Hesitation between two consecutive G33 codes
05 Jun 2017 19:27 #94117
by 10K
Hesitation between two consecutive G33 codes was created by 10K
I wrote some code to thread NPT pipe. I used two consecutive G33 commands to cut the effective and the vanish thread. When I run the code on my lathe, I get a slight hesitation between the two tapers. I can see the effect on the cut. If I run about 100 RPM, it's not too bad. More speed makes it worse.
I thought it might be because the threads didn't quite match up, so I made sure that the length of the first cut is an Integer multiple of the thread pitch. This made it better, but didn't completely eliminate it.
Here's the section out of the loop that cuts the tapers:
If LinuxCNC generates the code on the fly, maybe there's some processing time between the commands that results in a small delay?
Anybody have a solution or any ideas?
I thought it might be because the threads didn't quite match up, so I made sure that the length of the first cut is an Integer multiple of the thread pitch. This made it better, but didn't completely eliminate it.
Here's the section out of the loop that cuts the tapers:
G33 X[#<XX> + #<X1>] Z[#<ZZ> - #<Z1>] K#<Pitch> (cut effective thread)
G33 X[#<XX> + #<X1> + #<X2>] Z[#<ZZ> - #<Z1> - #<Z2>] K#<Pitch> (cut vanish thread)
If LinuxCNC generates the code on the fly, maybe there's some processing time between the commands that results in a small delay?
Anybody have a solution or any ideas?
Please Log in or Create an account to join the conversation.
06 Jun 2017 12:51 - 06 Jun 2017 12:53 #94155
by andypugh
Replied by andypugh on topic Hesitation between two consecutive G33 codes
Does G33 wait for an index? I am not sure, but I don't think it does.
However, it might be that there is no blending between two G33 moves, so the Z axis decelerates to a stop then accelerates.
I don't know what a "vanish" thread is.
However, it might be that there is no blending between two G33 moves, so the Z axis decelerates to a stop then accelerates.
I don't know what a "vanish" thread is.
Last edit: 06 Jun 2017 12:53 by andypugh.
Please Log in or Create an account to join the conversation.
06 Jun 2017 13:30 #94160
by PCW
Replied by PCW on topic Hesitation between two consecutive G33 codes
G33 waits for index so will likely pause between passes
but a pause (of any duration) should not matter as both cutting
paths should be correct
What is the actual problem?
but a pause (of any duration) should not matter as both cutting
paths should be correct
What is the actual problem?
Please Log in or Create an account to join the conversation.
06 Jun 2017 13:40 #94161
by 10K
Replied by 10K on topic Hesitation between two consecutive G33 codes
I believe that G33 waits for an index. If I rerun the program, it starts on the same point in the thread.
The names come from Machinery's Handbook. A vanish thread is the second part of the thread that blends the tapered thread to the OD of the pipe. Without it, the thread will stop at a point where it is still deeply cut into the pipe. It's a section of steeper taper, if you will.
The problem is that I get a hesitation in Z movement while the thread is being cut. If it is great enough, the Z pauses for some or most of a spindle revolution, and a non-threaded circumferential cut is made.
The explanation that there is a deceleration and acceleration between the two G33 codes may be the problem.
The names come from Machinery's Handbook. A vanish thread is the second part of the thread that blends the tapered thread to the OD of the pipe. Without it, the thread will stop at a point where it is still deeply cut into the pipe. It's a section of steeper taper, if you will.
The problem is that I get a hesitation in Z movement while the thread is being cut. If it is great enough, the Z pauses for some or most of a spindle revolution, and a non-threaded circumferential cut is made.
The explanation that there is a deceleration and acceleration between the two G33 codes may be the problem.
Please Log in or Create an account to join the conversation.
06 Jun 2017 13:57 - 06 Jun 2017 14:27 #94163
by PCW
Replied by PCW on topic Hesitation between two consecutive G33 codes
If it hesitates during the cut , thats either a G33 bug or perhaps an encoder issue
Edit: Maybe its related to his bug: (blending during G33)
github.com/LinuxCNC/linuxcnc/issues/68
Nots sure if this is fixed or if you could work around it by setting exact stop mode
or disabling the circular arc blend mode (ARC_BLEND_ENABLE = 0 in the ini)
Edit: Maybe its related to his bug: (blending during G33)
github.com/LinuxCNC/linuxcnc/issues/68
Nots sure if this is fixed or if you could work around it by setting exact stop mode
or disabling the circular arc blend mode (ARC_BLEND_ENABLE = 0 in the ini)
Last edit: 06 Jun 2017 14:27 by PCW.
The following user(s) said Thank You: 10K
Please Log in or Create an account to join the conversation.
06 Jun 2017 14:29 - 06 Jun 2017 14:34 #94168
by 10K
Replied by 10K on topic Hesitation between two consecutive G33 codes
Just to be clear, the hesitation is between the two G33 commands, not during a G33 command.
Edit: The link you provided appears to be similar to what I've experienced. Thanks! It'll take me a while to digest / try out the suggestions.
Edit: The link you provided appears to be similar to what I've experienced. Thanks! It'll take me a while to digest / try out the suggestions.
Last edit: 06 Jun 2017 14:34 by 10K.
Please Log in or Create an account to join the conversation.
06 Jun 2017 14:36 #94169
by andypugh
Replied by andypugh on topic Hesitation between two consecutive G33 codes
I wonder if you can use G76 with a combination of infeed and outfeed tapers?
I suspect not.
There is this:
www.wallacecompany.com/tmp/G76/
G76 with a taper option.
Looking at the standards the "vanish threads" are due to the die taper, and have a different (and variable) profile, so can't be single-point cut anyway.
I suspect not.
There is this:
www.wallacecompany.com/tmp/G76/
G76 with a taper option.
Looking at the standards the "vanish threads" are due to the die taper, and have a different (and variable) profile, so can't be single-point cut anyway.
Please Log in or Create an account to join the conversation.
06 Jun 2017 14:42 #94172
by PCW
Replied by PCW on topic Hesitation between two consecutive G33 codes
But since G33 waits for index (at least that's what the docs say) , it will always pause at the beginning
so It may not be possible to use it the way you are trying to use it
(merging 2 G33s instead of a separate pass), that is because of the wait for index sync
you cannot be cutting at the beginning of G33
so It may not be possible to use it the way you are trying to use it
(merging 2 G33s instead of a separate pass), that is because of the wait for index sync
you cannot be cutting at the beginning of G33
Please Log in or Create an account to join the conversation.
06 Jun 2017 15:09 #94175
by 10K
Replied by 10K on topic Hesitation between two consecutive G33 codes
Here's a picture of the threaded pipe:
There's a somewhat puny thread about three threads from the left end. That's the transition between G33 commands. It's not too bad on this sample, since I ran at about 100 RPM. If I run the lathe faster, it's much more pronounced.
I agree that the vanish taper is an artifact of the pipe die, and that this cut is an approximation. However, some transition is needed, or the thread would have to be cut too long. A steeper taper can help make up the pipe joint if the tolerance between the pipe and fitting is a little off.
There's a somewhat puny thread about three threads from the left end. That's the transition between G33 commands. It's not too bad on this sample, since I ran at about 100 RPM. If I run the lathe faster, it's much more pronounced.
I agree that the vanish taper is an artifact of the pipe die, and that this cut is an approximation. However, some transition is needed, or the thread would have to be cut too long. A steeper taper can help make up the pipe joint if the tolerance between the pipe and fitting is a little off.
Please Log in or Create an account to join the conversation.
06 Jun 2017 15:38 #94179
by andypugh
Replied by andypugh on topic Hesitation between two consecutive G33 codes
Is there a problem with just retracting the tool?
Certainly with G76 this works fine and leaves no groove. The thread just ends.
Certainly with G76 this works fine and leaves no groove. The thread just ends.
Please Log in or Create an account to join the conversation.
Time to create page: 0.130 seconds