I've read somewhere that Mach3 Plasma requires the switching on and off of a plasma torch with:

DoSpinCW()

and

DoSpinStop()

It was said that if you turn on the plasma without these commands, such as:

ActivateSignal(OUTPUT1)

then the behaviour of the machine will be incorrect ?? (did not say in what way it will be incorrect).

I know that Mach3 plasma will release motion if the THC is not switched on, but when THC is switched on it will not release motion until it gets an ARC OK signal. So assuming THC is switched ON:

Does anyone know if the DoSpinCW() command is tied to Mach3 releasing motion ONLY when it gets the ARC OK signal.

Put another way, if ActivateSignal(OUTPUT1) was used instead, would Mach3 release motion immediately at the first XY move regardless of whether the ARC OK signal is present.

Or is that last sentence untrue, and Mach3 Plasma will not release motion until it gets ARC OK, regardless of what command fired the torch.

Cheers,

Keith

I had very odd behaviour using ActivateSignal until I also used ArcOk then it all worked nicely and much faster.

What torch height controllers are you using?

If you have a THC that controls the z directly then that may be the case.... But ... Consider the sequence of events

Touch off, reset z coordinate
Retract to pierce height
Fire torch
Wait duration for pierce
Drop to z cut height and begin to move x and Y (or your hole will get bigger)
THC OK signal received once arc stabalised (defined as being plus or minus so many volts of the target voltage)
Then allow automated up and down motion once THC OK signal is within hysterysis of target voltage.
If the voltage spikes up, then you may be going over a kerf, so THCOK signal drops out (above hysterysis or rate of change greater than expected for blended tradjectory) and THCDn and THCUp are ignored.

Are you saying the x and Y should not move until the THCOK signal is received?.... If that is the case you could have blown a bigger hole in the plate than you need... Hence the THCOK signal will be wrong as the hole is oversize?

The activatesignal is a new one to me but I have not been doing it long at all... Seems strange to me though as the second last tab on the ports and pins settings controls which output is switched with m3 and m4... So cannot see what the difference is other than dospin probably controls the spindle settings if you have them setup... So if for example you were using a plasma that you could control the ampage on (hyperthem 65 or above or older bigger machines)... Then I guess you could motion control the arc ampage and gas pressure based on blended tradjectory of the x and Y motion.

Ie slow down for a corner... Not a problem with anti-dive, back off on the ampage and gas pressure a bit to maintain the torch voltage relative to the blended tradjectory.

See how this one pans out from the more knowledgable folk, but it would be nice to know which THC setup you are using.

Oh yeah, I missed out on one phase...

...fire torch...
ARC TRANSFER on some thcs...they may monitor the acr transfer and then start the counter for the pierce ....
...then pierce delay timer...
...then x Y motion move...
But this depends on the THC as all Mach is looking for is THCOK, THCUp and THCDn... If using the parallel port or ESS in order to control the z axis.

If your machine is setup to send a signal back to mach3 over whether the arc has transferred before starting the pierce counter of initiating motion, you could be creating a bigger pierce hole than nessesary because the loop speed of mach for inputs via macropump and brains is not very quick (10hz so 10 times per second) and the arc would have probably transfered quicker than that as some of the pierce times in the hyperthem manual recommended cutting speeds are near zero on thin plate.

If you have a stand alone THC with internal pierce timer function then it may use the above sort of strategy... But not if you are running stuff via mach3.

Mach4 offers some really great potential because of the loop speed, I'd just like someone to have a go at showing how they did it or an out of the box solution (screenset) from NFS... Plus need the new ESS plugin for m4...

He is using a teh Candcnc THC control. That THC only controls THC up /down, Ark ok and  and Gap inhibit of teh THC/down. Mach3 controls teh piere delay and uses the ark ok signal to start motion.

IF it is stalling on teh Arc ok signal starting motion something is a miss in Mach3.

Check that teh Gcode is NOT issueing a dwell on pierce.

Check your M3 macro for the proper start code for the torch.

Check your Spindles config to make sure there is NO spinup delay .  Reset teh value to ZERO and then resave and restart Mach3

 Then if that did not work the next thing I would do is repalce teh XML with a known good backup OR rebuild a new XML from stratch.

(;-)  TP

Thanks guys.

No dwell in gcode.

Spindle spinups all at zero.

I'm actually going to uninstall & reinstall Mach3 because it seems to have gone nuts on me during this little cut. So that will give a fresh install then I only have to check the command in the M3.

Candcnc told me there's "a bug in Mach3" that causes their plugin to "fight" with Machs THC DELAY and convert it to a pause in motion 

I think their more modern systems have put a lot of functions into the electronics because the plugin and Mach3 just didn't do what was intended.

Keith

YEP with CandCNC you have to use either one or the  other delays you cannot mix the two. BUT it has always been that way . WHAT did you change to cause a problem again ?? What did you do the first time you saw this problem to fix it.

Could be a corrupt XML mixing values that you cannot see.

(;-) TP

It never really got fixed Terry. I've hardly used my table the last couple of years and when I have it's been on thicker steel with a leadin so the delay in motion was just a slight nuisance during a pierce.

This time I've been given 2mm metal and I've nested the cut so tight I was using edge starts (won't be doing anything fancy like that again). Of course the motion delay just blows a big hole at the edge.

The other day I never got to the point of putting one of the delays to zero and testing that because Mach3 seems to also have gone wonky in the process. I've got a day off today so I'll re-install Mach3 and the MP3000 software and put the Mach3 THC DELAY to zero and do some test pierces.

Keith.