Joint2 following error, during homing switch reports on, 7i97

If its the index bug, speeding up the final homing will fix it.

(this does depend good tuning though so the actual motion doesn't slow below
1 count/servo thread)

Note: 2.8.4 definitely  has the index bug as it was released 2 months before the fix


 

ok, I'll update and see what happens

so, get this. I am not sure exactly where to report this bug, or if it even is a bug, but when the machine faults out from a failure to detect that the switch is active, IE the switch is having a issue with the way it is wired versus the 7i77, when the machine faults, it shows the input as being on. That's kind of a deceptive problem to chase.So basically the switch does not properly trigger the input " this is being remedied" when the joint faults out with  a motion error, the input lights up as active " switch is being read as active high 24v" 

That's a really bizarre behavior and very deceptive. 

I'm not sure what you are saying.

7I77 inputs should always represent the current
input status regardless of any motion faults.

The only case where they may not represent
the inputs faithfully is if communications
are lost (though this should be instantly reported)

I can 100% reproduce this behavior, IF the software is triggered into a motion fault state, with a following error, the switch is reporting active, with no change in state. all you have to do is create a following error during homing on the Z " all I have tested so far"

sequence of events

Set Z axis to homing, 
Z axis moves to home
Index counts go up
Machine reached end of mechanical travel
Following error faults machine into estop or other error mode
Input 8 is reported as High/On

I even disconnected the input terminal block to the card. This should be a bug, and a very easy to reproduce one at that. In the first page of the thread, I linked the youtube video showing this exact behavior. I thought the switch was working because it was reporting in Hal when I watched it, but that was a bad assumption. 

Can you post your hal/ini files?

 

on first page, I haven't changed them at all

OK so this makes sense now:

#net max-z <= [HMOT](CARD0).inmux.00.input-09
net min-home-z <= [HMOT](CARD0).gpio.008.in

You have  min-home-z connected to gpio.008 rather than
[HMOT](CARD0).inmux.00.input-08

gpio.008 is active low PWMENA so that would go high on a fault

the state shouldn't change in any situation ever. unless the switch state changes. this is a bug.