TLO measurement issue. Tool crashing into TCP touch plate.

[sebba]

I'll recheck section 8.11 ASAP and also I'll play again with some debounce settings.

Thanks for your fast reply

[sebba]

I raised the debounce for probing from 50 to 150. Keep testing but no errors so far.
Thanks

[DaveCVI]

it depends on how much testing you want to do - but ideally you want the debounce values as low as they can be and still work. Remember that doing this in software means we are ignoring the input signal for some amount of time - that causes the probe to continue moving during the debouce time - which results in overrun past the real trigger point. The amount of travel depends on the probing speed - but this can also result if damaged probes (particularly for Z axis probe motion).

The best approach is to clean up the probe input signal in the electronics so that software debounce is not required.

I realize that most people find a software debounce amount that works, add a tad for margin and forget about it. Just be aware of what you are trading off if you go that route.
Dave

[sebba]

Yes, I agree, I will try to lower it at much as possible but breaking tools is an expensive hobby and also a time consuming one.
I will try to design a Z-sensor, for measuring tools, with an E-Stop included function if the machine will try to go down more than usual.
At least, I'll not break more bits...

[DaveCVI]

Hi,
If the approach you decide on for a Z-sensor involves that sensor being spring loaded, please review this thread
http://www.calypsoventures.com/forums/v ... ?f=6&t=177
before settling on the sensor design.

Dave