QTPlasmac pierce delay handling

Hi,

I just got QTPlasmac (2.10 using buildbot instructions) going and starting to make some test cuts.  My plasma cutter (Thermal Dynamics Cutmaster 60i) has a ~2 second delay for gas purge before the pilot starts.  That means I have to add whatever actual pierce delay I need to the two seconds making it, for example, something like 2.5 seconds. After the cut is made there is post flow for 20 seconds and if you start a cut while post flow is active there is no delay and the pilot lights immediately making the pierce delay too long.  When the post flow stops after 20 seconds then the 2 second delay is back.  I know I could have two different materials... one for first cut and one for additional cuts and have my Fusion 360 post processor change materials but that seems like a bit of a hack.  Is there a way I could handle this in hal/script/component?

Is ARCOK signal wired to plasma or derived from THCAD?

ARCOK is derived from THCAD. My unit has arc ok output but the relay on the inverter board is not working so I'm using mode 0 for now. ark-ok-low is 60 and ark-ok-high is 200. Pilot voltage seems to be around 157 so maybe I could set ark-ok-high to 145 or 150. I've been testing on a scrap piece of 1/2" I had so not sure if the pilot voltage changes based on material thickness.

See what voltage you get when cutting and how much it fluctuates, then set the ARC low and high to about 20-30V above and below.
Do a lot of tests till you get it to have cuts without interruption and no false OK signal at start.
And yes, the voltage will differ with thickness and feed rate, but usually under 10V difference on all of my machines (over 15 machines) for materials from 1.8mm to 6mm.

The pierce delay starts from when qtplasmac gets an ArcOK signal signifying that arc transfer has occurred. If your synthesized arcOK is off, you will have problems.

Do yourself a favour and wire in the machine's arcOK signal. The time you will save stuffing around trying to get a synthesized arcOK working will more than cover the time you will spend wiring it in properly.

Unfortunately the relay in the unit is damaged and does not close. I ran through the diagnostics in the service manual to come to that conclusion and of course the service manual recommendation is to replace the inverter board which is half the cost of a new unit lol. It's also very difficult to access but some point down the road I will pull it apart and replace the relay on the board if possible. I'm going to try Tommy's suggestions and hopefully that will buy some time until I fix it or sell it and get another unit.

Get a "reed contact" or "reed relay", make some 3 turns of the clamp wire around it.
That is it, cheap simple and never fails.
I am sure it was in the PlasmaC documentation, no idea if it is in QtPlasmaC.
See attached.

kworm wrote:

Unfortunately the relay in the unit is damaged and does not close. I ran through the diagnostics in the service manual to come to that conclusion and of course the service manual recommendation is to replace the inverter board which is half the cost of a new unit lol. It's also very difficult to access but some point down the road I will pull it apart and replace the relay on the board if possible. I'm going to try Tommy's suggestions and hopefully that will buy some time until I fix it or sell it and get another unit.

the most likely cause of a "damaged" relay is oxidisation because the current through the relay has  been below the minimum operating spec. Installing a HS10 1 K pulldown resistor is likely to fix the issue after a few operations once the oxide is burnt off. This is covered in the qtplasmac docs.

Aforementioned docs references:

Reed relay:

linuxcnc.org/docs/devel/html/plasma/qtpl...l#plasma:reed-arc-ok

Contact loading schematic:

linuxcnc.org/docs/devel/html/plasma/qtpl...tact-load-schematics

Calculations:

linuxcnc.org/docs/devel/html/plasma/qtpl...asma:qt-contact-load

Thanks everyone. I'll give the resistor a try even though the unit fails the manufacturers test. If that doesn't work I'll use a reed relay until I fix or sell it.