Mits Not Following Directions

This is beyond bazaar...


1990 Tornos Top 200 with Mits 3 path control.

[background]
A few days ago I had a 5/8 bar of 4140 get bent at the back of the spindle after it came out of the bar feeder as the feeder collet did not have a hold on it like it should have. The bent bar took out the main spindle tach. I have replaced it, and all seemed fine, except that G28 C0 does not ever find home now. There is a secondary encoder on the same [belt driven] shaft as the tach, so I'm guessing that is the C axis encoder, and maybe the HOME pulse should be coming from that, and my EE was s'posed to come over and check to see if he could find a pulse fro the encoder - to see if it is the encoder that is damaged, or if somehow we buggard something in the control during the crash. But we are not dooing any milling on this job, so that is to happen as soon as I get this job off the machine.
[/background]


So, once I got it back running after puting the new (stock replacement?) tacho on the machine, I let it run.

I am pretty darn sure that I checked the first/next part off and all was fine.
It appears that either I did not check the last part from that bar when I fed it, and ran the whole next bar.
When I fed it the next time - I noticed a buggard up back end.
I chased it back up stream and it appears that it started mid way (maybe 2/3?) the way through that first bar.

You should be able to see in the pic - the end feature on the RH side is s'posed to be the same on the left, but the left (cut-off) end is all out of position. Well, not all.... I have a 35* tool coming in and breaking it down (says triangle in the code) and that is correct, and editable. However - at first my snap ring groove and the cut-off were +.100" in Z.

You see a big flange like feature that this has left on the part, and so finally I edited the 35* breakdown tool to go another .120" further to clean that off, and then my part got roughly another .120" longer yet!

So then it seemed like it was following where I left the 35* tool prior to changing tools, so I tried moving the 35* tool back .210" before changing tools, and seemed to get no change. So then I just stopped the 35* tool at that point, rather than going further and coming back, and now you can see the last part on the top of the pile is now .500" short! The breakdown is now on the front of the next part (still on the bar) and the groove and cut-off have moved .500".

I have killed all power to the machine and left set 1/2 hour (?) and rebooted. No change.
I have dumped the prog and reloaded it. (same file tho) No change.

I have edited the groove tool in N4 to T913 and used a different Z offset - no change.

I have went back to T909 and then edited the Z offset .100 in the offset reg and only the groove on the front end of the part moved.

I was able to move the T10 cut-off around, BUT the offset is .100 off of nominal.

It is like - even tho I edit anything after the 35* breakdown tool (triangle) nothing changes.
However - I can seemingly influence it by what I doo with the breakdown tool.


The short part is the last thing that I have done with it as of this posting.
To clarify the pic, the short one is the second down from the top, and the top 2 parts are a second tier on the stack - if that helps you see the pic better....

I have not tried setting up another job at this point yet.





I am only running path 1 right now as the rest of the program is only to run the sub-spindle for pick-off porpoises, and the sub spindle is currently not opperable. I will strip that code to lessen cornfuzsion, but that will at least explain the wait codes.

Has anyone ever experienced anything like this on an old Mits?



O250 (00-00 -0000 00:00)
$1
(CTG 3 AXLE)
G20
G94
G54(Z6.)
M99
G0 X17. T0
SYNC3(1)
N1 T101(TRIGON)
M3 S0=2000
G0 X.7 Z0
G1 X-.05 F5. M8
G0 X.44
G1 X.5 Z-.03 F5.
Z-.19 F20.
G0 T0 X17. M9
N2 T909(.047 GRV)
G0 X.7 Z-.19
G1 X.44 F1. M8
G0 X.7
Z-.179
G1 X.44 F1.
Z-.19
G0 T0 X17. M9
N3 T101(TRIGON)
G0 X.7 Z-.185
G1 X.55 F5. M8
X.575 Z-.25
Z-3. F20.
SYNC3(2)
Z-5.66
G0 T0 X17. M9
/
SYNC3(3)
N1 T202(TRIANGLE)
G0 X.7 Z-5.53
G1 X.5 F2. M8
Z-5.67 F15. M8
G0 X1. M9
G0 X17. T0
SYNC3(4)
N4 T909(.047 GRV)
G0 X.7 Z-5.415
G1 X.443 F1. M8
G0 X.7
Z-5.426
G1 X.44 F1.
Z-5.415
G0 T0 X17. M9
N5 T1010(CUTOFF)
G0 X.7 Z-5.685
G1 X.475 F2. M8
U.1 F10.
W.05
U-.1 W-.05 F1.
X.2 F2.
SYNC3(6)
X-.1
G0 T0 X17. M9
M5
SYNC3(7)
SYNC3(8)
N6 T707(STOP)
G0 X.2 Z.06
M11
G4 X1.
M10.
Z.1
T0 X17.
M20
M920
M930
M2
$2


----------------------

Think Snow Eh!
Ox

Closure:


Well, I finally found my problem.....

Between the breakdown tool and the last call for the groove tool, the bar needed to be pulled up into the collet a bit.
... apparently .100" !!!

IDK if it's that the tool just started making enough burr, or that I had to change the collet after the prior crack-up, but the part was pulling .100" out of the collet during this Z+ movement. (remember - Swiss - guide bushing)

A) This machine gets a poor grip on material as it's spring load / air release.

B) I am running 4140


On low carbon, any burr there usually just gets rolled back down from a Z+ movement, but the 4140 burr must be enough to pull it out of the collet.


---------------------

Think Snow Eh!
Ox