Lathe Threading - Spindle & Axis Sync (Resolved)

[hebs]

Fresh report attached. I've just noticed that I'm still running a BETA version on lathe! I will update that first thing in the morning and have another go.

[martyscncgarage]

Is your ball screw pitch correct? If its a 5TPI ball screw, then should it not be 5.08mm/rev?
Clarify your ball screw pitch.
I see no noise in your Error Log file

Axis Label Motor Encoder Lash Comp. Limit Home Dir Screw
mm/rev counts/rev. (mm) - + - + Rev Comp
1 Z 5.0000 6400 0.3070 0 0 0 0 N N
2 X 5.0000 6400 0.1060 0 0 0 0 Y N

[hebs]

Yes, 5mm pitch on each axis. Dimensions of parts produced are all spot on, also use the same settings when in mill mode. Steps per rev is correct. Backlash is more than I like but accurate.

[martyscncgarage]

Edited:
By chance have you gone into the PID screen and watched the spindle encoder count up and down and the asterisk flash once per revolution at the same count? Which I realize is hard to do. (Looks like you got a 1024 line encoder, so every 4096 counts from the index pulse, should be another index pulse, asterisk)

Have you tried to program the thread in Intercon and test?
Hopefully Tom or Marc will have some more ideas for you.

[cncsnw]

Yes, it is worth watching the PID screen to see if the index pulse comes around once, and only once, per revolution. Also verify that you get 4096 counts per revolution (since according to your Report, and the encoder link you posted last week, it is a 1024-line encoder).

Right now, though, I suspect it is your Z axis that is losing steps. That is based on your report that each successive thread pass is shifted slightly in Z, by similar amounts each time. If you were getting noise on the index pulse line, then each successive thread pass would start at a completely random place.

If you carefully set Z0 with your tool on the end of the stock; then run the offending threading program; then jog the same tool back to touch the end of the same piece of stock, does it still show Z0?

Does it make a difference in the threads if you extend your lead-in (starting Z before the G92 or G32) to 2mm or more?

Does it make a difference in the threads if you then increase Parameter 240 from 1mm to 2mm?

[hebs]

Thank you gents for the suggestions, because of your recommendations, I believe that I have identified the issue:

Having checked the PID screen, the pulse is good at every 4096 counts, however, having looked closely, it wasn't spot on with a single revolution (by eye it seemed to be ok but when using a fine marker pen on the spindle it was slightly off). I was getting apx 4030 counts per rev! It turns out that one of the 60T pulleys that I have, has 61 teeth!!! I can't believe it and have counted both pulleys many times, but sure enough, one of them has 61T!! outrageous! Doing the maths on the count error also checks out to a 60:61 ratio.

I have ordered 2 new pulleys, this time form the same supplier, so even if they are both 61 it won't be a problem.

Having been through 4 encoders now, I really think this has been the issue all along.

I will report back once I have installed the new pulleys.

Thanks again.

[martyscncgarage]

Thank you gents for the suggestions, because of your recommendations, I believe that I have identified the issue:

Having checked the PID screen, the pulse is good at every 4096 counts, however, having looked closely, it wasn't spot on with a single revolution (by eye it seemed to be ok but when using a fine marker pen on the spindle it was slightly off). I was getting apx 4030 counts per rev! It turns out that one of the 60T pulleys that I have, has 61 teeth!!! I can't believe it and have counted both pulleys many times, but sure enough, one of them has 61T!! outrageous! Doing the maths on the count error also checks out to a 60:61 ratio.

I have ordered 2 new pulleys, this time form the same supplier, so even if they are both 61 it won't be a problem.

Having been through 4 encoders now, I really think this has been the issue all along.

I will report back once I have installed the new pulleys.

Thanks again.

THAT would do it! Good catch!
And another one for us to remember!


Marty

[hebs]

Thank you all! I can confirm that all is working correctly now. It was down to that pesky extra tooth that sneaked in on one of my pulleys. From now on, I’m counting every tooth on every pulley that I use!

(If anyone in the UK is looking for an encoder, I have 3 lying around that I don’t need)

[martyscncgarage]

Glad you found the problem and its fixed. That was a tough one.