderX]

very strange indeed. what is your level of trust that you are getting a straight story?

jus sayin, machines usually dont just take off. and Okuma controlls are pretty solid.

[Leigh @ Kodiak]

Level of trust is pretty high. The operator is just a parts changer. Does not know how to even use the handwheel. I heard the bang from my office and was there within seconds. I have now changed the tool and restarted the program. got through the trouble area no problem. Even stranger, the point the machine took off from was in the middle of an arc moove. Could a power surge, or dirty power cause this? I have seen other issues with the machines lately that could possibly be related to dirty power.

[Reko]

+1 to Lazyboy

 

25+ years of CNC experience and I have never witnessed a machine tool randomly deviating from a program.

 

Sounds like someone was tinkering with the program.

[Leigh @ Kodiak]

There is NO way he could tinker with the program. I was there within seconds, and the program checks out as identical to what I have in my backup folder.

[gcode]

Years ago. I was a new hire in a shop, running Tree CNC mill

4 times in 2 days it took off and ran for home, breaking the tool and scrapping the part.

My head was on the chopping block and my final check was already cut when I figured it out.

The machine was on the same curcuit breaker as the the air compressor.

If the machine was in G02/G03 and the air compressor started, the power spike sent the machine

rapiding for home.

 

I wouldn;t think a modern Okuma would do this, but I'm thinking you and a service call in your future.

Something like this could cause heavy damage if it happened at the wrong time.

I'd have machine tech look into it.

[Brando]

definetly sounds shady

 

 

modern cnc machines just dont take off, and if there was a machine issue im 99% sure there would be some sort of alarm generated, besides okumas osp cncs are rock solid , only 2 things should be wrong , either you have a program issue or a operator issue, maybe he was foolin around with pulse handle interrupt button,

 

i suggest start using the program edit lock out key

and im pretty sure the osp control has a panel button tracing function, so if the operator is telling the truth you should be able to confirm that

 

and make sure you find out what it was too bc you'll regret it later if the operator thinks he can get away with that stuff..

[mig]

Make sure you are not doing any feed movement if you are on limit switches (z for example).

This could be a reason

[Mr. M]

and im pretty sure the osp control has a panel button tracing function, so if the operator is telling the truth you should be able to confirm that

That they do, what control is on this machine?

[Joe788]

Does the machine have it's own dedicated ground rod, driven 8 feet into the dirt?

[Leigh @ Kodiak]

Does the machine have it's own dedicated ground rod, driven 8 feet into the dirt?

 

Yea, this it definitely does not have. You don't want to know why I know this.

 

I checked in the operator history screen, and I don't see anything out of the ordinary.

[?Mark]

Funny, Just had a very similar thing happen here on the night shift. My program, Very expensive part, they were on 6th part of this batch already with no issues and the program was proven before (few dozens of parts made before). Total cycle time on this op is about 12 hrs and half way thru it the machine deviates slightly from the program, we get PS191 alarm and a tool and part are history...luckily it's all a "very" thin walled part so no other damage done -

 

Years ago I personally witnessed a kuraki vmc simply moving the z axis spindle down with no reason at least twice...scared the crap out of me but finally hitting the e-stop stopped it...we found out it was one of the lightning strikes that nuked it's communication board...I know...weird.

[Joe788]

Yea, this it definitely does not have. You don't want to know why I know this.

 

I checked in the operator history screen, and I don't see anything out of the ordinary.

 

The first thing a tech is going to say (other than questioning the operator's story), is "Does it have a ground rod?"

 

I'd start there and cross your fingers that it doesn't happen again.

[Brando]

well its possible with not having a ground rod installed (which is typical for most job shops),, youve had a electrical spike/strike/whatever and its already done some sort of damage to a circuit board and is now showing up

which really sucks bc intermittent electrical issues are near impossible to locate and a service tech is just gonna start replacing expensive boards until it stops doing what its doing

[chris m]

Ground issues typically show up in serial communication more than things like what you're describing, but you never know.

 

I have seen things like what you described with Fanucs when the control enclosure (the main electrical cabinet, usually) got really hot. I also caused something similar by turning the pulse handle on one of our Okumas during auto mode a few years ago, but I cannot remember exactly how.

[YoDoug®]

On older Okuma controls (non-P series) there are a few functions that are all hardwired vs software controlled that could cause the problem. The mid-auto manual function and feedrate override are all tied into the same circuits. We had a customer that had a button on an OSP7000L control that was shorting out intermittently and the result is the lathe would take off in X when it shorted.

[Leigh @ Kodiak]

For the record, it's an OSP-U100-M. Have not had any issues since.

[JerryBenoit]

Well if I might add. The first time that I ran a CNC was in 87" on a brand new Truab. In the two years that I work for this company we had issues with the X axis jumping 2mm at a time. NO offset adjusted and NO operator influence. Just take the parts out of the catcher and mic the parts. Now this went on for two years the tech came in and combed the machine over. This machine recorded all keyboard entries this is the only reason why the believed me that I made no changes. Just before I got done at that shop they found the problem, which was a power spike and over time all the x axis boards were fried.

 

So if you are comfortable with your operator, look hard at your power.

 

Jerry

[chris m]

For the record, it's an OSP-U100-M. Have not had any issues since

 

That is a pretty solid control, though not my favorite, and we've not experienced any problems with ours. With that control, the machine must be [10]-[12] years old or so? If you've not experienced this issue before I'd suggest looking hard at anything different in your shop recently

[beej]

we have all okumas here. I had it happen to me once on my okuma cadet mate. the problem ended up being caused by the excessive chips built up under the table. the chips compacted, put excessive pressure on the Y axis and the machine did a little bounce back action that made it look like it was taking off into the middle of the part. the problem was up under the way covers. over time chips had gotten blowed up under there. kind of a poor design.

[Del.]

we have all okumas here. I had it happen to me once on my okuma cadet mate. the problem ended up being caused by the excessive chips built up under the table. the chips compacted, put excessive pressure on the Y axis and the machine did a little bounce back action that made it look like it was taking off into the middle of the part. the problem was up under the way covers. over time chips had gotten blowed up under there. kind of a poor design.

 

 

Dang Beej, I thought you had fallen off the face of the Earth.

[Leigh @ Kodiak]

Oh lucky day. This happened to me again today.

[K2csq7]

Very odd...

I had an issue with a cadet mate like beej (hi beej), way covers damaged, chips built up, messed with limit switches.... but a machine zooming off for no apparent reason.... this I have never seen.

Call the power company, ask if anything abnormal is happening in your area.... you may not have noticed any "flickering" but they will certainly know for sure....

[K2csq7]

another step you could take...

alter the program to isolate that contour... make it it's own operation at the end of the program or something.... just to see if you can deduce/eliminate/gain any information.

[Leigh @ Kodiak]

Thats the great part. Even though its the same program, the machine took off at a completely different point in the program.