On 1/26/26 11:00, Jon Elson wrote:
On 1/26/26 06:42, gene heskett wrote:
A higher resolution encoder? What I have is probably overkill so let
me describe it.
The A/B signal portion of that encoder is a 1000 ppr omron, driven
from the rear of the motor, from an extension shaft I made.
Which has a variable scale in the .ini file depending on the gear the
go704 is in, detected & controlled by tally switches on the gear
shift knob.
OK that gearing MAY be the issue, or possibly Z axis backlash. If the
encoder senses motor movement before the spindle actually reverses as
the gear backlash is taken up, then the tap will start the Z
retraction before the tap rotation reverses. If the issue is Z axis
backlash, then the spindle reverse will get ahead of the Z motion.
There is something else that may affect this. The spindles direction
change is actually programmed in the hal file and is effective in the
reversals involved to any direction change, The sequence is as follows
in the .hal file:
1:When motion changes the direction, it is blocked by an xor2 and
initiates a slowdown in a limit3 while watching the encoder signal. The
dir signal applied to the pwm-servo is frozen in the previous state
2: this condition remains until a 50 millisecond timer triggered by the
encoder finally expires, indicating the spindle has essentially stopped.
2a: since your pwm-servo is a full 4 quadrant controller, the motors psu
filters are by then up to around 170 volts, 40 volts above their
rating. Its saved energy. The short pulse of overvoltage serves to
force a bit of leakage during this time, maintaining the chemical
formation of those caps at or possibly above their labeled voltage. Good
for the caps life IOW. They've now been abused like that for over a
decade, no failures.
3: then when the oneshot finally times out, the direction signal is
finally allowed to get to the dir input of the spindles pwm-servo amp.
4: at that same time, the commanded speed is reapplied to the limit3
input by a mux4, which then uses up that overcharge in the psu caps to
re-accelerate the motor to the requested speed.
5: no wall power is used during this reversal until the cap voltage is
back down to its normal 126 or so as the motor is spun back up by the
stored energy from 2a. This reversal is complete at tapping rpms, say
700 revs in low gear, in about 400 milliseconds.
That leaves a question: When does motion tell the TP the direction has
changed??. It should be fine if it waits to start the backout until the
encoder's out direction is updated, which will happen in the first
degree of a new direction or does it start the backout motion when it
has issued the direction change?? If the latter, there obviously is my
problem.
Andy suggests the necessary test to check Z backlash.
barring stiction on the post, might be a thou. I can usually see or hear
stiction, prompting me to squirt a fresh dose of vactra on the post.
Might be 4 thou if stiction has manifested, All 3 axises are 20mm ball
screws. So I've not yet read Andy's suggestion.
Jon
Thanks Jon.
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers
Cheers, Gene Heskett, CET.
--
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.
- Louis D. Brandeis
Don't poison our oceans, interdict drugs at the src.
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers
