Center Spot Sheetcam Post

thefabricator03 wrote:

rodw wrote:

thefabricator03 wrote: I have my setting in plasmac set at 400 milliseconds.

What happens at 999?

Ok tried 999 and G4 P10, works

999 G4 P3 works but the P10 created a better mark.

Strange you need to wait so long. You could back off slowly from 999 until it does not fire and then go up a bit. I did do a bit of trial and error.

eg. go to 500 and if its not a goer, go to 750, then keep bracketing until its as short as you can get away with.

Yeah Ill give it a try in the morning. Its funny I tried 999 with my gcode file I created and had no luck,

Phill, what exactly did you do to the file I posted?

I only added those G4 Pn lines

Ok, Ill give it a go without them tomorrow and see how it goes on my machine.

Well, I thought I'd post this pic which shows front and back of the donuts with the spotting function and front and backs of a part spotted with my normal wait until ArcOk method. Front sides are on the left.



The holes are much more evenly shaped (eg more circular).

When thinking about it, simply turning on torch for a fixed period is too simplistic because some relays in use will be very slow. I think some opto isolated relays have a 600 ms on delay. So the delay needs to commence from when the arc has been sensed by the THCAD. Therefore, this feature can't be used by Promas etc and will require a THCAD.

It might be able to be offloaded to my ohmic sensing component if you wanted to keep Plasmac more generic

I could do an experimental plasmac component that works on voltage sensing and uses Ark OK as a failsafe if you guys are willing to test it. It won't affect anything on an existing config, the only difference is that the Torch On Time would then be volts. If it works then folk with Promas could use it by ad justing the low level of the Arc OK.

Sounds like something worth trying, Let me know when you need it tested.

phillc54 wrote: I could do an experimental plasmac component that works on voltage sensing and uses Ark OK as a failsafe if you guys are willing to test it. It won't affect anything on an existing config, the only difference is that the Torch On Time would then be volts. If it works then folk with Promas could use it by ad justing the low level of the Arc OK.

To late at night to think hard! Willing to test, but could we also be able to disable the feature maybe by setting an external plasmac.spot-disable pin? And call the GUI field "Spot Threshold" so it could be volts or milliseconds

I am thinking that my ohmic sense component knows when the Arc is on (> 24.5 volts). So if it received spindle.2.on, it could wait until it had an arc and then count down the delay and turn it off. It would just need to have an output or2'ed with the torch on relay output. I can do that by unlinking pins in hal so it does not affect any Plasmac defaults. A proma user who chose to use a THCAD for ohmic sensing could still use the timer feature....

So rephrasing;
a seperate component called spot-delay receives as input:
Spindle.2.on
Arc_Is_On
Spot-Time-Delay

And outputs:
Torch-On

Torch-on goes true when spindle.2.on is true and turns off Spot-Time-Delay after Arc-Is-On goes true.
Torch-On is or2'ed with the plasmac torch on output
Spot_Time_Delay comes from the Plasmac GUI Spot Threshold.

Actually, thinking about Stefan's problem today, I wonder if the problem was the F5 turned the torch off before the delay was over becasue the programmed move between M3 and F5 was quicker than the pause.

What if you did a tiny move on the Z axis after M3 like:
G1 Z -0.008 F1

That move will take 0.5 seconds (500 ms) to complete so it the spot would be over by the time it finished....