Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: c30232 on December 31, 2014, 02:44:44 AM
-
When homing X with the home-off set to 0.800" the home-off sets to approximately 7.995". Set at .010" the result is approximately 0.070" Any ideas why? Additionally, homing triggers a reset. The only other axis, A, works fine. I swapped limit switches with the same result.
-
If the limit is triggering when homing then it is almost certainly bounce.
You could try setting a debounce interval (in General config) of 2000.
Hood
-
Thanks,
It seems to be unique to just the A axis and occurs when switches are swapped around. It is also resetting the home-off int the home screen. Is the reset an indication of bounce?
-
The debounce seems to have fixed the rest issue. However the home-off remains off. Not by much, but still off. The X axis is on, the A is off.
-
Are you looking at machine coords when reading the Home Off value?
Hood
-
Yes, positively. I did make an error in reporting. It is the A axis which has an issue. All was fine and then we changed stepper drives. That is when the issue surfaced. Just t to eliminate the drive as the ghost we re-installed the old drives . . . same issue. We swapped switches . . . same issue. We swapped computers . . . same issue. We rerouted the BOB . . . same issue.
The debounce was set at zero and changing it to 2000 seems to have solved the RESET issue. However, the A axis home-off is in error and the field in the home screen changes to match the error number. For example, if 1.00 is set and the resulting offset is .950 then the 1.00 is automatically changed in the home screen to .950 inches
We have a number of scripts that rely on machine zero to sync with other offset movements. Therefor it is important that all zeros actually represent zero.
-
If you attach your xml I will take a look and see if I can find the problem.
Hood
-
Thanks,
It will be several hours. The Sun isn't due in for a bit.
-
Attached is the xml
-
Seems to be working fine here.
I have tried two ways to simulat it, first I set up hot keys to simulate the home switch inputs and used them to simulate the home inputs and when viewing machine coords the X homed to 2.1800 and the A to 2.1711.
I then disabled the home switches for X and A and then jogged the axes and again pressed RefAll and again they were set to the above values.
So afraid I am not seeing your problem here and dont know what to suggest.
Hood
-
I see an issue or at least from here. Both home-offs were set to 2.180. The 2.171 is the result of the software setting itself after homing. The xml you have is from after the "auto" reset. The difference has been as much as 7.00"
Would please set A to 0.070 and see what occurs. In either case both .070 and 2.180 produce a re-write to the home screen when run on two different computers. The current error of 8 thou doesn't create a problem, but the fact that the software is creating settings worries me.
Thanks for your help!!
-
Ok I set to 0.07 and it set the value to 0.0661.
Now that was strange but I decided to try and see why it was doing such a thing. I opened the screenset in Screen4 and made the A DRO show 16 decimals. The value when homed is showing as 0.0660770833184247.
Ok so next I looked at your A axis steps per unit and they work out as 0.0094395833312035 degrees per step.
So if you divide your DRO value by the resolution (degree per step) then it comes out to 7, or almost. I suspect if I expanded the DRO to show more decimals it would be exactly 7 times your min step.
So what seems to be happening is Mach is saying to itself it can not possibly move to 0.070 degrees as the steps per unit wont allow it. Now of course Mach is not actually moving to that position, it is just setting the value of the Home position, it however means that it can not physically move to G53A0 if it accepts the 0.070 value you entered and thus is setting it to the nearest your steps per degree allow.
Hood
-
Well done! Thank you!! That explains everything. The issue surfaced when we installed new drives. The steps were well off, but our first move was to the home switch, hence the larger error. As we dialed in the steps the number settled at about 7 times by doing the head math on the fly with larger offsets, but it wasn't exactly so we 7.
Thank you again. When the machine is back up I will confirm your findings and let you know. I an confident you are correct.
-
Your conclusion proves spot on! The newly installed drives have more settings than the Cat Detector Van has aerials. This is our first programmable drive. What an improvement. We installed them with the default settings and thus the issue. On axis has a 20:1 screw and the other is nearly direct drive. Thus one axis worked out of the box and the other did not. When the drive resolution is changed the error changes proportionally. Both offsets are now zero-zero-zero. Thanks again!!!
-
Good you got things solved :)