Programming a microecetric grinder using mitshubushi

I have been trying to reprogram a machine 3 weeks now, I finally told my boss I was hitting a brick wall. Told him to call a program Integrator, got our 1st quote today it was close to 30 thousand dollars. My company wants to get something for nothing so I am afraid I will have to get a big hammer and try to knock that wall down. At least the integrator said he would delete program and start over,he said original program left a lot to be desired. He did spot something i missed E-stop, hooked into processor. I hardwired E-Stop already. I guess I was to overwhelmed by what I was seeing I missed the easy fix.
Now my question does anyone know where I may get additional assistance with this questionable program/crash course.
It's problem consists of flying data, things move when not told to move and it appears it all comes down to the data the original programmer setup. He liked to use DSUB, Dmov, DADD,COMP, all of which the integrator said was not necessary.
I have gone to shool on plc's about 10 years ago for about six months needles to say I need more education and experiance, but if I do back and educate myself I will probable never see a processor for another 10 years.

You could start by downloading the programming manuals from here

http://www.meau.com/eprise/main/Web_Site_Pages/Public/Documents_Downloads/P-DD-Technical-Manuals

If you post fairly detailed explanation of what machine
is supposed to do, you will get some tips and suggestions.
Once you have a good understanding of the process it should
be not easy but straight forwart to implement piece by piece.
Lot's of equipment I deal with uses analog I/O, communication,
motion, repetitive sequences etc. Try to elaborate on
what you have to achieve and someone will try to help.

monk,
How big is this program,steps? If its a big program and there is a lot of data being moved, try to break it down into steps. Don't look at the whole picture, pick it apart.
Good Luck

Hardware used for project above

The non working program consist of 5311 steps. It's general cycling abilities are correct. But it often does strange things. Here is a list of it's Hardware and sequence for machine. Fx2n32MR,4-FX2n-16EX-ES extensions,2-FX2N-16EYR extensions, 2-1PG cards,2-FX2N-4DA analog cards 2-MR-J2 sevo cards,1-E900 terminal 2-mitsubishi AC servo motors. Basic moves 2 slides infeed and dresser. Sequence in automatic as follows. Push cycle starts, After arm goes down for loading magnetcic chuck goes negative,after part nears shoes chuck voltage is positive,arm translates out, goes to pick up a new part,arm translates in,after arm translates in goes to rapid slide in,goes to course feed at value inserted into E900/predetermined,goes into fine feed at predetermined pos.,goes into tarry feed at predeterminded pos.,at tarry Sparkout timmer activates or dwells part on wheel,after a predetermined count /values again punched into E900 then dresses wheel compensates the infeed slide and dresser slide,during cycle only infeed slide is moved. Then above sequence is repeated load,feed,check if it ready for dress if not ready repeat cycle until ready to dress or cycle stop is energized. My problem lays in the different slide movements sometimes they will do things unexpected and crash machine. I expect something within data is either told to change or is changing at random. I expect told to change is my problem. But this action is random and may not occur for hours,but when it does it is quite distructive.

I feel for you.
Been there and done that.
My only thoughts are. Has this machine been running fine for quite some time now without any problems? Any modifications to the program lately?

program grinder

As of yet the machine has not ran it's 1st part it could. But with the problems I see I will not release it. To many things wrong. It would probably run some quality parts, but I prefer running without these bugs. It's sister machine has run, but it is weekly blowing wheels off and tearing up tooling as well as expensive spindles on dresser. I am afraid eventually it will fly apart and hurt someone and I do not want this to happen to anyone. My thinking is that at it's early stage of design they should have put a little more thought in it's hardware and perhaps bought a processor that you could use correct modules so a person may use the absolute encoders off of the servo motors, making the programming a little easier. Oh well sooner or later with help this machine may run correctly the way it is.

Re: program grinder

monkworm said:

As of yet the machine has not ran it's 1st part it could

Click to expand...

If this machine is new, then it needs to go right back to the manufacturer... :nodi:

If used, then the person in your company who purchased this 'gem' has some 'splainin' to do...

This mess is not YOUR problem. That is, unless it was YOU who authorized it's purchase... utoh

beerchug

-Eric

programming grinder

My Factory were having a retrofit company put these machines together, this company then proceeded to go out of business. Our company then had a option to have some of the people that were working for this out of buisness company finish the project, well to cut a story short our company was mad, we had just got bought out by a Japaneese company. They were very upset and told us to put these machine together at our facility, someone had told them we had the proper personnel to do this endevor. As you may relize blame gets passed down the ladder,nobody wants to make a decision on this project as of yet just keep passing it down the chain and right now I am on the bottom link. They cannot blame me at the moment because when these machine were received 3 years ago I told them in front of many witnesses to have a program integrator come in write program and then we would wire to their specifications. Right now it stands, try to make program work correctly, as reported above I really have no ideal at this time,but will keep trying. Hopefully though someone will make a decision to hire a integrator to do things right.
Thanks for your many responses.
Monk

hi monkworm

Have read your plight with a microcentric grinder.I have worked on and indeed rebuilt and modified these grinders. HOWEVER, i have only used rockwell controls for this. If you HAVE to sort this problem and can not walk away from it i will try to help.

Can you do the following?

Cycle the machine without a wheel fitted, this just makes things safer!!

Switch the dressing off and just let the machine dry grind cycle?

I am assuming of course that you can connect to the mitsi and de-bug the program, otherwise i am no help to you.

Let me know and i will try to help

regards

kevin

I know when I first used Mitsubishi (the first PLC I programmed) I had made the mistake of using a couple of registers that were part of the double registers used in other commands. In short the double uses two registers and only one is referenced, therefore it is possible to use the second register elsware without it being obvious that it's already used. I remember something being odd as well, I can't quite remember if it's that doubles reverse the order or if they use the register before the one referenced, you might want to look into how doubles are handled, it's been a while for me.

I think Desert dog has a very good point. It is a common mistake with mitsi That DMOV and DADD are double words. So if the command says something like DADD d10 d20 d30 (meaning add double word d10 to doubleword d20 and place the result in d30)
you have actually added d10, d11 to d20, d21 and the result has been placed in d30, d31.

It is possible to use the d11 d21 d31 elswhere without realising they are already used and this makes for incorrect maths popping up all over.