G32 G92 DIFFERENCE

[Species8472]

Leaving the formatting aside. There seems to be a difference in the machine's response to the "X" exit move. This can be best seen when making a blunt start thread (higbee). When using a G32 the to do the higbee the feather edge is removed but it makes a long ramp out of the first thread instead of a blunt start. When using a G92 cycle for the higbee the machine seems to get out of the way a lot faster on the the "X" re-track move so you get nice blunt start. It looks a lot the higbee on a thread gauge. What's going on in the controller to make this difference.

[g-codeguy]

I'm glad you brought up this question. We never machine higbee threads, but I tried it once to satisfy my curiosity. It pulled out way too slow. I used the G32 command. Nice to know that it will pull out faster with a G92...no matter the reason. I thought it was just our machine. Will have to give it another shot with the G92.

[ProProcess]

Originally Posted by Species8472 Leaving the formatting aside. There seems to be a difference in the machine's response to the "X" exit move. This can be best seen when making a blunt start thread (higbee). When using a G32 the to do the higbee the feather edge is removed but it makes a long ramp out of the first thread instead of a blunt start. When using a G92 cycle for the higbee the machine seems to get out of the way a lot faster on the the "X" re-track move so you get nice blunt start. It looks a lot the higbee on a thread gauge. What's going on in the controller to make this difference.

G32 is a thread cutting.
G92 is a threading cycle.


G32 make the machine behave a certain way (like G1 or G2 does).

When in G32...

• It ingores the Feedrate Override Control, performs at 100%
• It ingnores Single Block
• It looks at a "Dog" on the spindle and coordinates movement with rotation.

It's up to the programmer to control all it's movements...

G0 X1. Z.1
G32X.9F.0625
Z-1.
X1.
G0Z.1

G32 X.8 F.0625
Z-1.
X1.
G0Z.1

(SO ON AND SO FORTH)

The G92 cycle is a macro that performs thread cutting and looks to a machine paramater on how to behave at thread pullout. Too, it most likely uses G32 internally.

G0 X1. Z.1
G92X.1Z-1.F.0625
X.8
X.7
X.6
X.5
G0X1.

So depending on your paramater settings that effect G92 pull out and how you prgrammed the exit using G32, you can get a variety of different results.

Hope this helps.

[Daleb]

We programmed with G32 in the past, but have switched over to G92 because the quality of the last thread is so much better with the pull out faster.

[Species8472]

Originally Posted by ProProcess G32 is a thread cutting. G92 is a threading cycle. G32 make the machine behave a certain way (like G1 or G2 does). When in G32... It ingores the Feedrate Override Control, performs at 100% It ingnores Single Block It looks at a "Dog" on the spindle and coordinates movement with rotation. It's up to the programmer to control all it's movements... G0 X1. Z.1 G32X.9F.0625 Z-1. X1. G0Z.1 G32 X.8 F.0625 Z-1. X1. G0Z.1 (SO ON AND SO FORTH) The G92 cycle is a macro that performs thread cutting and looks to a machine paramater on how to behave at thread pullout. Too, it most likely uses G32 internally. G0 X1. Z.1 G92X.1Z-1.F.0625 X.8 X.7 X.6 X.5 G0X1. So depending on your paramater settings that effect G92 pull out and how you prgrammed the exit using G32, you can get a variety of different results. Hope this helps.

N105G0G99X6.8122Z.1
N106M8
N107Z-.4967
N108X6.5773
N109G32Z-3.0667F.125
N110G0X6.8122
N111Z-.5012
N112X6.5612
N113G32Z-3.0712F.125
N114G0X6.8122
N115Z-.4977
N116X6.5488
N117G32Z-3.0677F.125
N118G0X6.8122
N119Z-.5007
N120X6.5383

The format I'm using has a "G0" on the "X" re-track lines and it still leaves a ramp. In your code you don't issue a rapid on the re-track. If the G32 is modal then it should be re-tracking at the pitch feed rate. Or not? Then at .125 per rev it should be a little over 1/2 turn to clear with the thread depth at around .078. Looking at the ramp of the G32 higbee it seems that my G0 re-track isn't moving at the rapid feed rate.

[ProProcess]

Originally Posted by Species8472 ... The format I'm using has a "G0" on the "X" re-track lines and it still leaves a ramp. In your code you don't issue a rapid on the re-track. If the G32 is modal then it should be re-tracking at the pitch feed rate. Or not? Then at .125 per rev it should be a little over 1/2 turn to clear with the thread depth at around .078. Looking at the ramp of the G32 higbee it seems that my G0 re-track isn't moving at the rapid feed rate.

G32 is in deed modal (Group 1).
So, switching modes may be causing some issues with the pullout.
I'm under the impression that most threading cycles do remain in the threading mode as they are pulling out.
This will assure that the move exiting the thread is still "timed".

[sinha_nsit]

Straight pullout with G32 will always spoil the last thread.
G92 is a better choice. G32 is mainly useful for tapping with floating holder.

[Species8472]

Did some digging. My Mori NL2500Y lathe has a parameter (1182) that controls the re-track feed rate on non-chamfering threading. It was set to 0 this sets the re-track feed to the thread pitch. At that setting I get 122° of ramp. Set at 30000mm per minute it's about 89°. So at 300rpm at .125 per. rev. the threading feed rate is 37.5ipm with a thread height of .07785 the ramp should be 224°. At 1181ipm it would be about 7°. I would expect it to be bit longer because of the servo's acceleration ramp. 10x longer seems like a lot. I'm still missing something.