What's new
What's new

Help needed on intermittent ATC Halt... Waiting on FIN signal. Yasnac MX3 control.

gmoushon

Cast Iron
Joined
May 18, 2006
Location
Illinois
This is my first CNC machine, so forgive me if this is basic stuff.

Bought a used Mycenter 0 which has a Yasnac MX3 controller. (1995) I have the machine up and running and have done my first programming and test cut.

Everything went as expected except, towards the end of the code, the machine stopped when I called a tool change. This was a tool change that had been done multiple times before without incident, but I needed to adjust the program a little more and was running through it again when the ATC halted. An "M" was blinking on the screen and according to the book, this indicates the control is waiting for a "FIN" signal.

To recover, I placed the machine in handle, cycled the carousel a few times, navigated back to the tool change line in the code, and continued the program. This worked.

From what I've read on the net, I guessing it is a buffering issue. Some have suggested adding a couple of EOB commands after the tool change command to clear a buffer, but this sounds odd to me.

Wondering if someone could help me understand the likely cause.

Also, when I do a tool change, I call a subprogram to cut down a little on the amount of code. Here's what I execute to change tools:

T2 M98 P30;


O0030;
G0 G53 Z0.;
M6;
M9;
M1;
M99;


Don't know if doing a tool change this way is causing problems with this Yasnac control or not. I heard Yasnac can be a little difficult compared to others.

Thanks for any feedback.
gm
 
Warning, danger,danger!!! It could also be a sign of impending doom for the tool change motor inverter. The settings for it need to be manually removed, if it fails or is repaired it could likely lose those machine specific settings, and a lot of machine tool manufactures even if still in business no longer have that info available on a 20 year old machine. I would extract and write them down ASAP. What brand of machine is it? I hated my MX3 controller, was tired of the breakdowns and happily sold the machine for parts.
 
The machine is a Kitamura Mycenter 0.

When you say "tool change motor inverter", are you referring to the servo drive controller...the variable frequency drive? If so, this was replaced a few years before I purchased the machine. Also, I am fortunate enough to have the original factory documentation with all of the memory address settings etc., if that's what you're referring to.
 
I had the exact same machine if it is original it will say Mitsubishi Freqrol-Z024 on it. It is in the cabinet halfway up on the left side. I doubt it's settings are in the original docs usually they just list parameters and option codes. If it is a Mitsubishi Freqrol-Z024 it wasn't "replaced" it was repaired. Look for an alarm code on that drive at the next sign of trouble, I hope you have the book on it. I would clean all the proximity sensors first.
 
It also would help to know what halted the tool change process. Did the tool carousel index correctly? Did the tool holder make it out into the staging area? Did the change arm cause the problem by not being in home position?
 
It did not always halt at the same position. One time it halted without indexing the carousel. Another time it did move the tool holder to the staging area. The tool changer arm did not have any issues.

I just dug up the paperwork and found an old invoice dated 6-29-04 showing that the main servo control, a VMC25P5, was rebuilt.

The ATC control is a Mitsubishi Freqrol-Z024 and I cannot find any reference of work being done on it.

This is in-spite of the fact that the machine dealer stated this on his quote:
"New spindle drive installed 6 months ago"
"New Mitsubishi Vector drive Freqrol Z024 installed about one year ago"

Lovely.

I do have the book on the Freqrol, but unfortunately, all the buttons on the actual drive have only Japanese characters! You don't happen to speak Japanese do you? (Right)

Guess I'll start with the proximity sensors...if I can find them.

How many would I be looking for. Two?

gm
 
I no longer have anything to look at, the machine I got rid of had been down for years before I scrapped it so we are relying on my not so good memory. One good thing all the wires should be numbered at
the device they connect to on both ends. Without a wiring diagram I seem to recall that Freqrol may only run the gear motor that controls the arm and you said the arm is fine. There should be proximity switches for carousel position, position of the staging ram, and there will be an interlock on the door to the tool area. I think the staging ram is run by air so check for leaks, low pressure will alarm out the machine but inconsistent pressure will throw the timing off. Find the carousel gear motor and see what that hooks to. You could also have a lose or corroded connection somewhere. On the Freqrol even though the outside characters are in Japanese the button position dictates what they do, the reset button will be in the same place whether it says "reset" or is in Japaenese.
 
Thanks. I'll check it out.

One other thing. Do you see any issues with the way I program the tool change and do you think there may be a buffering issue?
 
Thanks. I'll check it out.

One other thing. Do you see any issues with the way I program the tool change and do you think there may be a buffering issue?

I never used sub programs, I will look around to see if I can find some old programs, I may have dumped them when I gave up and scrapped the machine. I remember running the rapids at 50% to avoid issues and also using cutter comp being a nuisance as you seemed to have to vary your approaches depending on part configuration to get it to work correctly. Buffering should not be an issue as there are delay settings to wait for signals built into the system. No matter how fast or how many times you tell it to do something, the control gives the machine X-amount of time to do it before alarming out. Not sure of your tech knowledge level, I would call myself "B grade." If you aren't aware all Yasnac manuals are free for download in PDF form from Yaskawa.com
Don't know if you have called Kitamura yet, but they would try to sell me a new machine anytime I ordered parts. Trying to help you is bringing back bad memories. I later noticed the Fanuc controlled Kitamuras went for an average of $5k more than the Yasnac ones when shopping used, now I know why.
 
This is my first CNC machine, so forgive me if this is basic stuff.

Bought a used Mycenter 0 which has a Yasnac MX3 controller. (1995) I have the machine up and running and have done my first programming and test cut.

Everything went as expected except, towards the end of the code, the machine stopped when I called a tool change. This was a tool change that had been done multiple times before without incident, but I needed to adjust the program a little more and was running through it again when the ATC halted. An "M" was blinking on the screen and according to the book, this indicates the control is waiting for a "FIN" signal.

To recover, I placed the machine in handle, cycled the carousel a few times, navigated back to the tool change line in the code, and continued the program. This worked.

From what I've read on the net, I guessing it is a buffering issue. Some have suggested adding a couple of EOB commands after the tool change command to clear a buffer, but this sounds odd to me.

Wondering if someone could help me understand the likely cause.

Also, when I do a tool change, I call a subprogram to cut down a little on the amount of code. Here's what I execute to change tools:

T2 M98 P30;


O0030;
G0 G53 Z0.;
M6;
M9;
M1;
M99;


Don't know if doing a tool change this way is causing problems with this Yasnac control or not. I heard Yasnac can be a little difficult compared to others.

Thanks for any feedback.
gm
I am not a programmer but it looks to me like you are precalling T2 and then going into the sub program to do what? If you put T2M6 on the same line doesn't the tool change normally? I have never seen a sub used for a tool change.
 
I don't think your programming is the issue, but it would be easy to test by writing it all out in your main program. As Dualkit said, it is not a "buffering" issue.

Sounds like a limit or proximity switch is not being turned on in some step of the toolchange. Since it stops at different steps in the process, I'd start looking at all the toolchange related inputs. A broken wire, bad connection, or numerous other things are possibilities. Unfortunately Kitamura did not write a very thorough ladder program. Yaskawa provides the machine builder with the ability to display ladder alarms and messages on the CRT if the builder chooses to write the logic for that.
 
You don't need all that stuff to do a tool change. The machine uses a macro to do that tool change. The macro is actually a subprogram of its own.

When the machine gets a M6T2 command, it will do the following things:
1) move Z to tool change position
2) turn off the spindle
3) orient the spindle
4) index the turret
5) engage the swing arm
6) unclamp the tool
7) swing the arm to change tools
8) clamp the tool

It may or may not turn off the coolant. It depends on the MTB. You can actually edit the tool change program. It's a 9000 series program. You have to do something to be able to edit it, I can't remember what. I had to change the one on my Tree long ago.

I'd forget staging the tool until you get it running right.


On these old controls, the control I/O requires a finish signal (FIN) for any command it issues. So, if the machine says "unclamp the tool", it will wait for the FIN signal from the clamping proximity switch before it moves on to the next task. If it does not get the FIN signal, it cannot move on.

I don't know on the MX3, but on some of the old controls, all of the FIN signals share the same input. This is the reason why you cannot program more than one M code on a singe line on the older controls. The machine can only process one FIN signal at a time since all the FIN signals share the same input. If you tried to do two M functions at the same time, the control would not know which function has finished.

On newer machines, this is no longer an issue because they have much more I/O capacity and you can have a unique FIN bit for every function. Consequently, it's much easier to diagnose a newer machine because it can tell you which FIN signal it did not receive. On the old controls, the only way to know what FIN is missing is to know what M code was being issued.

To further complicate things, many times multiple sensor or switch bits are used to make the FIN signal. There could be relays or flip flops to add or or the bits outside the control.

If I remember right, there is a VFD that runs the motor for the swing arm. There is also a servo drive that runs the servo for the carousel. There are several sensors that are used in the tool change process:

Tool unclamped
Swing arm home
Swing arm engaged
Tool pot down
Tool pot up
Carousel in position (from the servo encoder)

I can't remember how that one is set up. The tool changer is the most complicated part of a machining center. As a result, it's the most common area from problems to arise. Lathes are a little easier, but not by much. A side mount tool changer is one of the most complicated types of tool changer.


I'm in Mexico, but I'll be back Sunday if you need my help.
 
You can actually edit the tool change program. It's a 9000 series program. You have to do something to be able to edit it, I can't remember what.

Its a parameter that locks out viewing and / or editing 9,000 group programs. Forget the parameter number off the top of my head, But its called "NE9". There's also an "NE8" for locking 8,000 group programs. In the parameter's they are located close together.

Dead simple to find if you have a parameter manual in PDF. I'm also away from the office or I'd look it up.

Regards Phil.
 
I have a older MX2, 1987 Matsuura with a Yasnac,
my tool change is not easy but easier than your writing
so like this eh?
just out cut(I shut coolant off b4 tc)
T3
M09
G0
G91G28Z0
M06
T7(I stage next tool after TC)
G90G80G40G54G43H3M3S750;back to werk
Gw
 
Thanks a million!

The sensors on the pneumatic arm that slides the pot out of the carousel has led's at both extremes of travel. These are illuminated at the appropriate times.

The swing arm sensor is of the magnetic type and the arm is in the proper position when it halts.

The halt has occurred in two different scenarios:
1. With the pot still in the carousel. (Unsure if it properly indexed.)
2. With the pot extended fully out of the carousel.

So, would this eliminate the clamp/unclamp sensor?

gm
 
Do you have the book that shows the diagnostics? There should be one that shows what it thinks the carousel position is. Indexing the carousel by itself repeatedly while watching that should eliminate the carousel sensors, gear motor and what is driving it as an issue. I think I remember the carousel could index independently of everything and only stopped when waiting for the pneumatic arm.
 
....The sensors on the pneumatic arm that slides the pot out of the carousel has led's at both extremes of travel. These are illuminated at the appropriate times.

Just because the LED on a sensor lights up you can not assume the input signal is being received by the CNC. You should always look at the appropriate diagnostic bits to verify the CNC is really reading the signal.
 
How are you loading the tools into the machine? When it hangs up page thru the POS menu to check what tool it shows is in the spindle and what tool is waiting. If you request a different tool than the one waiting it will either hang up or it is possible to get the tools mixed up, I have seen that happen. I would keep the tool change command on 1 line only like TxxM6. I am not sure if you have a mechanical problem or not, first thing to eliminate is procedure.
 
There are timers involved in the tool change. Even if the prox switch is made, it may not be made at the right time. I'm not sure.
 








 
Back
Top