Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: Davek0974 on June 06, 2020, 10:41:22 AM

Title: Probe debounce...
Post by: Davek0974 on June 06, 2020, 10:41:22 AM
Hi all, been a few years since i was here :)

Needed to make a new surface probe for my plasma table, got the mechanics done but seen to have a little issue - if someone can confirm my diagnosis and then offer some help would one good :)

ISSUE:-
When running a Z axis probe, the switch will open and Mach3 wills sense the material but the axis does not appear to pull away from the material until the switch opens again.

DIAGNOSIS:-
Switch bounce - its closing and opening very rapidly i.e. bouncing and Mach3 senses it as having pulled away.

FIX:-
Adjust the debounce ion general config, trouble is there are two debounce settings - is this "Debounce" or "Index Debounce"

Thanks all

Dave


Title: Re: Probe debounce...
Post by: BR549 on June 06, 2020, 09:51:31 PM
A Mach3 probe does NOT automatically back away from the trip point . You have to program the macro to do so.

What does your Macro code look like ? IF you think Debonce will help then you use the Debounce setting.



Just a thought , (;-) TP
Title: Re: Probe debounce...
Post by: Davek0974 on June 07, 2020, 01:31:48 AM
It runs as a subroutine called from the main code before each plasma piercing.

Code: [Select]
G91 (switch to incremental mode)
G01 X24.00 Y2.00 F6000 (position probing point)
G90 (switch to absolute mode)
F2000
M1050 (trigger the probe cylinder)
G04 P0.3
G28.1 Z10.00 (probe the surface)
G92 Z#15045 (apply the probe switch offset from settings screen DRO)
M1051 (raise the probe)
G91 (switch to incremental mode)
G01 X-24.00 Y-2.00 F6000 (put the axes back to where we came from)
G90 (switch to absolute mode)
M99

I cant see why anything else would change - the subroutine has been running for a few years now, all i did was build a new mechanism and fit a new microswitch, tried two. It seems it sometimes backs away but not always??

The probe is offset from the torch so it needs the axes moving as in the sub.
Title: Re: Probe debounce...
Post by: TPS on June 07, 2020, 03:33:04 AM
your macrocode M1050 and M1051 would be interesting.
Title: Re: Probe debounce...
Post by: Davek0974 on June 07, 2020, 03:36:28 AM
M1050 =
ActivateSignal(OutPut3)
Sleep(300)

M1051 =
DeActivateSignal(OutPut3)
Sleep(400)
Title: Re: Probe debounce...
Post by: TPS on June 07, 2020, 03:45:29 AM
ok nothing Spezial there.

you are doing the probe with G28.1 (referencing) i see.

have you tryed a Ref Z in the diagostic page, witch should prizipial do the same Thing, just to verify.

before you start the probe/reference is the M3Home LED on diagnostic page on or off?
Title: Re: Probe debounce...
Post by: Davek0974 on June 07, 2020, 03:49:15 AM
Not tried the ref z yet, will do that shortly.

The Home LED is off before the routine runs.

I'll be going out to play shortly, been a very long time since i touched this code, machine just runs and lately has been getting hammered hence the new probe build :)

Brings back lots of memories and headaches :) :)
Title: Re: Probe debounce...
Post by: Davek0974 on June 07, 2020, 06:00:40 AM
I think i have it now.

Probing speed was the issue - this particular microswitch does not like approaching trigger point slowly, after messing with debounce and macros i turned up the speed for the G28.1 setting in homing/limits.

It was on 20% now on 50% and accuracy seems to have returned, plus i get the added bonus of faster probing :)

I guess the switch internals are a little grey and not as good as the old one i took off.

After many dozens of runs and one test cut job i am satisfied that she'll be working ok Monday now.

Thanks for the tips guys.
Title: Re: Probe debounce...
Post by: Davek0974 on June 13, 2020, 09:44:25 AM
Even better, fitted a longer air cylinder today, i can now get the probe tip way up out of the path of the plasma muck, I have also increased the speed to 100% rapid and accuracy is amazing - ran over 50 cycles with a DTI on the Z axis and it maintained +/- 0.05mm each time - plenty good enough for plasma work and very good for a common microswitch.

My earlier post was a little off  i am not probing - its using a ref-z instead and the ref process does need to back-off from the switch closure.
Title: Re: Probe debounce...
Post by: BR549 on June 14, 2020, 07:59:45 PM
I remember back when you were working on it.

(;-) TP
Title: Re: Probe debounce...
Post by: Davek0974 on June 15, 2020, 01:34:19 AM
Hiya, yes it was a blast from the past, seeing all the familiar names pop up in searches :)

She's been working well over the years, lately doing full days in production since i retired last year.

Hard to beat Mach 3 for a bit of customisation ;)

Hope you're keeping well.
Title: Re: Probe debounce...
Post by: BR549 on June 15, 2020, 11:27:25 AM
Well it is good that you finally pulled the plug and got a real job (;-)

(;-) TP
Title: Re: Probe debounce...
Post by: Davek0974 on June 15, 2020, 11:33:54 AM
Well it is good that you finally pulled the plug and got a real job (;-)

(;-) TP

LOL, it was supposed to be an easy top-up for the early pension/retirement - i'm doing more work now than i was in full-time work :)

Its good though, nice not having to drive to work etc ;)