What's new
What's new

New vf6 has one random peck in a drilling cycle go very slow but the rest are normal.

Dicks4fingers

Aluminum
Joined
Dec 24, 2019
Hoping I'm just being an idiot. But sometimes peck drilling it'll just do one peck in the cycle very very slow even though the current feed rate keeps showing its unchanged. Happens with random values but when it happens it'll do it in the same part over and over.

Here's a drill cycle I got to do it. I hope I'm just being dumb


M08
T6 M06 (0.625 DRILL)
m8
G55 G17 G90
S1833 M03
G00 A90. C90.
M10
M12
G254
T20
X0. Y-2.0849
G43 H6 Z2.25
G94
G98 G83 Z-.4 R1.95 Q0.06 F15.5
G80
G255
G53 Z0.

Also does it with out dwo active. If I play with f and q I can get to happen or not happen.
 

Dicks4fingers

Aluminum
Joined
Dec 24, 2019
Great so new update fixes it and kills dwo. I'll pass. Fucking shit show
 

thesidetalker

Hot Rolled
Joined
Jan 11, 2015
Location
Bay Area, CA
Great so new update fixes it and kills dwo. I'll pass. Fucking shit show
It doesn't necessarily kill DWO.
... probably. Who the hell knows? This is why I'm still using my own DWO macro and not the built in one. Was a bug a few years back where if drilling different depths with canned cycles, the final retract would be wrong. Sometimes close enough not to notice. Sometimes Z+ and sometimes Z-
In my case ruined a nice carbide drill and my collet chuck pushed the part out of the vise.

The problem I found right now is in TCPC. With G234 active, if you make a rapid or feed move, then call a workoffset, the next move could be your last, depending how your code is formatted.

For example:
(G234 is already active)
G0 XYZBC
G54
(this line is potentially crash-tastic)
If you call any other letters besides X, Y, AND Z, (all 3!) the movement is unpredictable. Even calling the exact same position you're already in with only X, only XY, only YZ, etc. or even BC angles, without all three XYZ, you will get unpredictable movements. Sometimes left, sometimes up, sometimes Z-9.

G0 X1 Y1 Z1 B50 C50
G54
X1

Will produce random movement in Y and Z directions. Even though you're already at X1, the machine will move in Y and Z to some unknown values.

Or

G0 X1 Y1 Z1 B50 C50
G54
B50

Will move XYZ some unknown amount.


So... make sure your code doesn't call a duplicate workoffset between movements. (my formatting had other codes between operations, but I've narrowed it down to simply calling the workoffset which causes the problem) Cancelling tool length with G49 and reapplying G234 would work fine. Or just don't call workoffsets while in G234 mode.


Probably my last Haas machine. The bugs are just ridiculous. Every time the control software is updated, are we supposed to prove out every possible operation and combination of operations each time, to be sure there's no other bugs? I just can't trust anything they do after seeing this. The sad part is if you have a problem with something on the machine, the first troubleshooting step for the service guy is a software update. No thanks... who knows what other undiscovered problems you'll find next.
 
Last edited:

13engines

Hot Rolled
Joined
Jun 30, 2015
Location
Saint Paul
Here's a drill cycle I got to do it. I hope I'm just being dumb
......................
T6 M06 (0.625 DRILL)
...........
G43 H6 Z2.25
G94
G98 G83 Z-.4 R1.95 Q0.06 F15.5
G80
I hope this is just some playing around trying to get the machine to act poorly. If not... maybe it's simply the machine's way of complaining that someone just asked it to full retract peck 40 times at less then 1/10th diameter drilling a hole only 3 3/4 diameters deep. I wouldn't want to do that either... under any circumstance! : -) [elbow-elbow]
 

Dicks4fingers

Aluminum
Joined
Dec 24, 2019
......................

I hope this is just some playing around trying to get the machine to act poorly. If not... maybe it's simply the machine's way of complaining that someone just asked it to full retract peck 40 times at less then 1/10th diameter drilling a hole only 3 3/4 diameters deep. I wouldn't want to do that either... under any circumstance! : -) [elbow-elbow]
It was playing around with an angry machine. I'm not the dumb yet. The issue does seem to be a weird peck limit though
 

Dicks4fingers

Aluminum
Joined
Dec 24, 2019
It doesn't necessarily kill DWO.
... probably. Who the hell knows? This is why I'm still using my own DWO macro and not the built in one. Was a bug a few years back where if drilling different depths with canned cycles, the final retract would be wrong. Sometimes close enough not to notice. Sometimes Z+ and sometimes Z-
In my case ruined a nice carbide drill and my collet chuck pushed the part out of the vise.

The problem I found right now is in TCPC. With G234 active, if you make a rapid or feed move, then call a workoffset, the next move could be your last, depending how your code is formatted.

For example:
(G234 is already active)
G0 XYZBC
G54
(this line is potentially crash-tastic)
If you call any other letters besides X, Y, AND Z, (all 3!) the movement is unpredictable. Even calling the exact same position you're already in with only X, only XY, only YZ, etc. or even BC angles, without all three XYZ, you will get unpredictable movements. Sometimes left, sometimes up, sometimes Z-9.

G0 X1 Y1 Z1 B50 C50
G54
X1

Will produce random movement in Y and Z directions. Even though you're already at X1, the machine will move in Y and Z to some unknown values.

Or

G0 X1 Y1 Z1 B50 C50
G54
B50

Will move XYZ some unknown amount.


So... make sure your code doesn't call a duplicate workoffset between movements. (my formatting had other codes between operations, but I've narrowed it down to simply calling the workoffset which causes the problem) Cancelling tool length with G49 and reapplying G234 would work fine. Or just don't call workoffsets while in G234 mode.


Probably my last Haas machine. The bugs are just ridiculous. Every time the control software is updated, are we supposed to prove out every possible operation and combination of operations each time, to be sure there's no other bugs? I just can't trust anything they do after seeing this. The sad part is if you have a problem with something on the machine, the first troubleshooting step for the service guy is a software update. No thanks... who knows what other undiscovered problems you'll find next.
I've had bugs all in their lathe too. The dual spindle lathe wouldn't synce for months.
 

Dicks4fingers

Aluminum
Joined
Dec 24, 2019
It doesn't necessarily kill DWO.
... probably. Who the hell knows? This is why I'm still using my own DWO macro and not the built in one. Was a bug a few years back where if drilling different depths with canned cycles, the final retract would be wrong. Sometimes close enough not to notice. Sometimes Z+ and sometimes Z-
In my case ruined a nice carbide drill and my collet chuck pushed the part out of the vise.

The problem I found right now is in TCPC. With G234 active, if you make a rapid or feed move, then call a workoffset, the next move could be your last, depending how your code is formatted.

For example:
(G234 is already active)
G0 XYZBC
G54
(this line is potentially crash-tastic)
If you call any other letters besides X, Y, AND Z, (all 3!) the movement is unpredictable. Even calling the exact same position you're already in with only X, only XY, only YZ, etc. or even BC angles, without all three XYZ, you will get unpredictable movements. Sometimes left, sometimes up, sometimes Z-9.

G0 X1 Y1 Z1 B50 C50
G54
X1

Will produce random movement in Y and Z directions. Even though you're already at X1, the machine will move in Y and Z to some unknown values.

Or

G0 X1 Y1 Z1 B50 C50
G54
B50

Will move XYZ some unknown amount.


So... make sure your code doesn't call a duplicate workoffset between movements. (my formatting had other codes between operations, but I've narrowed it down to simply calling the workoffset which causes the problem) Cancelling tool length with G49 and reapplying G234 would work fine. Or just don't call workoffsets while in G234 mode.


Probably my last Haas machine. The bugs are just ridiculous. Every time the control software is updated, are we supposed to prove out every possible operation and combination of operations each time, to be sure there's no other bugs? I just can't trust anything they do after seeing this. The sad part is if you have a problem with something on the machine, the first troubleshooting step for the service guy is a software update. No thanks... who knows what other undiscovered problems you'll find next.
Tcpc shiting the bed is how things get expensive fast.
 








 
Top