Printable View
Everyone should know that Greg has a second video on YouTube that shows his vacuum clamping setup. The vac stuff is very interesting in it own right, and in that short video, he uses his autozero program, showing just how easy it is to use.
http://youtube.com/watch?v=U9rqj96b4XE
I have just got round to making my autozero plate by following Greg's instructions. I now wonder why I took so long to get around to doing it.
Thanks Greg.
OK got zeroplate built. Got modified screen installed now my probleme is i need to set to inches not mnetric
Code "G31Z-20 F100" 'probing move, can set the feed rate here as well as how far to move---is metric right for inches would i change it to this
Code "G31Z-20.0 F50.0" 'probing move, can set the feed rate here as well as how far to move
To make it the same but in inches.
This is a probe move of one inch and at 4 inches per min feed rate.
Code "G31Z-1 F4" 'probing move, set the feed rate here as well as how far to move
Also further down you need to change the retract distance, this is one inch
Code "G0 Z1" 'put the Z retract height you want here
Greg
Just further to the issue of overrun on zero devices chipping tools.
It occurred to me that some people may have much slower acceleration settings than I have, on their Z axis.
This would mean that when the G31 probe move ends on probe signal (tool touches the plate) the axis needs to decelerate at the set rate. If this is slow it will overrun.
I have my Z axis set to approx 0.1 G's and probe feed at 100mm/min so the stop is virtually instant.
If your Z axis is a bit underpowered or machine is not so rigid and you need slow acceleration settings then make sure your probe speed is not too fast.
Don't have probe feed rate set at 20 or 30 ipm. 4 ipm would be more appropriate.
First jog the Z down close to the target surface (1/4") then run the Auto Tool Zero. No need to start 8 or 10 inches in the air.
Hope this helps, Greg
Greg, et-al,
I just upgraded my Mach3 software and now my z-zero routine won't work. Here's what happening. The probe works fine, as I can touch the router bit and the screen lights up as desired. When I run the program, the zaxis moves downward about 7-20 thousands of an inch then it stops and retracts. Like it sees the probe touch. Of course the probe didn't touch the bit.
Here's the code I'm using. Units are inches.
I hope you can offer a suggestion.
Dave
---------------------
PlateThickness = GetUserDRO(1151) 'Z-plate thickness DRO
If GetOemLed (825)=0 Then 'Check to see if the probe is already grounded or faulty
DoOEMButton (1010) 'zero the Z axis so the probe move will start from here
Code "G4 P5" ' this delay gives me time to get from computer to hold probe in place
Code "G31Z-1 F2" 'probing move, can set the feed rate here as well as how far to move
While IsMoving() 'wait while it happens
Wend
ZProbePos = GetVar(2002) 'get the axact point the probe was hit
Code "G0 Z" &ZProbePos 'go back to that point, always a very small amount of overrun
While IsMoving ()
Wend
Call SetDro (2, PlateThickness) 'set the Z axis DRO to whatever is set as plate thickness
Code "G4 P0.25" 'Pause for Dro to update.
Code "G0 Z1" 'Go to the Z retract height you want here
Code "(Z axis is now zeroed)" 'puts this message in the status bar
Else
Code "(Z-Plate is grounded, check connection and try again)" 'this goes in the status bar if applicable
Exit Sub
End If
Dave
I can't see anything odd about the code.
Does it stop at a consistent spot?
Sounds a bit like "noise" triggering it. Has anything electrical been changed?
It is possible that Art has made some change in the latest release that has affected this. It has happened before.
If your sure it is not noise producing a false trigger then I would go back to the older release and see if that fixes it.
Greg
I"m running version 2.6. How about you?
The touch light on your screen works perfectly. When I touch the probe to the bit, I get a light. Otherwise, I don't get any indication of a spurious touch according to the touch plate light you prove in your screen setup. Maybe a stray signal is so fast that the light doesn't register. I wonder if that's possible?
Dave
Can some of you other folks out there check your version and let me know if you have R2.60?
Thanks,
Dave
Yeah 2.6 here too.
If it is noise you wont see it on the LED. Too short.
The fact that the problem only started now is strange but I would still suspect noise.
Do you have a pullup on the input? I can't get away without it but every setup is different.
Greg
Greg, does increasing debounce help in this instance?
I put the volt meter on the BreakOutBoard test points and measured 4.1 volts. I have a resistor of unknown rating between the 5volt power and the BOB Probe port. The resistor must be reducing the voltage. I forgot why I need that resistor.
I followed the probe lead and separated it from all the other wires in the area. Still no go. At one point I did get a screen indication of the probe grounding out, with no apparent ground contact. So I've got a problem, and it's not Mach3.
I turned off the motor driver power, and ran the mach3 script. This test proved a success, in that the z movement ran the entire 1 inch in the negative direction. The probe even worked in that test. So there's something about the 36 volt power to the drivers that is causing the problem. Without that power, the probe is not getting the noise.
Tell me more about that debounce.
Not sure if it works for probing, but try putting 5000 in the debounce interval box in the general config page, top right corner.Quote:
Originally Posted by Glidergider
About three months ago Art decoupled the probe from the debounce.
So from that version on it has no effect on G31
Greg
That would explain it! I had debounce already set and prior to the recent mach3 upgrade, it work perfectly. After the upgrade, it hasn't worked, not even once.Quote:
Originally Posted by Greolt
Who do I ask about eliminating noise? I wonder if I put a capacitor in parallel with the resistor? Remember, I have a pullup resistor on the probe input.
Dave here is the entry relating to probe debounce in the Mach revisions list.
Did you previously have Ver 2.4 or older?
Greg
Quote:
Aug 1st 2007 V2.4 ( development update)
Fixes to Plugin variables with SDK. Fix to slave homing selection
Fix to debounce. Removed probe from debounce code.
No what you can try is a capacitor between the input and 0 volts (or ground)Quote:
Originally Posted by Glidergider
I would try 0.01uf. Just remember this slows down the probe response. Probably only the same as debounce was doing.
With your 1"/min probe rate it will most likely be fine.
Greg
Greg,
I don't see that cap helping me. Putting a cap between 0v and the probe? My router is my ground, so in my instance, the probe never touches ground. The stray signals trigger a command before the router and probe touch.
I'm not an EE, so this electrical stuff is a little foreign to me. I don't mean to be dense.
In the ports and pins, I have the probe set as "active low"
I was wondering if the resistor I'm using is too much. Since the voltage drop at the probe pin is just 4 volts, that's pretty close to the trigger voltage, which I'm guessing is 2.5 volts.
Regarding my M3 previous version, its been a real long time since I last updated my Mach3. I'm guessing it was May 07 since the last update. I'm sure its not after August.
Hey, wait a minute, I think I might know where you are going with the cap to ground. I forgot that a capacitor doesn't actually transmit any current. So a cap between the probe input and bus ground would be invisible in terms of current and voltage drop at the pin. Yet, the fluctuations in voltage would react to ground. Cool.
Hi GG,
Been a very long time, but I'm finally done with my Solslyva buid. I ran into the same type of problem for my limit/home switches because I needed to have pretty long leads from my BOB to the machine. Noise was so bad that I couldn't even reset Mach3 to allow enough time to give the motors a whirl. The noise was occuring so fast that the LED's on the diagnostic page didn't even light up. The only way to get it work "properly" was to set the debounce to 5000! That's almost a full second of noise filtration - no good long term. After reading several solutions posted by others with similar problems I tried soldering a 0.01uF cap across the leads at the BOB. It worked so well that I was able to reset the debounce back to 0. So, you might just start with playing with the debounce level to show that it is noise and if it is then give the capacitor a shot. By the way, did you need to use a second wire to GND as part of your zeroing system? I wasn't so lucky so I have to attach a wire to the bit to complete the circuit. If you didn't then maybe that might be another experiment you can try.
Good luck and best regards,
Mike
Whoops! I should have backed up and read a few posts before posting. Sorry for being redundent.
Best regards,
Mike