lathe G76 internal and external threading questions

More
23 May 2019 14:50 #134706 by OT-CNC
I just completed a job running internal and external threads. Linuxcnc performed flawlessly. A big thanks to all the developers and Mesa to make this such a kick ass control system.

I initially tried coding following the G76 outlines in G Codes section and found it a bit confusing.
www.linuxcnc.org/docs/2.5/html/gcode/gco...G76-Threading-Canned

the J and K values seem to throw me off sometimes. I would like to see an actual example with values for quick reference and to make the illustration a bit clearer. Question 1; could we add a sample code to the illustration. Some of us learn better that way.

I also tried fusion's posts for linuxcnc which when set to cycles didn't output correct code. I prefer it in cycles as it's easier to edit to adjust final pitch dia. The post for tormach worked. Both in cycles and caming out all the points. Question 2; is the linuxcnc post processor for lathe being improved/developed for fusion? I have no problem using the tormach one but it would be good to know what works and what doesn't as it eats up time.

Q parameter in G76 cycles question 3; if turning on the back of a part, is this angle set to a neg value? I assume it doesn't matter as the code I ran looked the same with 29.5 and -29.5. I didn't see a tool path graphical difference on the display in axis. I ran it at 29.5. To clarify, this was an unusual approach as I was using an internal threading tool to cut an external thread spinning the spindle backwards and running the tool from the chuck out toward the tail stock.

BTW, I was totally blown away with linuxcnc being able to sync up at that op since there really was not much room to start the thread near a shoulder. I usually like to leave a bit of room for that. That leads me to the last question 4; on 2 occasions in over 100 threading cycles I noticed a pause in the start of threading (on a conventional internal op) what would cause that? To clarify, the threading synced up correctly and cut perfect threads but took a bit longer to sync up. I didn't notice a drop in spindle rpm running at 750. It happened twice in a row then didn't occur again. Computer timing issues?? Linuxcnc going on coffee break? I'm running 5i25 and mesa hardware.

Please Log in or Create an account to join the conversation.

More
23 May 2019 15:42 #134712 by BigJohnT
G76 and other G codes that depend on spindle synchronization will wait at the start for the index pulse from the encoder as well as spindle at speed. Your spindle at speed tolerance may need to be a bit wider as sometimes a VFD will not give you the exact RPM every time.

You should not be looking at out of date documents for older versions of LinuxCNC unless your still running 2.5.
linuxcnc.org/docs/2.7/html/gcode/g-code.html#gcode:g76

There is an example of G76 at the bottom of that section.

JT

Please Log in or Create an account to join the conversation.

More
24 May 2019 15:32 #134818 by OT-CNC
Good point. Which ini values should I look at for the spindle at speed tolerance? Although currently the spindle is capable of indexing and is a bit of a thoroughbred in CSS mode. I'm not noticing any obvious rpm changes during threading set at 750 rpm. This only happened twice say in the middle of a 10 pass threading cycle where it paused briefly twice in a row before picking up again synced.

Please Log in or Create an account to join the conversation.

More
24 May 2019 16:20 #134823 by BigJohnT
It won't be in the ini file it's a hal pin.

motion.spindle−at−speed IN BIT

Motion will pause until this pin is TRUE, under the following conditions: before the first feed move after each spindle start or speed change; before the start of every chain of spindle-synchronized moves; and if in CSS mode, at every rapid->feed transition.

linuxcnc.org/docs/2.7/html/man/man9/motion.9.html

JT

Please Log in or Create an account to join the conversation.

More
28 May 2019 21:26 #135198 by andypugh

on 2 occasions in over 100 threading cycles I noticed a pause in the start of threading (on a conventional internal op) what would cause that?


This could just be luck, depending on whether the index pulse happens just after the system starts to look for it, or fractionally after (in which case there will be a full rotation of paused motion).

Incidentally, I have concluded that it is best to turn off Constant Surface Speed for the duration of threading.

Please Log in or Create an account to join the conversation.

Time to create page: 0.141 seconds
Powered by Kunena Forum