What was the default index debounce setting.

[Derek]

I noticed that when I upgraded to the version that had selectable debounce settings that the default was zero. both of my mills stopped ridged tapping. I set the index debounce to 1 and they started working. The Bridgeport just had a problem seeing the index signal on a hole. I restarted the program and it worked fine. I'm curious as to what the debounce figure was before this feature was added.

Thanks
Derek

[cncdrive]

There was no debouncing like that before. Debounce was only the electrtical RC filter which is on the board, but it gives only a few microseconds of debounce.
I think the debounce is not applied on encoder signals, I mean if I recall encoder signals are excluded from the debounce function, because they are short (usually high freq.) signals, so even 1msec long debounce could mess that up.
Could you please post a profile file in the state when the rigid tapping is not working?
Then I will ask my collegue to debug with that profile to see what the issue exactly is.

[cncdrive]

Hi Derek,

My collegue just finished debugging the issue you reported here.
We have used your profile which you posted earlier here: viewtopic.php?f=3&t=882&p=6723#p6619
What we have checked is if the controller is doing the G33.1 rigid tapping moves with the debounce = 0 setting as in your profile.
And it works without problem, the tapping works OK here with the 0 debounce.
My collegue also confirmed that the debounce is not applied on encoder signals, they are excluded from the debouncing, because they are fast signals, so even 1msec debounce would be too much.

So, we do not see the problem yet, but maybe you ran this with another profile which we do not have?
If so can you please post that profile file with which the tapping is not working?

[Derek]

Hi Balazs

I use a custom screen Terry made for me. Often times UCCNC will have new features when I update but the controls are not on the screen. Case in point is the debounce settings. Terry sent me a new screen set that had them. The values were all zeros. I fired up the mill and coincidently the first thing I wanted to do was check a tapping program. It would start the cycle but just sit there waiting for an index signal. Tried it a few times. Keep in mind the day before I tapped 384 holes which is my normal production day with this machine. I had remembered seeing the debounce settings on the new screen so I set index to 1 just to see if it made a difference. I didn't restart or anything. I hit run and the machine tapped perfectly.It hasn't messed up since. Ran into the same issue with the Bridgeport when I updated it.

I think I have a before and after update on the Alliant profile. I'll see if I can put that together for you.
Thanks
Derek

[cncdrive]

Hi Derek,

Yes, what you described is basicly the same what we tested here.
The only thing is missing for us is your up to date profile file. So, we used the profile file which you posted earlier, but I'm not sure which of your machine it was for, if the issue happened with that profile or with a different one.

So, what we tested with that profile is that with 0 debounce we have tested if the controller is recognising the index signal and if it is doing the tapping cycle properly and it does, it recognised the index and did the tapping cycles. This is why I said that we could not reproduce, we do not see the issue yet. I mean we understood your description, just cannot reproduce the issue here.
So, maybe you can post that profile file with which the issue is happening at you?
Maybe it is pin dependent and if you running a different profile with the encoder configured to different pins might be the reason that you see a problem and we do not, or it could be any other settings which we have no idea about what it could be. So, seeing and trying with that profile could be helpful.

We also checked if the debounce influances the encoder signals A, B or index and it does not, the controller reads these signals directly from the input pins, the debounce is not applied,
but maybe there is a mistake in the code for some pins for the index which we could not locate so far, but seeing your profile could probably help, because then we will at least know what pins to exactly check and test with...