Page 1 of 2
Hickory - Slave axis moves independently (BUG?)
Posted: Tue Aug 19, 2025 7:51 pm
by spumco
Just discovered what may be a potentially serious issue.
Trying to align/square the gantry (paired Y/W) and I accidentally had the WMPG knob on '4th' axis. Y and W are paired (P64=2).
I moved the handwheel a few clicks at 10x (0.001") and only the W servo moved,
despite being paired.
With the MPG set on "Y" jog both servos move the same as expected. But if 4th is selected only the W moves.
The discrepancy is also visible on the DRO if W is displayed.
Further testing using MDI revealed that I could command a W-axis move
independently of the Y-axis. G1 W0.0001 F20 resulted in the W servo ONLY moving, and the DRO showed the same discrepancy.
Note the DRO in the screenshot, as well as the last MDI command. This isn't noise or a bad drive setting; I've got EL8's with identical parameters and the W-axis is moving exactly as commanded.
I've no clue if this is a Hickory-only thing, but if it was on Acorn/6 I'd expect someone would have racked the heck out of a gantry by now.
Can someone please check this and see if I've got a setting off, or if this is a bug?
Re: Hickory - Slave axis moves independently (BUG?)
Posted: Tue Aug 19, 2025 8:32 pm
by cnckeith
spumco wrote: ↑Tue Aug 19, 2025 7:51 pm
Just discovered what may be a potentially serious issue.
Trying to align/square the gantry (paired Y/W) and I accidentally had the WMPG knob on '4th' axis. Y and W are paired (P64=2).
I moved the handwheel a few clicks at 10x (0.001") and only the W servo moved,
despite being paired.
With the MPG set on "Y" jog both servos move the same as expected. But if 4th is selected only the W moves.
The discrepancy is also visible on the DRO if W is displayed.
Further testing using MDI revealed that I could command a W-axis move
independently of the Y-axis. G1 W0.0001 F20 resulted in the W servo ONLY moving, and the DRO showed the same discrepancy.
Note the DRO in the screenshot, as well as the last MDI command. This isn't noise or a bad drive setting; I've got EL8's with identical parameters and the W-axis is moving exactly as commanded.
I've no clue if this is a Hickory-only thing, but if it was on Acorn/6 I'd expect someone would have racked the heck out of a gantry by now.
Screenshot 2025-08-19 193556.png
Can someone please check this and see if I've got a setting off, or if this is a bug?
there are massive improvements to axis pairing and auto squaring in v5.4 would you like to become a beta tester? if so, dm me
Re: Hickory - Slave axis moves independently (BUG?)
Posted: Tue Aug 19, 2025 8:33 pm
by cnckeith
spumco wrote: ↑Tue Aug 19, 2025 7:51 pm
Just discovered what may be a potentially serious issue.
Trying to align/square the gantry (paired Y/W) and I accidentally had the WMPG knob on '4th' axis. Y and W are paired (P64=2).
I moved the handwheel a few clicks at 10x (0.001") and only the W servo moved,
despite being paired.
With the MPG set on "Y" jog both servos move the same as expected. But if 4th is selected only the W moves.
The discrepancy is also visible on the DRO if W is displayed.
Further testing using MDI revealed that I could command a W-axis move
independently of the Y-axis. G1 W0.0001 F20 resulted in the W servo ONLY moving, and the DRO showed the same discrepancy.
Note the DRO in the screenshot, as well as the last MDI command. This isn't noise or a bad drive setting; I've got EL8's with identical parameters and the W-axis is moving exactly as commanded.
I've no clue if this is a Hickory-only thing, but if it was on Acorn/6 I'd expect someone would have racked the heck out of a gantry by now.
Screenshot 2025-08-19 193556.png
Can someone please check this and see if I've got a setting off, or if this is a bug?
if you have the axis labeled as W its gonna move if you command it. slave motors don't have axis labels.
Re: Hickory - Slave axis moves independently (BUG?)
Posted: Tue Aug 19, 2025 8:53 pm
by spumco
cnckeith wrote: ↑Tue Aug 19, 2025 8:33 pm
if you have the axis labeled as W its gonna move if you command it. slave motors don't have axis labels.
It is labeled as W. Slaves should be labeled as "N"?
EDIT:
I knew I wasn't crazy. Hickory installation manual Rev5, page 42
Which begs the question, how does one do an M26/W if there's no axis label?
Re: Hickory - Slave axis moves independently (BUG?)
Posted: Tue Aug 19, 2025 9:35 pm
by spumco
So changing the slave axis from W to N didn't work as hoped.
The W axis disappeared from the DRO, unsurprisingly. And jogging the master Y axis worked.
But on power down and restart I got the following errors:
And can't do anything but estop and power down.
After reassigning the axis label as W, I'm now getting error 9013 W-axis communication fault when repowering the machine (all power off).
Now I'm stuck. If the axis is labeled "N", I can start but not get through homing/resynching. If it's labeled "W", I get communication faults.
Re: Hickory - Slave axis moves independently (BUG?)
Posted: Wed Aug 20, 2025 9:37 am
by centroid467
Keith was speaking in too general of terms. Pairing on Hickory prior to V5.40 is different from our other controllers. You do need to give it an axis label.
For some reason, a concise 1 page summary of pairing setup and operation was omitted from newer versions of the Hickory drive TBs. I'll reproduce the page here in full.
Step 9 under 12.C explains how to hide the W axis. This does not prevent it from being commanded to move but makes it less likely that someone will try.
Code: Select all
12) Axis Pairing Setup
A) Basic requirements
1. Basic machine setup completed
▪ Home positions set
▪ Absolute encoder setup
2. Both master and slave of paired axis must have an Axis Label assigned
▪ Example: Master is Y and slave is W
B) Procedure overview
1. Determine adjustment distance required to align/square the paired axes
2. Move slave axis that amount to align the slave axis with the master axis
3. Verify alignment
4. Move the paired axis to the home position
5. Unpair the axis
6. Rehome both the master and slave at this position
7. Re-pair the master and slave axis
8. Power cycle the control.
C) Step by step setup procedure
1. Measure by how much the paired axis is out of square and in which direction.
Example: Y axis gantry is out of square by 0.1” in positive Y direction.
2. Command only the slave axis to slowly move that distance and in that direction.
Example: Press F3 MDI and execute G1 G91 W0.1 F20
3. Re-measure and repeat steps 1 and 2 until gantry is squared. Avoid pressing E-Stop as this will prompt CNC12 to move the axes to their previous alignment and lose the corrections made in steps 1 and 2.
4. Move paired axis to previously set home position.
Example: In MDI, execute G53 Y0 L20
5. Disable the axis pairing. This can be accomplished by setting parameter 64 to 0 in the parameters menu or by executing G10 P64 R0 in MDI.
6. Set the new home position for both the master and slave axis.
Example: In MDI, execute M26 /Y /W
7. Confirm that DRO shows both Y and W axes set to 0. Go to the parameters menu and enable axis pairing by setting parameter 64 back to the appropriate value. See page 5 for a table of settings.
Example: 4th axis paired to 2nd axis requires parameter 64 = 2
8. Shut down CNC12 and power cycle the system. Proceed with normal startup procedures. The squared position will be remembered as long as the absolute encoder position is retained.
9. Recommended: Set appropriate axis properties parameter to hide slaved axis. These are parameters 91-94 and 166-169. Add 512 to the value of the corresponding parameter.
Example: 4th axis is parameter 94. Set value to current value + 512.
D) Day to day operation
1. Power up the CNC control and press cycle start to execute the absolute encoder homing file.
2. Press and release E-Stop to synchronize the paired axes.
▪ Re-sync is automatic after E-Stop release for alignment corrections under 1/64” (0.4mm).
▪ Corrections greater than 1/64” (0.4mm) will prompt for cycle start to initiate re-sync.
3. The machine tool is squared and ready to work.
Re: Hickory - Slave axis moves independently (BUG?)
Posted: Wed Aug 20, 2025 10:19 am
by spumco
centroid467 wrote: ↑Wed Aug 20, 2025 9:37 am
Step 9 under 12.C explains how to hide the W axis. This does not prevent it from being commanded to move but makes it less likely that someone will try.
Does the ability to independently command a slave axis change in 5.40? If not, at the very least I'll have to remove the MPG from the machine as it's entirely possible an operator would put the axis selector on '4th' and rack the machine.
Unless there's a way to edit the WMPG files/settings/black-box to eliminate (or restrict) the 4th/5th/6th axis selector positions?
More importantly, how do I get my machine working now?
If I name the slave "W" (as before) I get the drive communication fault (9013) - seems like CNC12 lost the Ethercat address (or similar) when the axis label changed from W to N and back to W.
Re: Hickory - Slave axis moves independently (BUG?)
Posted: Wed Aug 20, 2025 11:16 am
by centroid467
spumco wrote: ↑Wed Aug 20, 2025 10:19 am
centroid467 wrote: ↑Wed Aug 20, 2025 9:37 am
Step 9 under 12.C explains how to hide the W axis. This does not prevent it from being commanded to move but makes it less likely that someone will try.
Does the ability to independently command a slave axis change in 5.40? If not, at the very least I'll have to remove the MPG from the machine as it's entirely possible an operator would put the axis selector on '4th' and rack the machine.
Unless there's a way to edit the WMPG files/settings/black-box to eliminate (or restrict) the 4th/5th/6th axis selector positions?
More importantly, how do I get my machine working now?
If I name the slave "W" (as before) I get the drive communication fault (9013) - seems like CNC12 lost the Ethercat address (or similar) when the axis label changed from W to N and back to W.
Yes, the pairing rework makes it so that slave axes cannot be independently commanded to move while paired. Only when pairing is disabled is it possible to command motion.
I'm not aware of any facility lock out the MPG axes beyond editing the PLC program. The axis selection logic in the "WirelessMpgStage" could be modified to keep the last valid axis (presumably Z) selected if the knob is rotated to the W position or beyond.
Does the 9013 W AXIS IN FAULT error persist after a power cycle of the Hickory and drives? I don't typically expect to see issues like this when changing axis labels. You did not modify parameters 303 or 311 when changing the label, correct?
Re: Hickory - Slave axis moves independently (BUG?)
Posted: Wed Aug 20, 2025 11:41 am
by spumco
centroid467 wrote: ↑Wed Aug 20, 2025 11:16 am
Yes, the pairing rework makes it so that slave axes cannot be independently commanded to move while paired. Only when pairing is disabled is it possible to command motion.
That's exceedingly good news.
centroid467 wrote: ↑Wed Aug 20, 2025 11:16 am
I'm not aware of any facility lock out the MPG axes beyond editing the PLC program. The axis selection logic in the "WirelessMpgStage" could be modified to keep the last valid axis (presumably Z) selected if the knob is rotated to the W position or beyond.
I just found thid:
IF ExtUsb_MPGAxis4Select_SV && SV_AXIS_VALID_4 THEN ExtUsb_MPGAxisSelect_W = 4, (ExtUsb_MPGAxis4SelectOut_SV)
I'll try commenting it out and compiling once I get the axis working again.
centroid467 wrote: ↑Wed Aug 20, 2025 11:16 am
Does the 9013 W AXIS IN FAULT error persist after a power cycle of the Hickory and drives? I don't typically expect to see issues like this when changing axis labels. You did not modify parameters 303 or 311 when changing the label, correct?
9013 persists after power cycling drives and Hickory (multiple attempts).
Did not knowingly modify 303 or 311, but I forgot to pull a report last night before shutting everything down for the night so I can't be 100% certain that something didn't modify it. I will be able to verify when I get back to it late afternoon.
Re: Hickory - Slave axis moves independently (BUG?) SOLVED
Posted: Wed Aug 20, 2025 9:04 pm
by spumco
Backed everything up, reinstalled CNC12, and went through the process of copying over all the params and wizard settings.
W-axis is now working again, so we can call this one closed.
I'll follow the instructions on leaving the slave axis labeled W, and hiding it on the DRO until 5.40 comes out.
Before and after reports attached in case anyone can find something useful in determining what happened. Before was taken immediately prior to greasing CNC12, so if there's a bad setting it'd be in there.