Probe calibration cycle confusion. G78 bore locate works but G72 calibration doesn't
Adding probing to a sabre 1000.
OMI machine interface and OMP400 probe. Probe is set to OMP700 in acramatic. Otherwise wired 100% to the wiring diagram
G74 height calibration cycle works great
G78 bore locating cycle works great
I can't get G72 to run correctly. It starts but doesn't finish successfully - outcome depends on how I start the cycle.
probe is set as a probe in tools, rough length set and rough tip diameter set (6mm)
two different starting situations with different bad outcomes:
1. The help menu in RAP says to position the probe 'over' the exact center of the bore. Use G78 to center on bore - pull up 1 inch (arbitrary!) - press cycle start. the table will run to the right until the limit switch (basically the entire table) and then return a no-hit. (this seems wrong)
2. operating manual 3.0 pub 91203809--001 PDF says to position probe at the exact center inside the bore. again use G78 to center on bore. G72 cycle start. table runs right to bore edge. Over run the probe and report an unintended hit (40-55) (this setup seems right)
G72 & G78 both call for diameter - I'm using a calibrated ring gauge and using the same diameter for both operations.
What am I doing wrong?!
1 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
G72.pdf Attached. Probe is below top of hole as to contact Diameter. I always use a lever type indicator and tram the stylus to run within .0005" (.013mm)
TIR (Total indicator runout) before running a calibration. Also tram the precision bore to be centered before calibration.
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Thank you Greg!
That confirms the #2 operation is the right one.
Unfortunately I'm still running in to the unexpected hit alarm.
Here is a video
https://youtu.be/C6Uj-UPkx4c
code run - straight out of RAP
Code:
G78 G90 Z0 P4.8001
G72 G90 P4.8001
I'm not sure I can get anymore centered in the bore.
the probe tip is reading ~.0001 TIR, I'll triple check again
my only thought is using the output of the G78 to feed the P value from the system probe value for the G72
1 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
1) Check that the X & Y Probe Offsets in the tool table are set to Zero and that the Nominal Probe Diameter is correct in the Tool Table.
If any changes are made data reset and reactivate the tool in MDI.
2) Use an indicator and tram the bore to read the same at all four positions in X & Y.
3) Re-run just the G72.
If your still getting unexpected hits, the feedrate in the Probe Cycle Parameter table may be set too high.
After the G72 to Calibrate; then you can use the G78 to measure a bore, etc...
3 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Quote:
Originally Posted by
gwood
1) Check that the X & Y Probe Offsets in the tool table are set to Zero and that the Nominal Probe Diameter is correct in the Tool Table.
If any changes are made data reset and reactivate the tool in MDI.
2) Use an indicator and tram the bore to read the same at all four positions in X & Y.
3) Re-run just the G72.
If your still getting unexpected hits, the feedrate in the Probe Cycle Parameter table may be set too high.
After the G72 to Calibrate; then you can use the G78 to measure a bore, etc...
Thank you again Greg!
1. confirmed both values are 0 in the tool table - diameter of probe ball end is confirmed & set
2. done - the G78 bore cycle was within half a thou center - but the measured diameter in PRB_AVG_DIA was small. I realized that was because the aborted/failed G72 was clearing the X_NEG_TIP and related - causing the diameter to be off effectively by the diameter of the probe tip.
3. same outcome :(
Adjusted the cycle probe speeds way, way, down (from default of 60 and 6 to 10 and 2) - no difference.
Tried adjusting the G72 diameter value up and down - behavior is unchanged.
The unexpected hit on G72 seems... different - it registers the hit (red led) and then seems to keeps moving in the same direction for another ~1/8" until the probe is really bent over (as seen in the first vid.) This long over travel still occurs with the greatly reduced probe cycle speed.
Attached a few config screens
1 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Are you sure the tool table is in Metric when you put the 6mm probe diameter in? Normally 4 decimal digits is Metric and 5 is Inch
Possibly a lot of backlash in the Axis. Attached: _Troubleshooting_Axis_Positioning_Problems-VMC-110518GEW.pdf
With the OP700 and OPM400 these are digital probes and you should check with Renishaw in Chicago; but normally a Graphite probe stylus is required.
The ceramic is too heavy. The Calculated probe offsets are about 0.01573 larger on the diameter.
You trammed the bore with an indicator and then ran just the G72 1st?
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Quote:
Originally Posted by
servtech
Clearly G72 is not looking for the trigger point during the calibration cycle. It is just overtravelling.
I would upload a fresh copy of the probe calibration cycle in case of corruption.
That 100% sounds in-line with the observed behavior!
Any pointers on how to do that? Is this a 'reload software' from the acramatic service menu, or something in programware? I'm failing at finding anything in manuals or via search.
1 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Attached: 91203884-010_MP700_SSP_A2100-Rel3.x.pdf
P.S. The Control always works in Metric in the Background (System Registers)
In tool Table you probe Diam is close to 0.23622 Inches
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Reload Fresh Copy of A2100 Base-NCX-MAI software
________________________________________
To Reload the A2100 Base, NCX and MAI software; but NOT WiNNT.
Perform an “Emergency Uninstall” during Boot or E-Stop the machine and select “Uninstall” in the Service screen
Wait until it is on the “Software Uninstall Utility” screen which shows the correct Current software in Blue on the Left.
You will see a COMMIT button Icon below Grayed Out.
Push and hold the “Shift” key on the keyboard and select the “New Install” icon.
Commit icon turns from grey to black; Select “Commit” and OK it will back everything up; then if found will un-compress and re-install
a fresh copy of the Base, NCX and MAI software; then it will prompt:
“Do you have more software to install? Press No to run the System”.
Press “No” It will restore the Machine Data Backup, Tool Tables and Offsets and load with a fresh install of A2100 software.
________________________________________
If the A2100 compressed software is corrupt, it will ask if you want to go back to the current junk...
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Quote:
Originally Posted by
gwood
Attached: 91203884-010_MP700_SSP_A2100-Rel3.x.pdf
P.S. The Control always works in Metric in the Background (System Registers)
In tool Table you probe Diam is close to 0.23622 Inches
super helpful info, thank you!
Quote:
Originally Posted by
gwood
Reload Fresh Copy of A2100 Base-NCX-MAI software
________________________________________
To Reload the A2100 Base, NCX and MAI software; but NOT WiNNT.
Perform an “Emergency Uninstall” during Boot or E-Stop the machine and select “Uninstall” in the Service screen
Wait until it is on the “Software Uninstall Utility” screen which shows the correct Current software in Blue on the Left.
You will see a COMMIT button Icon below Grayed Out.
Push and hold the “Shift” key on the keyboard and select the “New Install” icon.
Commit icon turns from grey to black; Select “Commit” and OK it will back everything up; then if found will un-compress and re-install
a fresh copy of the Base, NCX and MAI software; then it will prompt:
“Do you have more software to install? Press No to run the System”.
Press “No” It will restore the Machine Data Backup, Tool Tables and Offsets and load with a fresh install of A2100 software.
________________________________________
If the A2100 compressed software is corrupt, it will ask if you want to go back to the current junk...
very straight forward and quick process, thank you again!
Unfortunately no change, exact same behavior as before :(
Quote:
Originally Posted by
servtech
I've watched your video and the first pass at S0 runs through fine, the rotation to S90 seems to be the issue. Does it always do that ? If so I would have the probe and tip runout checked in a tool setter. Check ball runout adjust to reduce the tip runout to a minimum.
Very easy to look for the complicated solution when basics haven't been checked ????
I've knocked it out of a few thou a few times from how many times I've run it in to the side of the bore testing. But it goes right back in with a few adjustments and indicates dead zero.
here is a terrible video of the runout
https://youtu.be/mkjSpzgZ33E
still to pull the way covers and check backlash
@gwood Renishaw confirmed your suggestion on the styli - I've got a carbon fiber one in the mail. They said the ceramic will work almost always but sometimes has some issues. They didn't believe the behavior as I described to them sounds like a false or double hit but suggested the CF to confirm.
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
I'm confused on the rotation of the spindle. I should has seen this earlier...
The UK-VMC software did NOT rotate the spindle. Normally was oriented at M19 S0 by software cycles. On the Horizontals it did rotate.
Instead the probe was trammed in to within .0005" runout using very light pressure on a lever style Test Indicator.
Verify the MAI Spindle Probe MP700 settings I attached earlier. During the MP700 calibration it makes a total of 12 hits with "No" rotation of the spindle.
Always run the G72 fist. After being calibrated; the System Parameters for Spindle Probe should show 12 values for 360-Calibration.
After calibration; then just your measuring routines.
Take a photo of left side of Service screen to show the Software "Loaded" Software Options should be YES and enabled for the SSP (Spindle Probe and Cycles).
3 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Quote:
Originally Posted by
gwood
I'm confused on the rotation of the spindle. I should has seen this earlier...
The UK-VMC software did NOT rotate the spindle. Normally was oriented at M19 S0 by software cycles. On the Horizontals it did rotate.
Instead the probe was trammed in to within .0005" runout using very light pressure on a lever style Test Indicator.
Verify the MAI Spindle Probe MP700 settings I attached earlier. During the MP700 calibration it makes a total of 12 hits with "No" rotation of the spindle.
Always run the G72 fist. After being calibrated; the System Parameters for Spindle Probe should show 12 values for 360-Calibration.
After calibration; then just your measuring routines.
Take a photo of left side of Service screen to show the Software "Loaded" Software Options should be YES and enabled for the SSP (Spindle Probe and Cycles).
Thank you again for the thoughtful replies! Apologies for the delayed response, got distracted with some other shop projects while waiting for the CF probe tip.
Super interesting on the rotation!
I went back through my various rambling videos of me trying to figure out what's going on. I *think* the spindle rotation only occurs when the spindle gets rotated out of the start/fixed position and it's just a return to configured orientation. It does not occur every time I run a G72.
I can never get the probe calibration values to display anything but the 4 values as shown in my previous calibration value screenshot. x pos/neg & y pos/neg
Attached are screenshots of the service screen & license screen as well as the two relevant config value screens.
I received the 50mm CF probe tip - same result :(
I've run through the mp700 calibration doc you had previously sent again and everything looks to match up (there doesn't seem to be much to config there beyond the 360 deg type checkbox)
Aside from checking backlash I think my next best straight forward test is to try and switch to a non OMP700 probe in settings and see if behavior is any different.
I'm going to put together a video of the whole bore centering and g72 operation from start to finish to see if anything else stands out that might be wrong that isn't captured in text or short videos.
Anything else that comes to mind? I keep scratching my head on this whole process...
2 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Your Machine application Configs are not even close to what I sent earlier.
You've got the wrong port wired, Probe Contact incorrect.
Make sure that the time out in "Axes and Spindle Setup" is set to 60 seconds.
The probe internal settings may be wrong as well.
I've attached a single page notes that also has some register values.
Hopefully you have wiring diagrams that match the 9120xxxxy on the serial plate.
Your Software is correct!
Hopefully this will help.
P.S. Make sure that when the tool is loaded from the drum as a tool type PROBE that if the stylus is deflected you get an unexpected hit.
3 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Quote:
Originally Posted by
gwood
Your Machine application Configs are not even close to what I sent earlier.
You've got the wrong port wired, Probe Contact incorrect.
Make sure that the time out in "Axes and Spindle Setup" is set to 60 seconds.
The probe internal settings may be wrong as well.
I've attached a single page notes that also has some register values.
Hopefully you have wiring diagrams that match the 9120xxxxy on the serial plate.
Your Software is correct!
Hopefully this will help.
P.S. Make sure that when the tool is loaded from the drum as a tool type PROBE that if the stylus is deflected you get an unexpected hit.
probe contacts & wiring - you're very much on to what is likely the issue, thank you again!
You're right, things don't match up, I fiddled with normally open/normally closed/MI8-4 and OMM configs until everything more or less was doing what it was supposed to. I figured the differences were caused by me using am OMM instead of the older OMI & the MI8-4 instead of the MI8 with subtly different wiring and configs. I've got the wiring on two pins flipped around for sure. The probe grounds & probe 0 pins
on P66 going in to the probe connector on the acramatic controller - I flipped pins 1 & 3. This is absolutely causing 75+% of the different normally open/normally closed differences in configs. I broke my pin removal tool so I'll get that swapped this week and adjust everything to 100% match your configs.
https://www.cnczone.com/forums/attac...1&d=1678069929
First off I triple checked wiring diagrams match up - they sure look like they do... physical copies and a PDF came with the machine. I believe these were re-orders from ~2017
https://www.cnczone.com/forums/attac...1&d=1678069534
I'm coming up with a difference between the wiring diagram and your one pager - I try to read the wiring diagram any other way but I keep coming back to tool setter probe 0 input, spindle probe probe 1 input. Does it matter beyond the configuration of one or the other in the config values? I can certainly swap the pins around but I wanted to confirm which is the intended/'right' - spindle probe, tool probe 1, or vice versa?
https://www.cnczone.com/forums/attac...1&d=1678069534
4 Attachment(s)
Re: Probe calibration cycle confusion. G78 bore locate works but G72 calibration does
Various wiring diagrams attached relative to Rel 2.3x & 3.2 all are the same Physical Wiring. One shows the switch setting for OMI-1 & MP700 as well.
Also the Configs for Rel 2.3x with latest updates Configs all show Software ports 0 for Spindle and 1 for tool Probe.
As long as the Spindle Probe gets loaded by M6 with tool type set to "Probe"; then if stylus is push with string or by hand with the door sw closed;
then you should get an "Unexpected Hit" Alarm.
