O/D Threading help on HAAS ST-30

[Anonymous]

I've been trying to cut an M12 thread but is't as if it's the wrong pitch?

The O/D is good, the thread depth is good, but it looks/feels like the pitch is slightly off.
It's feeding at 1.75 in the G code but a nut won't go on.

I've tried CSS and fixed, same problem. Canned cycle (G92) and without (G32). (without is a better thread* but still not correct)
When I use the machines G76, it's good.


Could it be related to the way the tool geo is set-up? It's set at 1.75, would the Tip compensation have an effect?




* With the canned cycle, it tends to damage the last-but-one thread slightly, it doesn't happen without it.

[Laurens-3DTechDraw]

KaCad wrote:

I've been trying to cut an M12 thread but is't as if it's the wrong pitch? The O/D is good, the thread depth is good, but it looks/feels like the pitch is slightly off. It's feeding at 1.75 in the G code but a nut won't go on. I've tried CSS and fixed, same problem. Canned cycle (G92) and without (G32). (without is a better thread* but still not correct) When I use the machines G76, it's good. Could it be related to the way the tool geo is set-up? It's set at 1.75, would the Tip compensation have an effect? * With the canned cycle, it tends to damage the last-but-one thread slightly, it doesn't happen without it.

Do you have an infeed angle set?

[Anonymous]

Only thing I can suggest is lay a bolt up next to your threads, and see whats different.

[Anonymous]

Laurens-3DTechDraw wrote:

KaCad wrote: I've been trying to cut an M12 thread but is't as if it's the wrong pitch? The O/D is good, the thread depth is good, but it looks/feels like the pitch is slightly off. It's feeding at 1.75 in the G code but a nut won't go on. I've tried CSS and fixed, same problem. Canned cycle (G92) and without (G32). (without is a better thread* but still not correct) When I use the machines G76, it's good. Could it be related to the way the tool geo is set-up? It's set at 1.75, would the Tip compensation have an effect? * With the canned cycle, it tends to damage the last-but-one thread slightly, it doesn't happen without it. Do you have an infeed angle set?

It's feeding in from 1mm off the job, the angle's set at 0 at the moment.

[Anonymous]

dthuren wrote:
Only thing I can suggest is lay a bolt up next to your threads, and see whats different.

The threads look very close but don't quite mesh. A nut will go on about 1 turn before locking.

[Rob_Lockwood]

KaCad wrote:
When I use the machines G76, it's good.

then perhaps the fix is to have your post processor altered to output threading cycles using g76.

[Anonymous]

Rob Lockwood wrote:
KaCad wrote: When I use the machines G76, it's good. then perhaps the fix is to have your post processor altered to output threading cycles using g76.

I could do that, yes, but I'm just trying to understand why it's doing what it is.



Does the In-feed angle have to be half the insert angle? eg, 30° for a 60° 1.75mm ISO insert?

[Laurens-3DTechDraw]

KaCad wrote:
Rob Lockwood wrote: KaCad wrote: When I use the machines G76, it's good. then perhaps the fix is to have your post processor altered to output threading cycles using g76. I could do that, yes, but I'm just trying to understand why it's doing what it is. Does the In-feed angle have to be half the insert angle? eg, 30° for a 60° 1.75mm ISO insert?

We run the infeed angle usually anywhere between 26 and 30.

[jeff.walters]

KaCad wrote:
dthuren wrote: Only thing I can suggest is lay a bolt up next to your threads, and see whats different. The threads look very close but don't quite mesh. A nut will go on about 1 turn before locking.

I have seen this before where a machine works (reacts) differently when doing synchronized turning (G92, G32, G33) and running a canned cycle (G76, G78). Sometimes it’s a matter of slowing down the spindle to give the machine time to sync the feed better, the other thing is the decimal for the feed may need to be carried out a bit farther to be more “exact”. In a canned cycle (G76) you are giving the control the pitch and then it is doing all the synchronized turning calculations and will adjust the actual feed to what it needs.

[Anonymous · Accepted solution]

Just a follow-up on this.

The thread was off due to my error. ::)


I wasn't giving it enough of a lead to get in sync before cutting. Since increasing that, it's cutting great.  🙂




Thanks all for the help/suggestions.